##// END OF EJS Templates
moved getdefaultencoding from text to py3compat
Brandon Parsons -
Show More
@@ -1,702 +1,702 b''
1 """A simple configuration system.
1 """A simple configuration system.
2
2
3 Authors
3 Authors
4 -------
4 -------
5 * Brian Granger
5 * Brian Granger
6 * Fernando Perez
6 * Fernando Perez
7 * Min RK
7 * Min RK
8 """
8 """
9
9
10 #-----------------------------------------------------------------------------
10 #-----------------------------------------------------------------------------
11 # Copyright (C) 2008-2011 The IPython Development Team
11 # Copyright (C) 2008-2011 The IPython Development Team
12 #
12 #
13 # Distributed under the terms of the BSD License. The full license is in
13 # Distributed under the terms of the BSD License. The full license is in
14 # the file COPYING, distributed as part of this software.
14 # the file COPYING, distributed as part of this software.
15 #-----------------------------------------------------------------------------
15 #-----------------------------------------------------------------------------
16
16
17 #-----------------------------------------------------------------------------
17 #-----------------------------------------------------------------------------
18 # Imports
18 # Imports
19 #-----------------------------------------------------------------------------
19 #-----------------------------------------------------------------------------
20
20
21 import __builtin__ as builtin_mod
21 import __builtin__ as builtin_mod
22 import os
22 import os
23 import re
23 import re
24 import sys
24 import sys
25
25
26 from IPython.external import argparse
26 from IPython.external import argparse
27 from IPython.utils.path import filefind, get_ipython_dir
27 from IPython.utils.path import filefind, get_ipython_dir
28 from IPython.utils import py3compat, text, warn
28 from IPython.utils import py3compat, text, warn
29
29
30 #-----------------------------------------------------------------------------
30 #-----------------------------------------------------------------------------
31 # Exceptions
31 # Exceptions
32 #-----------------------------------------------------------------------------
32 #-----------------------------------------------------------------------------
33
33
34
34
35 class ConfigError(Exception):
35 class ConfigError(Exception):
36 pass
36 pass
37
37
38 class ConfigLoaderError(ConfigError):
38 class ConfigLoaderError(ConfigError):
39 pass
39 pass
40
40
41 class ConfigFileNotFound(ConfigError):
41 class ConfigFileNotFound(ConfigError):
42 pass
42 pass
43
43
44 class ArgumentError(ConfigLoaderError):
44 class ArgumentError(ConfigLoaderError):
45 pass
45 pass
46
46
47 #-----------------------------------------------------------------------------
47 #-----------------------------------------------------------------------------
48 # Argparse fix
48 # Argparse fix
49 #-----------------------------------------------------------------------------
49 #-----------------------------------------------------------------------------
50
50
51 # Unfortunately argparse by default prints help messages to stderr instead of
51 # Unfortunately argparse by default prints help messages to stderr instead of
52 # stdout. This makes it annoying to capture long help screens at the command
52 # stdout. This makes it annoying to capture long help screens at the command
53 # line, since one must know how to pipe stderr, which many users don't know how
53 # line, since one must know how to pipe stderr, which many users don't know how
54 # to do. So we override the print_help method with one that defaults to
54 # to do. So we override the print_help method with one that defaults to
55 # stdout and use our class instead.
55 # stdout and use our class instead.
56
56
57 class ArgumentParser(argparse.ArgumentParser):
57 class ArgumentParser(argparse.ArgumentParser):
58 """Simple argparse subclass that prints help to stdout by default."""
58 """Simple argparse subclass that prints help to stdout by default."""
59
59
60 def print_help(self, file=None):
60 def print_help(self, file=None):
61 if file is None:
61 if file is None:
62 file = sys.stdout
62 file = sys.stdout
63 return super(ArgumentParser, self).print_help(file)
63 return super(ArgumentParser, self).print_help(file)
64
64
65 print_help.__doc__ = argparse.ArgumentParser.print_help.__doc__
65 print_help.__doc__ = argparse.ArgumentParser.print_help.__doc__
66
66
67 #-----------------------------------------------------------------------------
67 #-----------------------------------------------------------------------------
68 # Config class for holding config information
68 # Config class for holding config information
69 #-----------------------------------------------------------------------------
69 #-----------------------------------------------------------------------------
70
70
71
71
72 class Config(dict):
72 class Config(dict):
73 """An attribute based dict that can do smart merges."""
73 """An attribute based dict that can do smart merges."""
74
74
75 def __init__(self, *args, **kwds):
75 def __init__(self, *args, **kwds):
76 dict.__init__(self, *args, **kwds)
76 dict.__init__(self, *args, **kwds)
77 # This sets self.__dict__ = self, but it has to be done this way
77 # This sets self.__dict__ = self, but it has to be done this way
78 # because we are also overriding __setattr__.
78 # because we are also overriding __setattr__.
79 dict.__setattr__(self, '__dict__', self)
79 dict.__setattr__(self, '__dict__', self)
80
80
81 def _merge(self, other):
81 def _merge(self, other):
82 to_update = {}
82 to_update = {}
83 for k, v in other.iteritems():
83 for k, v in other.iteritems():
84 if not self.has_key(k):
84 if not self.has_key(k):
85 to_update[k] = v
85 to_update[k] = v
86 else: # I have this key
86 else: # I have this key
87 if isinstance(v, Config):
87 if isinstance(v, Config):
88 # Recursively merge common sub Configs
88 # Recursively merge common sub Configs
89 self[k]._merge(v)
89 self[k]._merge(v)
90 else:
90 else:
91 # Plain updates for non-Configs
91 # Plain updates for non-Configs
92 to_update[k] = v
92 to_update[k] = v
93
93
94 self.update(to_update)
94 self.update(to_update)
95
95
96 def _is_section_key(self, key):
96 def _is_section_key(self, key):
97 if key[0].upper()==key[0] and not key.startswith('_'):
97 if key[0].upper()==key[0] and not key.startswith('_'):
98 return True
98 return True
99 else:
99 else:
100 return False
100 return False
101
101
102 def __contains__(self, key):
102 def __contains__(self, key):
103 if self._is_section_key(key):
103 if self._is_section_key(key):
104 return True
104 return True
105 else:
105 else:
106 return super(Config, self).__contains__(key)
106 return super(Config, self).__contains__(key)
107 # .has_key is deprecated for dictionaries.
107 # .has_key is deprecated for dictionaries.
108 has_key = __contains__
108 has_key = __contains__
109
109
110 def _has_section(self, key):
110 def _has_section(self, key):
111 if self._is_section_key(key):
111 if self._is_section_key(key):
112 if super(Config, self).__contains__(key):
112 if super(Config, self).__contains__(key):
113 return True
113 return True
114 return False
114 return False
115
115
116 def copy(self):
116 def copy(self):
117 return type(self)(dict.copy(self))
117 return type(self)(dict.copy(self))
118
118
119 def __copy__(self):
119 def __copy__(self):
120 return self.copy()
120 return self.copy()
121
121
122 def __deepcopy__(self, memo):
122 def __deepcopy__(self, memo):
123 import copy
123 import copy
124 return type(self)(copy.deepcopy(self.items()))
124 return type(self)(copy.deepcopy(self.items()))
125
125
126 def __getitem__(self, key):
126 def __getitem__(self, key):
127 # We cannot use directly self._is_section_key, because it triggers
127 # We cannot use directly self._is_section_key, because it triggers
128 # infinite recursion on top of PyPy. Instead, we manually fish the
128 # infinite recursion on top of PyPy. Instead, we manually fish the
129 # bound method.
129 # bound method.
130 is_section_key = self.__class__._is_section_key.__get__(self)
130 is_section_key = self.__class__._is_section_key.__get__(self)
131
131
132 # Because we use this for an exec namespace, we need to delegate
132 # Because we use this for an exec namespace, we need to delegate
133 # the lookup of names in __builtin__ to itself. This means
133 # the lookup of names in __builtin__ to itself. This means
134 # that you can't have section or attribute names that are
134 # that you can't have section or attribute names that are
135 # builtins.
135 # builtins.
136 try:
136 try:
137 return getattr(builtin_mod, key)
137 return getattr(builtin_mod, key)
138 except AttributeError:
138 except AttributeError:
139 pass
139 pass
140 if is_section_key(key):
140 if is_section_key(key):
141 try:
141 try:
142 return dict.__getitem__(self, key)
142 return dict.__getitem__(self, key)
143 except KeyError:
143 except KeyError:
144 c = Config()
144 c = Config()
145 dict.__setitem__(self, key, c)
145 dict.__setitem__(self, key, c)
146 return c
146 return c
147 else:
147 else:
148 return dict.__getitem__(self, key)
148 return dict.__getitem__(self, key)
149
149
150 def __setitem__(self, key, value):
150 def __setitem__(self, key, value):
151 # Don't allow names in __builtin__ to be modified.
151 # Don't allow names in __builtin__ to be modified.
152 if hasattr(builtin_mod, key):
152 if hasattr(builtin_mod, key):
153 raise ConfigError('Config variable names cannot have the same name '
153 raise ConfigError('Config variable names cannot have the same name '
154 'as a Python builtin: %s' % key)
154 'as a Python builtin: %s' % key)
155 if self._is_section_key(key):
155 if self._is_section_key(key):
156 if not isinstance(value, Config):
156 if not isinstance(value, Config):
157 raise ValueError('values whose keys begin with an uppercase '
157 raise ValueError('values whose keys begin with an uppercase '
158 'char must be Config instances: %r, %r' % (key, value))
158 'char must be Config instances: %r, %r' % (key, value))
159 else:
159 else:
160 dict.__setitem__(self, key, value)
160 dict.__setitem__(self, key, value)
161
161
162 def __getattr__(self, key):
162 def __getattr__(self, key):
163 try:
163 try:
164 return self.__getitem__(key)
164 return self.__getitem__(key)
165 except KeyError, e:
165 except KeyError, e:
166 raise AttributeError(e)
166 raise AttributeError(e)
167
167
168 def __setattr__(self, key, value):
168 def __setattr__(self, key, value):
169 try:
169 try:
170 self.__setitem__(key, value)
170 self.__setitem__(key, value)
171 except KeyError, e:
171 except KeyError, e:
172 raise AttributeError(e)
172 raise AttributeError(e)
173
173
174 def __delattr__(self, key):
174 def __delattr__(self, key):
175 try:
175 try:
176 dict.__delitem__(self, key)
176 dict.__delitem__(self, key)
177 except KeyError, e:
177 except KeyError, e:
178 raise AttributeError(e)
178 raise AttributeError(e)
179
179
180
180
181 #-----------------------------------------------------------------------------
181 #-----------------------------------------------------------------------------
182 # Config loading classes
182 # Config loading classes
183 #-----------------------------------------------------------------------------
183 #-----------------------------------------------------------------------------
184
184
185
185
186 class ConfigLoader(object):
186 class ConfigLoader(object):
187 """A object for loading configurations from just about anywhere.
187 """A object for loading configurations from just about anywhere.
188
188
189 The resulting configuration is packaged as a :class:`Struct`.
189 The resulting configuration is packaged as a :class:`Struct`.
190
190
191 Notes
191 Notes
192 -----
192 -----
193 A :class:`ConfigLoader` does one thing: load a config from a source
193 A :class:`ConfigLoader` does one thing: load a config from a source
194 (file, command line arguments) and returns the data as a :class:`Struct`.
194 (file, command line arguments) and returns the data as a :class:`Struct`.
195 There are lots of things that :class:`ConfigLoader` does not do. It does
195 There are lots of things that :class:`ConfigLoader` does not do. It does
196 not implement complex logic for finding config files. It does not handle
196 not implement complex logic for finding config files. It does not handle
197 default values or merge multiple configs. These things need to be
197 default values or merge multiple configs. These things need to be
198 handled elsewhere.
198 handled elsewhere.
199 """
199 """
200
200
201 def __init__(self):
201 def __init__(self):
202 """A base class for config loaders.
202 """A base class for config loaders.
203
203
204 Examples
204 Examples
205 --------
205 --------
206
206
207 >>> cl = ConfigLoader()
207 >>> cl = ConfigLoader()
208 >>> config = cl.load_config()
208 >>> config = cl.load_config()
209 >>> config
209 >>> config
210 {}
210 {}
211 """
211 """
212 self.clear()
212 self.clear()
213
213
214 def clear(self):
214 def clear(self):
215 self.config = Config()
215 self.config = Config()
216
216
217 def load_config(self):
217 def load_config(self):
218 """Load a config from somewhere, return a :class:`Config` instance.
218 """Load a config from somewhere, return a :class:`Config` instance.
219
219
220 Usually, this will cause self.config to be set and then returned.
220 Usually, this will cause self.config to be set and then returned.
221 However, in most cases, :meth:`ConfigLoader.clear` should be called
221 However, in most cases, :meth:`ConfigLoader.clear` should be called
222 to erase any previous state.
222 to erase any previous state.
223 """
223 """
224 self.clear()
224 self.clear()
225 return self.config
225 return self.config
226
226
227
227
228 class FileConfigLoader(ConfigLoader):
228 class FileConfigLoader(ConfigLoader):
229 """A base class for file based configurations.
229 """A base class for file based configurations.
230
230
231 As we add more file based config loaders, the common logic should go
231 As we add more file based config loaders, the common logic should go
232 here.
232 here.
233 """
233 """
234 pass
234 pass
235
235
236
236
237 class PyFileConfigLoader(FileConfigLoader):
237 class PyFileConfigLoader(FileConfigLoader):
238 """A config loader for pure python files.
238 """A config loader for pure python files.
239
239
240 This calls execfile on a plain python file and looks for attributes
240 This calls execfile on a plain python file and looks for attributes
241 that are all caps. These attribute are added to the config Struct.
241 that are all caps. These attribute are added to the config Struct.
242 """
242 """
243
243
244 def __init__(self, filename, path=None):
244 def __init__(self, filename, path=None):
245 """Build a config loader for a filename and path.
245 """Build a config loader for a filename and path.
246
246
247 Parameters
247 Parameters
248 ----------
248 ----------
249 filename : str
249 filename : str
250 The file name of the config file.
250 The file name of the config file.
251 path : str, list, tuple
251 path : str, list, tuple
252 The path to search for the config file on, or a sequence of
252 The path to search for the config file on, or a sequence of
253 paths to try in order.
253 paths to try in order.
254 """
254 """
255 super(PyFileConfigLoader, self).__init__()
255 super(PyFileConfigLoader, self).__init__()
256 self.filename = filename
256 self.filename = filename
257 self.path = path
257 self.path = path
258 self.full_filename = ''
258 self.full_filename = ''
259 self.data = None
259 self.data = None
260
260
261 def load_config(self):
261 def load_config(self):
262 """Load the config from a file and return it as a Struct."""
262 """Load the config from a file and return it as a Struct."""
263 self.clear()
263 self.clear()
264 try:
264 try:
265 self._find_file()
265 self._find_file()
266 except IOError as e:
266 except IOError as e:
267 raise ConfigFileNotFound(str(e))
267 raise ConfigFileNotFound(str(e))
268 self._read_file_as_dict()
268 self._read_file_as_dict()
269 self._convert_to_config()
269 self._convert_to_config()
270 return self.config
270 return self.config
271
271
272 def _find_file(self):
272 def _find_file(self):
273 """Try to find the file by searching the paths."""
273 """Try to find the file by searching the paths."""
274 self.full_filename = filefind(self.filename, self.path)
274 self.full_filename = filefind(self.filename, self.path)
275
275
276 def _read_file_as_dict(self):
276 def _read_file_as_dict(self):
277 """Load the config file into self.config, with recursive loading."""
277 """Load the config file into self.config, with recursive loading."""
278 # This closure is made available in the namespace that is used
278 # This closure is made available in the namespace that is used
279 # to exec the config file. It allows users to call
279 # to exec the config file. It allows users to call
280 # load_subconfig('myconfig.py') to load config files recursively.
280 # load_subconfig('myconfig.py') to load config files recursively.
281 # It needs to be a closure because it has references to self.path
281 # It needs to be a closure because it has references to self.path
282 # and self.config. The sub-config is loaded with the same path
282 # and self.config. The sub-config is loaded with the same path
283 # as the parent, but it uses an empty config which is then merged
283 # as the parent, but it uses an empty config which is then merged
284 # with the parents.
284 # with the parents.
285
285
286 # If a profile is specified, the config file will be loaded
286 # If a profile is specified, the config file will be loaded
287 # from that profile
287 # from that profile
288
288
289 def load_subconfig(fname, profile=None):
289 def load_subconfig(fname, profile=None):
290 # import here to prevent circular imports
290 # import here to prevent circular imports
291 from IPython.core.profiledir import ProfileDir, ProfileDirError
291 from IPython.core.profiledir import ProfileDir, ProfileDirError
292 if profile is not None:
292 if profile is not None:
293 try:
293 try:
294 profile_dir = ProfileDir.find_profile_dir_by_name(
294 profile_dir = ProfileDir.find_profile_dir_by_name(
295 get_ipython_dir(),
295 get_ipython_dir(),
296 profile,
296 profile,
297 )
297 )
298 except ProfileDirError:
298 except ProfileDirError:
299 return
299 return
300 path = profile_dir.location
300 path = profile_dir.location
301 else:
301 else:
302 path = self.path
302 path = self.path
303 loader = PyFileConfigLoader(fname, path)
303 loader = PyFileConfigLoader(fname, path)
304 try:
304 try:
305 sub_config = loader.load_config()
305 sub_config = loader.load_config()
306 except ConfigFileNotFound:
306 except ConfigFileNotFound:
307 # Pass silently if the sub config is not there. This happens
307 # Pass silently if the sub config is not there. This happens
308 # when a user s using a profile, but not the default config.
308 # when a user s using a profile, but not the default config.
309 pass
309 pass
310 else:
310 else:
311 self.config._merge(sub_config)
311 self.config._merge(sub_config)
312
312
313 # Again, this needs to be a closure and should be used in config
313 # Again, this needs to be a closure and should be used in config
314 # files to get the config being loaded.
314 # files to get the config being loaded.
315 def get_config():
315 def get_config():
316 return self.config
316 return self.config
317
317
318 namespace = dict(load_subconfig=load_subconfig, get_config=get_config)
318 namespace = dict(load_subconfig=load_subconfig, get_config=get_config)
319 fs_encoding = sys.getfilesystemencoding() or 'ascii'
319 fs_encoding = sys.getfilesystemencoding() or 'ascii'
320 conf_filename = self.full_filename.encode(fs_encoding)
320 conf_filename = self.full_filename.encode(fs_encoding)
321 py3compat.execfile(conf_filename, namespace)
321 py3compat.execfile(conf_filename, namespace)
322
322
323 def _convert_to_config(self):
323 def _convert_to_config(self):
324 if self.data is None:
324 if self.data is None:
325 ConfigLoaderError('self.data does not exist')
325 ConfigLoaderError('self.data does not exist')
326
326
327
327
328 class CommandLineConfigLoader(ConfigLoader):
328 class CommandLineConfigLoader(ConfigLoader):
329 """A config loader for command line arguments.
329 """A config loader for command line arguments.
330
330
331 As we add more command line based loaders, the common logic should go
331 As we add more command line based loaders, the common logic should go
332 here.
332 here.
333 """
333 """
334
334
335 def _exec_config_str(self, lhs, rhs):
335 def _exec_config_str(self, lhs, rhs):
336 """execute self.config.<lhs>=<rhs>
336 """execute self.config.<lhs>=<rhs>
337
337
338 * expands ~ with expanduser
338 * expands ~ with expanduser
339 * tries to assign with raw exec, otherwise assigns with just the string,
339 * tries to assign with raw exec, otherwise assigns with just the string,
340 allowing `--C.a=foobar` and `--C.a="foobar"` to be equivalent. *Not*
340 allowing `--C.a=foobar` and `--C.a="foobar"` to be equivalent. *Not*
341 equivalent are `--C.a=4` and `--C.a='4'`.
341 equivalent are `--C.a=4` and `--C.a='4'`.
342 """
342 """
343 rhs = os.path.expanduser(rhs)
343 rhs = os.path.expanduser(rhs)
344 exec_str = 'self.config.' + lhs + '=' + rhs
344 exec_str = 'self.config.' + lhs + '=' + rhs
345 try:
345 try:
346 # Try to see if regular Python syntax will work. This
346 # Try to see if regular Python syntax will work. This
347 # won't handle strings as the quote marks are removed
347 # won't handle strings as the quote marks are removed
348 # by the system shell.
348 # by the system shell.
349 exec exec_str in locals(), globals()
349 exec exec_str in locals(), globals()
350 except (NameError, SyntaxError):
350 except (NameError, SyntaxError):
351 # This case happens if the rhs is a string but without
351 # This case happens if the rhs is a string but without
352 # the quote marks. Use repr, to get quote marks, and
352 # the quote marks. Use repr, to get quote marks, and
353 # 'u' prefix and see if
353 # 'u' prefix and see if
354 # it succeeds. If it still fails, we let it raise.
354 # it succeeds. If it still fails, we let it raise.
355 exec_str = u'self.config.' + lhs + '= rhs'
355 exec_str = u'self.config.' + lhs + '= rhs'
356 exec exec_str in locals(), globals()
356 exec exec_str in locals(), globals()
357
357
358 def _load_flag(self, cfg):
358 def _load_flag(self, cfg):
359 """update self.config from a flag, which can be a dict or Config"""
359 """update self.config from a flag, which can be a dict or Config"""
360 if isinstance(cfg, (dict, Config)):
360 if isinstance(cfg, (dict, Config)):
361 # don't clobber whole config sections, update
361 # don't clobber whole config sections, update
362 # each section from config:
362 # each section from config:
363 for sec,c in cfg.iteritems():
363 for sec,c in cfg.iteritems():
364 self.config[sec].update(c)
364 self.config[sec].update(c)
365 else:
365 else:
366 raise TypeError("Invalid flag: %r" % cfg)
366 raise TypeError("Invalid flag: %r" % cfg)
367
367
368 # raw --identifier=value pattern
368 # raw --identifier=value pattern
369 # but *also* accept '-' as wordsep, for aliases
369 # but *also* accept '-' as wordsep, for aliases
370 # accepts: --foo=a
370 # accepts: --foo=a
371 # --Class.trait=value
371 # --Class.trait=value
372 # --alias-name=value
372 # --alias-name=value
373 # rejects: -foo=value
373 # rejects: -foo=value
374 # --foo
374 # --foo
375 # --Class.trait
375 # --Class.trait
376 kv_pattern = re.compile(r'\-\-[A-Za-z][\w\-]*(\.[\w\-]+)*\=.*')
376 kv_pattern = re.compile(r'\-\-[A-Za-z][\w\-]*(\.[\w\-]+)*\=.*')
377
377
378 # just flags, no assignments, with two *or one* leading '-'
378 # just flags, no assignments, with two *or one* leading '-'
379 # accepts: --foo
379 # accepts: --foo
380 # -foo-bar-again
380 # -foo-bar-again
381 # rejects: --anything=anything
381 # rejects: --anything=anything
382 # --two.word
382 # --two.word
383
383
384 flag_pattern = re.compile(r'\-\-?\w+[\-\w]*$')
384 flag_pattern = re.compile(r'\-\-?\w+[\-\w]*$')
385
385
386 class KeyValueConfigLoader(CommandLineConfigLoader):
386 class KeyValueConfigLoader(CommandLineConfigLoader):
387 """A config loader that loads key value pairs from the command line.
387 """A config loader that loads key value pairs from the command line.
388
388
389 This allows command line options to be gives in the following form::
389 This allows command line options to be gives in the following form::
390
390
391 ipython --profile="foo" --InteractiveShell.autocall=False
391 ipython --profile="foo" --InteractiveShell.autocall=False
392 """
392 """
393
393
394 def __init__(self, argv=None, aliases=None, flags=None):
394 def __init__(self, argv=None, aliases=None, flags=None):
395 """Create a key value pair config loader.
395 """Create a key value pair config loader.
396
396
397 Parameters
397 Parameters
398 ----------
398 ----------
399 argv : list
399 argv : list
400 A list that has the form of sys.argv[1:] which has unicode
400 A list that has the form of sys.argv[1:] which has unicode
401 elements of the form u"key=value". If this is None (default),
401 elements of the form u"key=value". If this is None (default),
402 then sys.argv[1:] will be used.
402 then sys.argv[1:] will be used.
403 aliases : dict
403 aliases : dict
404 A dict of aliases for configurable traits.
404 A dict of aliases for configurable traits.
405 Keys are the short aliases, Values are the resolved trait.
405 Keys are the short aliases, Values are the resolved trait.
406 Of the form: `{'alias' : 'Configurable.trait'}`
406 Of the form: `{'alias' : 'Configurable.trait'}`
407 flags : dict
407 flags : dict
408 A dict of flags, keyed by str name. Vaues can be Config objects,
408 A dict of flags, keyed by str name. Vaues can be Config objects,
409 dicts, or "key=value" strings. If Config or dict, when the flag
409 dicts, or "key=value" strings. If Config or dict, when the flag
410 is triggered, The flag is loaded as `self.config.update(m)`.
410 is triggered, The flag is loaded as `self.config.update(m)`.
411
411
412 Returns
412 Returns
413 -------
413 -------
414 config : Config
414 config : Config
415 The resulting Config object.
415 The resulting Config object.
416
416
417 Examples
417 Examples
418 --------
418 --------
419
419
420 >>> from IPython.config.loader import KeyValueConfigLoader
420 >>> from IPython.config.loader import KeyValueConfigLoader
421 >>> cl = KeyValueConfigLoader()
421 >>> cl = KeyValueConfigLoader()
422 >>> cl.load_config(["--A.name='brian'","--B.number=0"])
422 >>> cl.load_config(["--A.name='brian'","--B.number=0"])
423 {'A': {'name': 'brian'}, 'B': {'number': 0}}
423 {'A': {'name': 'brian'}, 'B': {'number': 0}}
424 """
424 """
425 self.clear()
425 self.clear()
426 if argv is None:
426 if argv is None:
427 argv = sys.argv[1:]
427 argv = sys.argv[1:]
428 self.argv = argv
428 self.argv = argv
429 self.aliases = aliases or {}
429 self.aliases = aliases or {}
430 self.flags = flags or {}
430 self.flags = flags or {}
431
431
432
432
433 def clear(self):
433 def clear(self):
434 super(KeyValueConfigLoader, self).clear()
434 super(KeyValueConfigLoader, self).clear()
435 self.extra_args = []
435 self.extra_args = []
436
436
437
437
438 def _decode_argv(self, argv, enc=None):
438 def _decode_argv(self, argv, enc=None):
439 """decode argv if bytes, using stin.encoding, falling back on default enc"""
439 """decode argv if bytes, using stin.encoding, falling back on default enc"""
440 uargv = []
440 uargv = []
441 if enc is None:
441 if enc is None:
442 enc = text.getdefaultencoding()
442 enc = py3compat.getdefaultencoding()
443 for arg in argv:
443 for arg in argv:
444 if not isinstance(arg, unicode):
444 if not isinstance(arg, unicode):
445 # only decode if not already decoded
445 # only decode if not already decoded
446 arg = arg.decode(enc)
446 arg = arg.decode(enc)
447 uargv.append(arg)
447 uargv.append(arg)
448 return uargv
448 return uargv
449
449
450
450
451 def load_config(self, argv=None, aliases=None, flags=None):
451 def load_config(self, argv=None, aliases=None, flags=None):
452 """Parse the configuration and generate the Config object.
452 """Parse the configuration and generate the Config object.
453
453
454 After loading, any arguments that are not key-value or
454 After loading, any arguments that are not key-value or
455 flags will be stored in self.extra_args - a list of
455 flags will be stored in self.extra_args - a list of
456 unparsed command-line arguments. This is used for
456 unparsed command-line arguments. This is used for
457 arguments such as input files or subcommands.
457 arguments such as input files or subcommands.
458
458
459 Parameters
459 Parameters
460 ----------
460 ----------
461 argv : list, optional
461 argv : list, optional
462 A list that has the form of sys.argv[1:] which has unicode
462 A list that has the form of sys.argv[1:] which has unicode
463 elements of the form u"key=value". If this is None (default),
463 elements of the form u"key=value". If this is None (default),
464 then self.argv will be used.
464 then self.argv will be used.
465 aliases : dict
465 aliases : dict
466 A dict of aliases for configurable traits.
466 A dict of aliases for configurable traits.
467 Keys are the short aliases, Values are the resolved trait.
467 Keys are the short aliases, Values are the resolved trait.
468 Of the form: `{'alias' : 'Configurable.trait'}`
468 Of the form: `{'alias' : 'Configurable.trait'}`
469 flags : dict
469 flags : dict
470 A dict of flags, keyed by str name. Values can be Config objects
470 A dict of flags, keyed by str name. Values can be Config objects
471 or dicts. When the flag is triggered, The config is loaded as
471 or dicts. When the flag is triggered, The config is loaded as
472 `self.config.update(cfg)`.
472 `self.config.update(cfg)`.
473 """
473 """
474 from IPython.config.configurable import Configurable
474 from IPython.config.configurable import Configurable
475
475
476 self.clear()
476 self.clear()
477 if argv is None:
477 if argv is None:
478 argv = self.argv
478 argv = self.argv
479 if aliases is None:
479 if aliases is None:
480 aliases = self.aliases
480 aliases = self.aliases
481 if flags is None:
481 if flags is None:
482 flags = self.flags
482 flags = self.flags
483
483
484 # ensure argv is a list of unicode strings:
484 # ensure argv is a list of unicode strings:
485 uargv = self._decode_argv(argv)
485 uargv = self._decode_argv(argv)
486 for idx,raw in enumerate(uargv):
486 for idx,raw in enumerate(uargv):
487 # strip leading '-'
487 # strip leading '-'
488 item = raw.lstrip('-')
488 item = raw.lstrip('-')
489
489
490 if raw == '--':
490 if raw == '--':
491 # don't parse arguments after '--'
491 # don't parse arguments after '--'
492 # this is useful for relaying arguments to scripts, e.g.
492 # this is useful for relaying arguments to scripts, e.g.
493 # ipython -i foo.py --pylab=qt -- args after '--' go-to-foo.py
493 # ipython -i foo.py --pylab=qt -- args after '--' go-to-foo.py
494 self.extra_args.extend(uargv[idx+1:])
494 self.extra_args.extend(uargv[idx+1:])
495 break
495 break
496
496
497 if kv_pattern.match(raw):
497 if kv_pattern.match(raw):
498 lhs,rhs = item.split('=',1)
498 lhs,rhs = item.split('=',1)
499 # Substitute longnames for aliases.
499 # Substitute longnames for aliases.
500 if lhs in aliases:
500 if lhs in aliases:
501 lhs = aliases[lhs]
501 lhs = aliases[lhs]
502 if '.' not in lhs:
502 if '.' not in lhs:
503 # probably a mistyped alias, but not technically illegal
503 # probably a mistyped alias, but not technically illegal
504 warn.warn("Unrecognized alias: '%s', it will probably have no effect."%lhs)
504 warn.warn("Unrecognized alias: '%s', it will probably have no effect."%lhs)
505 try:
505 try:
506 self._exec_config_str(lhs, rhs)
506 self._exec_config_str(lhs, rhs)
507 except Exception:
507 except Exception:
508 raise ArgumentError("Invalid argument: '%s'" % raw)
508 raise ArgumentError("Invalid argument: '%s'" % raw)
509
509
510 elif flag_pattern.match(raw):
510 elif flag_pattern.match(raw):
511 if item in flags:
511 if item in flags:
512 cfg,help = flags[item]
512 cfg,help = flags[item]
513 self._load_flag(cfg)
513 self._load_flag(cfg)
514 else:
514 else:
515 raise ArgumentError("Unrecognized flag: '%s'"%raw)
515 raise ArgumentError("Unrecognized flag: '%s'"%raw)
516 elif raw.startswith('-'):
516 elif raw.startswith('-'):
517 kv = '--'+item
517 kv = '--'+item
518 if kv_pattern.match(kv):
518 if kv_pattern.match(kv):
519 raise ArgumentError("Invalid argument: '%s', did you mean '%s'?"%(raw, kv))
519 raise ArgumentError("Invalid argument: '%s', did you mean '%s'?"%(raw, kv))
520 else:
520 else:
521 raise ArgumentError("Invalid argument: '%s'"%raw)
521 raise ArgumentError("Invalid argument: '%s'"%raw)
522 else:
522 else:
523 # keep all args that aren't valid in a list,
523 # keep all args that aren't valid in a list,
524 # in case our parent knows what to do with them.
524 # in case our parent knows what to do with them.
525 self.extra_args.append(item)
525 self.extra_args.append(item)
526 return self.config
526 return self.config
527
527
528 class ArgParseConfigLoader(CommandLineConfigLoader):
528 class ArgParseConfigLoader(CommandLineConfigLoader):
529 """A loader that uses the argparse module to load from the command line."""
529 """A loader that uses the argparse module to load from the command line."""
530
530
531 def __init__(self, argv=None, aliases=None, flags=None, *parser_args, **parser_kw):
531 def __init__(self, argv=None, aliases=None, flags=None, *parser_args, **parser_kw):
532 """Create a config loader for use with argparse.
532 """Create a config loader for use with argparse.
533
533
534 Parameters
534 Parameters
535 ----------
535 ----------
536
536
537 argv : optional, list
537 argv : optional, list
538 If given, used to read command-line arguments from, otherwise
538 If given, used to read command-line arguments from, otherwise
539 sys.argv[1:] is used.
539 sys.argv[1:] is used.
540
540
541 parser_args : tuple
541 parser_args : tuple
542 A tuple of positional arguments that will be passed to the
542 A tuple of positional arguments that will be passed to the
543 constructor of :class:`argparse.ArgumentParser`.
543 constructor of :class:`argparse.ArgumentParser`.
544
544
545 parser_kw : dict
545 parser_kw : dict
546 A tuple of keyword arguments that will be passed to the
546 A tuple of keyword arguments that will be passed to the
547 constructor of :class:`argparse.ArgumentParser`.
547 constructor of :class:`argparse.ArgumentParser`.
548
548
549 Returns
549 Returns
550 -------
550 -------
551 config : Config
551 config : Config
552 The resulting Config object.
552 The resulting Config object.
553 """
553 """
554 super(CommandLineConfigLoader, self).__init__()
554 super(CommandLineConfigLoader, self).__init__()
555 self.clear()
555 self.clear()
556 if argv is None:
556 if argv is None:
557 argv = sys.argv[1:]
557 argv = sys.argv[1:]
558 self.argv = argv
558 self.argv = argv
559 self.aliases = aliases or {}
559 self.aliases = aliases or {}
560 self.flags = flags or {}
560 self.flags = flags or {}
561
561
562 self.parser_args = parser_args
562 self.parser_args = parser_args
563 self.version = parser_kw.pop("version", None)
563 self.version = parser_kw.pop("version", None)
564 kwargs = dict(argument_default=argparse.SUPPRESS)
564 kwargs = dict(argument_default=argparse.SUPPRESS)
565 kwargs.update(parser_kw)
565 kwargs.update(parser_kw)
566 self.parser_kw = kwargs
566 self.parser_kw = kwargs
567
567
568 def load_config(self, argv=None, aliases=None, flags=None):
568 def load_config(self, argv=None, aliases=None, flags=None):
569 """Parse command line arguments and return as a Config object.
569 """Parse command line arguments and return as a Config object.
570
570
571 Parameters
571 Parameters
572 ----------
572 ----------
573
573
574 args : optional, list
574 args : optional, list
575 If given, a list with the structure of sys.argv[1:] to parse
575 If given, a list with the structure of sys.argv[1:] to parse
576 arguments from. If not given, the instance's self.argv attribute
576 arguments from. If not given, the instance's self.argv attribute
577 (given at construction time) is used."""
577 (given at construction time) is used."""
578 self.clear()
578 self.clear()
579 if argv is None:
579 if argv is None:
580 argv = self.argv
580 argv = self.argv
581 if aliases is None:
581 if aliases is None:
582 aliases = self.aliases
582 aliases = self.aliases
583 if flags is None:
583 if flags is None:
584 flags = self.flags
584 flags = self.flags
585 self._create_parser(aliases, flags)
585 self._create_parser(aliases, flags)
586 self._parse_args(argv)
586 self._parse_args(argv)
587 self._convert_to_config()
587 self._convert_to_config()
588 return self.config
588 return self.config
589
589
590 def get_extra_args(self):
590 def get_extra_args(self):
591 if hasattr(self, 'extra_args'):
591 if hasattr(self, 'extra_args'):
592 return self.extra_args
592 return self.extra_args
593 else:
593 else:
594 return []
594 return []
595
595
596 def _create_parser(self, aliases=None, flags=None):
596 def _create_parser(self, aliases=None, flags=None):
597 self.parser = ArgumentParser(*self.parser_args, **self.parser_kw)
597 self.parser = ArgumentParser(*self.parser_args, **self.parser_kw)
598 self._add_arguments(aliases, flags)
598 self._add_arguments(aliases, flags)
599
599
600 def _add_arguments(self, aliases=None, flags=None):
600 def _add_arguments(self, aliases=None, flags=None):
601 raise NotImplementedError("subclasses must implement _add_arguments")
601 raise NotImplementedError("subclasses must implement _add_arguments")
602
602
603 def _parse_args(self, args):
603 def _parse_args(self, args):
604 """self.parser->self.parsed_data"""
604 """self.parser->self.parsed_data"""
605 # decode sys.argv to support unicode command-line options
605 # decode sys.argv to support unicode command-line options
606 enc = text.getdefaultencoding()
606 enc = py3compat.getdefaultencoding()
607 uargs = [py3compat.cast_unicode(a, enc) for a in args]
607 uargs = [py3compat.cast_unicode(a, enc) for a in args]
608 self.parsed_data, self.extra_args = self.parser.parse_known_args(uargs)
608 self.parsed_data, self.extra_args = self.parser.parse_known_args(uargs)
609
609
610 def _convert_to_config(self):
610 def _convert_to_config(self):
611 """self.parsed_data->self.config"""
611 """self.parsed_data->self.config"""
612 for k, v in vars(self.parsed_data).iteritems():
612 for k, v in vars(self.parsed_data).iteritems():
613 exec "self.config.%s = v"%k in locals(), globals()
613 exec "self.config.%s = v"%k in locals(), globals()
614
614
615 class KVArgParseConfigLoader(ArgParseConfigLoader):
615 class KVArgParseConfigLoader(ArgParseConfigLoader):
616 """A config loader that loads aliases and flags with argparse,
616 """A config loader that loads aliases and flags with argparse,
617 but will use KVLoader for the rest. This allows better parsing
617 but will use KVLoader for the rest. This allows better parsing
618 of common args, such as `ipython -c 'print 5'`, but still gets
618 of common args, such as `ipython -c 'print 5'`, but still gets
619 arbitrary config with `ipython --InteractiveShell.use_readline=False`"""
619 arbitrary config with `ipython --InteractiveShell.use_readline=False`"""
620
620
621 def _convert_to_config(self):
621 def _convert_to_config(self):
622 """self.parsed_data->self.config"""
622 """self.parsed_data->self.config"""
623 for k, v in vars(self.parsed_data).iteritems():
623 for k, v in vars(self.parsed_data).iteritems():
624 self._exec_config_str(k, v)
624 self._exec_config_str(k, v)
625
625
626 def _add_arguments(self, aliases=None, flags=None):
626 def _add_arguments(self, aliases=None, flags=None):
627 self.alias_flags = {}
627 self.alias_flags = {}
628 # print aliases, flags
628 # print aliases, flags
629 if aliases is None:
629 if aliases is None:
630 aliases = self.aliases
630 aliases = self.aliases
631 if flags is None:
631 if flags is None:
632 flags = self.flags
632 flags = self.flags
633 paa = self.parser.add_argument
633 paa = self.parser.add_argument
634 for key,value in aliases.iteritems():
634 for key,value in aliases.iteritems():
635 if key in flags:
635 if key in flags:
636 # flags
636 # flags
637 nargs = '?'
637 nargs = '?'
638 else:
638 else:
639 nargs = None
639 nargs = None
640 if len(key) is 1:
640 if len(key) is 1:
641 paa('-'+key, '--'+key, type=unicode, dest=value, nargs=nargs)
641 paa('-'+key, '--'+key, type=unicode, dest=value, nargs=nargs)
642 else:
642 else:
643 paa('--'+key, type=unicode, dest=value, nargs=nargs)
643 paa('--'+key, type=unicode, dest=value, nargs=nargs)
644 for key, (value, help) in flags.iteritems():
644 for key, (value, help) in flags.iteritems():
645 if key in self.aliases:
645 if key in self.aliases:
646 #
646 #
647 self.alias_flags[self.aliases[key]] = value
647 self.alias_flags[self.aliases[key]] = value
648 continue
648 continue
649 if len(key) is 1:
649 if len(key) is 1:
650 paa('-'+key, '--'+key, action='append_const', dest='_flags', const=value)
650 paa('-'+key, '--'+key, action='append_const', dest='_flags', const=value)
651 else:
651 else:
652 paa('--'+key, action='append_const', dest='_flags', const=value)
652 paa('--'+key, action='append_const', dest='_flags', const=value)
653
653
654 def _convert_to_config(self):
654 def _convert_to_config(self):
655 """self.parsed_data->self.config, parse unrecognized extra args via KVLoader."""
655 """self.parsed_data->self.config, parse unrecognized extra args via KVLoader."""
656 # remove subconfigs list from namespace before transforming the Namespace
656 # remove subconfigs list from namespace before transforming the Namespace
657 if '_flags' in self.parsed_data:
657 if '_flags' in self.parsed_data:
658 subcs = self.parsed_data._flags
658 subcs = self.parsed_data._flags
659 del self.parsed_data._flags
659 del self.parsed_data._flags
660 else:
660 else:
661 subcs = []
661 subcs = []
662
662
663 for k, v in vars(self.parsed_data).iteritems():
663 for k, v in vars(self.parsed_data).iteritems():
664 if v is None:
664 if v is None:
665 # it was a flag that shares the name of an alias
665 # it was a flag that shares the name of an alias
666 subcs.append(self.alias_flags[k])
666 subcs.append(self.alias_flags[k])
667 else:
667 else:
668 # eval the KV assignment
668 # eval the KV assignment
669 self._exec_config_str(k, v)
669 self._exec_config_str(k, v)
670
670
671 for subc in subcs:
671 for subc in subcs:
672 self._load_flag(subc)
672 self._load_flag(subc)
673
673
674 if self.extra_args:
674 if self.extra_args:
675 sub_parser = KeyValueConfigLoader()
675 sub_parser = KeyValueConfigLoader()
676 sub_parser.load_config(self.extra_args)
676 sub_parser.load_config(self.extra_args)
677 self.config._merge(sub_parser.config)
677 self.config._merge(sub_parser.config)
678 self.extra_args = sub_parser.extra_args
678 self.extra_args = sub_parser.extra_args
679
679
680
680
681 def load_pyconfig_files(config_files, path):
681 def load_pyconfig_files(config_files, path):
682 """Load multiple Python config files, merging each of them in turn.
682 """Load multiple Python config files, merging each of them in turn.
683
683
684 Parameters
684 Parameters
685 ==========
685 ==========
686 config_files : list of str
686 config_files : list of str
687 List of config files names to load and merge into the config.
687 List of config files names to load and merge into the config.
688 path : unicode
688 path : unicode
689 The full path to the location of the config files.
689 The full path to the location of the config files.
690 """
690 """
691 config = Config()
691 config = Config()
692 for cf in config_files:
692 for cf in config_files:
693 loader = PyFileConfigLoader(cf, path=path)
693 loader = PyFileConfigLoader(cf, path=path)
694 try:
694 try:
695 next_config = loader.load_config()
695 next_config = loader.load_config()
696 except ConfigFileNotFound:
696 except ConfigFileNotFound:
697 pass
697 pass
698 except:
698 except:
699 raise
699 raise
700 else:
700 else:
701 config._merge(next_config)
701 config._merge(next_config)
702 return config
702 return config
@@ -1,59 +1,59 b''
1 """Support for interactive macros in IPython"""
1 """Support for interactive macros in IPython"""
2
2
3 #*****************************************************************************
3 #*****************************************************************************
4 # Copyright (C) 2001-2005 Fernando Perez <fperez@colorado.edu>
4 # Copyright (C) 2001-2005 Fernando Perez <fperez@colorado.edu>
5 #
5 #
6 # Distributed under the terms of the BSD License. The full license is in
6 # Distributed under the terms of the BSD License. The full license is in
7 # the file COPYING, distributed as part of this software.
7 # the file COPYING, distributed as part of this software.
8 #*****************************************************************************
8 #*****************************************************************************
9
9
10 import re
10 import re
11 import sys
11 import sys
12
12
13 from IPython.utils import py3compat
13 from IPython.utils import py3compat
14
14
15 coding_declaration = re.compile(r"#\s*coding[:=]\s*([-\w.]+)")
15 coding_declaration = re.compile(r"#\s*coding[:=]\s*([-\w.]+)")
16
16
17 class Macro(object):
17 class Macro(object):
18 """Simple class to store the value of macros as strings.
18 """Simple class to store the value of macros as strings.
19
19
20 Macro is just a callable that executes a string of IPython
20 Macro is just a callable that executes a string of IPython
21 input when called.
21 input when called.
22
22
23 Args to macro are available in _margv list if you need them.
23 Args to macro are available in _margv list if you need them.
24 """
24 """
25
25
26 def __init__(self,code):
26 def __init__(self,code):
27 """store the macro value, as a single string which can be executed"""
27 """store the macro value, as a single string which can be executed"""
28 lines = []
28 lines = []
29 enc = None
29 enc = None
30 for line in code.splitlines():
30 for line in code.splitlines():
31 coding_match = coding_declaration.match(line)
31 coding_match = coding_declaration.match(line)
32 if coding_match:
32 if coding_match:
33 enc = coding_match.group(1)
33 enc = coding_match.group(1)
34 else:
34 else:
35 lines.append(line)
35 lines.append(line)
36 code = "\n".join(lines)
36 code = "\n".join(lines)
37 if isinstance(code, bytes):
37 if isinstance(code, bytes):
38 code = code.decode(enc or sys.getdefaultencoding())
38 code = code.decode(enc or py3compat.getdefaultencoding())
39 self.value = code + '\n'
39 self.value = code + '\n'
40
40
41 def __str__(self):
41 def __str__(self):
42 return py3compat.unicode_to_str(self.value)
42 return py3compat.unicode_to_str(self.value)
43
43
44 def __unicode__(self):
44 def __unicode__(self):
45 return self.value
45 return self.value
46
46
47 def __repr__(self):
47 def __repr__(self):
48 return 'IPython.macro.Macro(%s)' % repr(self.value)
48 return 'IPython.macro.Macro(%s)' % repr(self.value)
49
49
50 def __getstate__(self):
50 def __getstate__(self):
51 """ needed for safe pickling via %store """
51 """ needed for safe pickling via %store """
52 return {'value': self.value}
52 return {'value': self.value}
53
53
54 def __add__(self, other):
54 def __add__(self, other):
55 if isinstance(other, Macro):
55 if isinstance(other, Macro):
56 return Macro(self.value + other.value)
56 return Macro(self.value + other.value)
57 elif isinstance(other, basestring):
57 elif isinstance(other, basestring):
58 return Macro(self.value + other)
58 return Macro(self.value + other)
59 raise TypeError
59 raise TypeError
@@ -1,3798 +1,3798 b''
1 # encoding: utf-8
1 # encoding: utf-8
2 """Magic functions for InteractiveShell.
2 """Magic functions for InteractiveShell.
3 """
3 """
4
4
5 #-----------------------------------------------------------------------------
5 #-----------------------------------------------------------------------------
6 # Copyright (C) 2001 Janko Hauser <jhauser@zscout.de> and
6 # Copyright (C) 2001 Janko Hauser <jhauser@zscout.de> and
7 # Copyright (C) 2001-2007 Fernando Perez <fperez@colorado.edu>
7 # Copyright (C) 2001-2007 Fernando Perez <fperez@colorado.edu>
8 # Copyright (C) 2008-2011 The IPython Development Team
8 # Copyright (C) 2008-2011 The IPython Development Team
9
9
10 # Distributed under the terms of the BSD License. The full license is in
10 # Distributed under the terms of the BSD License. The full license is in
11 # the file COPYING, distributed as part of this software.
11 # the file COPYING, distributed as part of this software.
12 #-----------------------------------------------------------------------------
12 #-----------------------------------------------------------------------------
13
13
14 #-----------------------------------------------------------------------------
14 #-----------------------------------------------------------------------------
15 # Imports
15 # Imports
16 #-----------------------------------------------------------------------------
16 #-----------------------------------------------------------------------------
17
17
18 import __builtin__ as builtin_mod
18 import __builtin__ as builtin_mod
19 import __future__
19 import __future__
20 import bdb
20 import bdb
21 import inspect
21 import inspect
22 import imp
22 import imp
23 import io
23 import io
24 import os
24 import os
25 import sys
25 import sys
26 import shutil
26 import shutil
27 import re
27 import re
28 import time
28 import time
29 import gc
29 import gc
30 from StringIO import StringIO
30 from StringIO import StringIO
31 from getopt import getopt,GetoptError
31 from getopt import getopt,GetoptError
32 from pprint import pformat
32 from pprint import pformat
33 from xmlrpclib import ServerProxy
33 from xmlrpclib import ServerProxy
34
34
35 # cProfile was added in Python2.5
35 # cProfile was added in Python2.5
36 try:
36 try:
37 import cProfile as profile
37 import cProfile as profile
38 import pstats
38 import pstats
39 except ImportError:
39 except ImportError:
40 # profile isn't bundled by default in Debian for license reasons
40 # profile isn't bundled by default in Debian for license reasons
41 try:
41 try:
42 import profile,pstats
42 import profile,pstats
43 except ImportError:
43 except ImportError:
44 profile = pstats = None
44 profile = pstats = None
45
45
46 import IPython
46 import IPython
47 from IPython.core import debugger, oinspect
47 from IPython.core import debugger, oinspect
48 from IPython.core.error import TryNext
48 from IPython.core.error import TryNext
49 from IPython.core.error import UsageError
49 from IPython.core.error import UsageError
50 from IPython.core.error import StdinNotImplementedError
50 from IPython.core.error import StdinNotImplementedError
51 from IPython.core.fakemodule import FakeModule
51 from IPython.core.fakemodule import FakeModule
52 from IPython.core.profiledir import ProfileDir
52 from IPython.core.profiledir import ProfileDir
53 from IPython.core.macro import Macro
53 from IPython.core.macro import Macro
54 from IPython.core import magic_arguments, page
54 from IPython.core import magic_arguments, page
55 from IPython.core.prefilter import ESC_MAGIC
55 from IPython.core.prefilter import ESC_MAGIC
56 from IPython.core.pylabtools import mpl_runner
56 from IPython.core.pylabtools import mpl_runner
57 from IPython.testing.skipdoctest import skip_doctest
57 from IPython.testing.skipdoctest import skip_doctest
58 from IPython.utils import py3compat
58 from IPython.utils import py3compat
59 from IPython.utils import openpy
59 from IPython.utils import openpy
60 from IPython.utils.io import file_read, nlprint
60 from IPython.utils.io import file_read, nlprint
61 from IPython.utils.module_paths import find_mod
61 from IPython.utils.module_paths import find_mod
62 from IPython.utils.path import get_py_filename, unquote_filename
62 from IPython.utils.path import get_py_filename, unquote_filename
63 from IPython.utils.process import arg_split, abbrev_cwd
63 from IPython.utils.process import arg_split, abbrev_cwd
64 from IPython.utils.terminal import set_term_title
64 from IPython.utils.terminal import set_term_title
65 from IPython.utils.text import LSString, SList, format_screen
65 from IPython.utils.text import LSString, SList, format_screen
66 from IPython.utils.timing import clock, clock2
66 from IPython.utils.timing import clock, clock2
67 from IPython.utils.warn import warn, error
67 from IPython.utils.warn import warn, error
68 from IPython.utils.ipstruct import Struct
68 from IPython.utils.ipstruct import Struct
69 from IPython.config.application import Application
69 from IPython.config.application import Application
70
70
71 #-----------------------------------------------------------------------------
71 #-----------------------------------------------------------------------------
72 # Utility functions
72 # Utility functions
73 #-----------------------------------------------------------------------------
73 #-----------------------------------------------------------------------------
74
74
75 def on_off(tag):
75 def on_off(tag):
76 """Return an ON/OFF string for a 1/0 input. Simple utility function."""
76 """Return an ON/OFF string for a 1/0 input. Simple utility function."""
77 return ['OFF','ON'][tag]
77 return ['OFF','ON'][tag]
78
78
79 class Bunch: pass
79 class Bunch: pass
80
80
81 def compress_dhist(dh):
81 def compress_dhist(dh):
82 head, tail = dh[:-10], dh[-10:]
82 head, tail = dh[:-10], dh[-10:]
83
83
84 newhead = []
84 newhead = []
85 done = set()
85 done = set()
86 for h in head:
86 for h in head:
87 if h in done:
87 if h in done:
88 continue
88 continue
89 newhead.append(h)
89 newhead.append(h)
90 done.add(h)
90 done.add(h)
91
91
92 return newhead + tail
92 return newhead + tail
93
93
94 def needs_local_scope(func):
94 def needs_local_scope(func):
95 """Decorator to mark magic functions which need to local scope to run."""
95 """Decorator to mark magic functions which need to local scope to run."""
96 func.needs_local_scope = True
96 func.needs_local_scope = True
97 return func
97 return func
98
98
99
99
100 # Used for exception handling in magic_edit
100 # Used for exception handling in magic_edit
101 class MacroToEdit(ValueError): pass
101 class MacroToEdit(ValueError): pass
102
102
103 #***************************************************************************
103 #***************************************************************************
104 # Main class implementing Magic functionality
104 # Main class implementing Magic functionality
105
105
106 # XXX - for some odd reason, if Magic is made a new-style class, we get errors
106 # XXX - for some odd reason, if Magic is made a new-style class, we get errors
107 # on construction of the main InteractiveShell object. Something odd is going
107 # on construction of the main InteractiveShell object. Something odd is going
108 # on with super() calls, Configurable and the MRO... For now leave it as-is, but
108 # on with super() calls, Configurable and the MRO... For now leave it as-is, but
109 # eventually this needs to be clarified.
109 # eventually this needs to be clarified.
110 # BG: This is because InteractiveShell inherits from this, but is itself a
110 # BG: This is because InteractiveShell inherits from this, but is itself a
111 # Configurable. This messes up the MRO in some way. The fix is that we need to
111 # Configurable. This messes up the MRO in some way. The fix is that we need to
112 # make Magic a configurable that InteractiveShell does not subclass.
112 # make Magic a configurable that InteractiveShell does not subclass.
113
113
114 class Magic:
114 class Magic:
115 """Magic functions for InteractiveShell.
115 """Magic functions for InteractiveShell.
116
116
117 Shell functions which can be reached as %function_name. All magic
117 Shell functions which can be reached as %function_name. All magic
118 functions should accept a string, which they can parse for their own
118 functions should accept a string, which they can parse for their own
119 needs. This can make some functions easier to type, eg `%cd ../`
119 needs. This can make some functions easier to type, eg `%cd ../`
120 vs. `%cd("../")`
120 vs. `%cd("../")`
121
121
122 ALL definitions MUST begin with the prefix magic_. The user won't need it
122 ALL definitions MUST begin with the prefix magic_. The user won't need it
123 at the command line, but it is is needed in the definition. """
123 at the command line, but it is is needed in the definition. """
124
124
125 # class globals
125 # class globals
126 auto_status = ['Automagic is OFF, % prefix IS needed for magic functions.',
126 auto_status = ['Automagic is OFF, % prefix IS needed for magic functions.',
127 'Automagic is ON, % prefix NOT needed for magic functions.']
127 'Automagic is ON, % prefix NOT needed for magic functions.']
128
128
129
129
130 configurables = None
130 configurables = None
131 #......................................................................
131 #......................................................................
132 # some utility functions
132 # some utility functions
133
133
134 def __init__(self,shell):
134 def __init__(self,shell):
135
135
136 self.options_table = {}
136 self.options_table = {}
137 if profile is None:
137 if profile is None:
138 self.magic_prun = self.profile_missing_notice
138 self.magic_prun = self.profile_missing_notice
139 self.shell = shell
139 self.shell = shell
140 if self.configurables is None:
140 if self.configurables is None:
141 self.configurables = []
141 self.configurables = []
142
142
143 # namespace for holding state we may need
143 # namespace for holding state we may need
144 self._magic_state = Bunch()
144 self._magic_state = Bunch()
145
145
146 def profile_missing_notice(self, *args, **kwargs):
146 def profile_missing_notice(self, *args, **kwargs):
147 error("""\
147 error("""\
148 The profile module could not be found. It has been removed from the standard
148 The profile module could not be found. It has been removed from the standard
149 python packages because of its non-free license. To use profiling, install the
149 python packages because of its non-free license. To use profiling, install the
150 python-profiler package from non-free.""")
150 python-profiler package from non-free.""")
151
151
152 def default_option(self,fn,optstr):
152 def default_option(self,fn,optstr):
153 """Make an entry in the options_table for fn, with value optstr"""
153 """Make an entry in the options_table for fn, with value optstr"""
154
154
155 if fn not in self.lsmagic():
155 if fn not in self.lsmagic():
156 error("%s is not a magic function" % fn)
156 error("%s is not a magic function" % fn)
157 self.options_table[fn] = optstr
157 self.options_table[fn] = optstr
158
158
159 def lsmagic(self):
159 def lsmagic(self):
160 """Return a list of currently available magic functions.
160 """Return a list of currently available magic functions.
161
161
162 Gives a list of the bare names after mangling (['ls','cd', ...], not
162 Gives a list of the bare names after mangling (['ls','cd', ...], not
163 ['magic_ls','magic_cd',...]"""
163 ['magic_ls','magic_cd',...]"""
164
164
165 # FIXME. This needs a cleanup, in the way the magics list is built.
165 # FIXME. This needs a cleanup, in the way the magics list is built.
166
166
167 # magics in class definition
167 # magics in class definition
168 class_magic = lambda fn: fn.startswith('magic_') and \
168 class_magic = lambda fn: fn.startswith('magic_') and \
169 callable(Magic.__dict__[fn])
169 callable(Magic.__dict__[fn])
170 # in instance namespace (run-time user additions)
170 # in instance namespace (run-time user additions)
171 inst_magic = lambda fn: fn.startswith('magic_') and \
171 inst_magic = lambda fn: fn.startswith('magic_') and \
172 callable(self.__dict__[fn])
172 callable(self.__dict__[fn])
173 # and bound magics by user (so they can access self):
173 # and bound magics by user (so they can access self):
174 inst_bound_magic = lambda fn: fn.startswith('magic_') and \
174 inst_bound_magic = lambda fn: fn.startswith('magic_') and \
175 callable(self.__class__.__dict__[fn])
175 callable(self.__class__.__dict__[fn])
176 magics = filter(class_magic,Magic.__dict__.keys()) + \
176 magics = filter(class_magic,Magic.__dict__.keys()) + \
177 filter(inst_magic,self.__dict__.keys()) + \
177 filter(inst_magic,self.__dict__.keys()) + \
178 filter(inst_bound_magic,self.__class__.__dict__.keys())
178 filter(inst_bound_magic,self.__class__.__dict__.keys())
179 out = []
179 out = []
180 for fn in set(magics):
180 for fn in set(magics):
181 out.append(fn.replace('magic_','',1))
181 out.append(fn.replace('magic_','',1))
182 out.sort()
182 out.sort()
183 return out
183 return out
184
184
185 def extract_input_lines(self, range_str, raw=False):
185 def extract_input_lines(self, range_str, raw=False):
186 """Return as a string a set of input history slices.
186 """Return as a string a set of input history slices.
187
187
188 Parameters
188 Parameters
189 ----------
189 ----------
190 range_str : string
190 range_str : string
191 The set of slices is given as a string, like "~5/6-~4/2 4:8 9",
191 The set of slices is given as a string, like "~5/6-~4/2 4:8 9",
192 since this function is for use by magic functions which get their
192 since this function is for use by magic functions which get their
193 arguments as strings. The number before the / is the session
193 arguments as strings. The number before the / is the session
194 number: ~n goes n back from the current session.
194 number: ~n goes n back from the current session.
195
195
196 Optional Parameters:
196 Optional Parameters:
197 - raw(False): by default, the processed input is used. If this is
197 - raw(False): by default, the processed input is used. If this is
198 true, the raw input history is used instead.
198 true, the raw input history is used instead.
199
199
200 Note that slices can be called with two notations:
200 Note that slices can be called with two notations:
201
201
202 N:M -> standard python form, means including items N...(M-1).
202 N:M -> standard python form, means including items N...(M-1).
203
203
204 N-M -> include items N..M (closed endpoint)."""
204 N-M -> include items N..M (closed endpoint)."""
205 lines = self.shell.history_manager.\
205 lines = self.shell.history_manager.\
206 get_range_by_str(range_str, raw=raw)
206 get_range_by_str(range_str, raw=raw)
207 return "\n".join(x for _, _, x in lines)
207 return "\n".join(x for _, _, x in lines)
208
208
209 def arg_err(self,func):
209 def arg_err(self,func):
210 """Print docstring if incorrect arguments were passed"""
210 """Print docstring if incorrect arguments were passed"""
211 print 'Error in arguments:'
211 print 'Error in arguments:'
212 print oinspect.getdoc(func)
212 print oinspect.getdoc(func)
213
213
214 def format_latex(self,strng):
214 def format_latex(self,strng):
215 """Format a string for latex inclusion."""
215 """Format a string for latex inclusion."""
216
216
217 # Characters that need to be escaped for latex:
217 # Characters that need to be escaped for latex:
218 escape_re = re.compile(r'(%|_|\$|#|&)',re.MULTILINE)
218 escape_re = re.compile(r'(%|_|\$|#|&)',re.MULTILINE)
219 # Magic command names as headers:
219 # Magic command names as headers:
220 cmd_name_re = re.compile(r'^(%s.*?):' % ESC_MAGIC,
220 cmd_name_re = re.compile(r'^(%s.*?):' % ESC_MAGIC,
221 re.MULTILINE)
221 re.MULTILINE)
222 # Magic commands
222 # Magic commands
223 cmd_re = re.compile(r'(?P<cmd>%s.+?\b)(?!\}\}:)' % ESC_MAGIC,
223 cmd_re = re.compile(r'(?P<cmd>%s.+?\b)(?!\}\}:)' % ESC_MAGIC,
224 re.MULTILINE)
224 re.MULTILINE)
225 # Paragraph continue
225 # Paragraph continue
226 par_re = re.compile(r'\\$',re.MULTILINE)
226 par_re = re.compile(r'\\$',re.MULTILINE)
227
227
228 # The "\n" symbol
228 # The "\n" symbol
229 newline_re = re.compile(r'\\n')
229 newline_re = re.compile(r'\\n')
230
230
231 # Now build the string for output:
231 # Now build the string for output:
232 #strng = cmd_name_re.sub(r'\n\\texttt{\\textsl{\\large \1}}:',strng)
232 #strng = cmd_name_re.sub(r'\n\\texttt{\\textsl{\\large \1}}:',strng)
233 strng = cmd_name_re.sub(r'\n\\bigskip\n\\texttt{\\textbf{ \1}}:',
233 strng = cmd_name_re.sub(r'\n\\bigskip\n\\texttt{\\textbf{ \1}}:',
234 strng)
234 strng)
235 strng = cmd_re.sub(r'\\texttt{\g<cmd>}',strng)
235 strng = cmd_re.sub(r'\\texttt{\g<cmd>}',strng)
236 strng = par_re.sub(r'\\\\',strng)
236 strng = par_re.sub(r'\\\\',strng)
237 strng = escape_re.sub(r'\\\1',strng)
237 strng = escape_re.sub(r'\\\1',strng)
238 strng = newline_re.sub(r'\\textbackslash{}n',strng)
238 strng = newline_re.sub(r'\\textbackslash{}n',strng)
239 return strng
239 return strng
240
240
241 def parse_options(self,arg_str,opt_str,*long_opts,**kw):
241 def parse_options(self,arg_str,opt_str,*long_opts,**kw):
242 """Parse options passed to an argument string.
242 """Parse options passed to an argument string.
243
243
244 The interface is similar to that of getopt(), but it returns back a
244 The interface is similar to that of getopt(), but it returns back a
245 Struct with the options as keys and the stripped argument string still
245 Struct with the options as keys and the stripped argument string still
246 as a string.
246 as a string.
247
247
248 arg_str is quoted as a true sys.argv vector by using shlex.split.
248 arg_str is quoted as a true sys.argv vector by using shlex.split.
249 This allows us to easily expand variables, glob files, quote
249 This allows us to easily expand variables, glob files, quote
250 arguments, etc.
250 arguments, etc.
251
251
252 Options:
252 Options:
253 -mode: default 'string'. If given as 'list', the argument string is
253 -mode: default 'string'. If given as 'list', the argument string is
254 returned as a list (split on whitespace) instead of a string.
254 returned as a list (split on whitespace) instead of a string.
255
255
256 -list_all: put all option values in lists. Normally only options
256 -list_all: put all option values in lists. Normally only options
257 appearing more than once are put in a list.
257 appearing more than once are put in a list.
258
258
259 -posix (True): whether to split the input line in POSIX mode or not,
259 -posix (True): whether to split the input line in POSIX mode or not,
260 as per the conventions outlined in the shlex module from the
260 as per the conventions outlined in the shlex module from the
261 standard library."""
261 standard library."""
262
262
263 # inject default options at the beginning of the input line
263 # inject default options at the beginning of the input line
264 caller = sys._getframe(1).f_code.co_name.replace('magic_','')
264 caller = sys._getframe(1).f_code.co_name.replace('magic_','')
265 arg_str = '%s %s' % (self.options_table.get(caller,''),arg_str)
265 arg_str = '%s %s' % (self.options_table.get(caller,''),arg_str)
266
266
267 mode = kw.get('mode','string')
267 mode = kw.get('mode','string')
268 if mode not in ['string','list']:
268 if mode not in ['string','list']:
269 raise ValueError,'incorrect mode given: %s' % mode
269 raise ValueError,'incorrect mode given: %s' % mode
270 # Get options
270 # Get options
271 list_all = kw.get('list_all',0)
271 list_all = kw.get('list_all',0)
272 posix = kw.get('posix', os.name == 'posix')
272 posix = kw.get('posix', os.name == 'posix')
273 strict = kw.get('strict', True)
273 strict = kw.get('strict', True)
274
274
275 # Check if we have more than one argument to warrant extra processing:
275 # Check if we have more than one argument to warrant extra processing:
276 odict = {} # Dictionary with options
276 odict = {} # Dictionary with options
277 args = arg_str.split()
277 args = arg_str.split()
278 if len(args) >= 1:
278 if len(args) >= 1:
279 # If the list of inputs only has 0 or 1 thing in it, there's no
279 # If the list of inputs only has 0 or 1 thing in it, there's no
280 # need to look for options
280 # need to look for options
281 argv = arg_split(arg_str, posix, strict)
281 argv = arg_split(arg_str, posix, strict)
282 # Do regular option processing
282 # Do regular option processing
283 try:
283 try:
284 opts,args = getopt(argv,opt_str,*long_opts)
284 opts,args = getopt(argv,opt_str,*long_opts)
285 except GetoptError,e:
285 except GetoptError,e:
286 raise UsageError('%s ( allowed: "%s" %s)' % (e.msg,opt_str,
286 raise UsageError('%s ( allowed: "%s" %s)' % (e.msg,opt_str,
287 " ".join(long_opts)))
287 " ".join(long_opts)))
288 for o,a in opts:
288 for o,a in opts:
289 if o.startswith('--'):
289 if o.startswith('--'):
290 o = o[2:]
290 o = o[2:]
291 else:
291 else:
292 o = o[1:]
292 o = o[1:]
293 try:
293 try:
294 odict[o].append(a)
294 odict[o].append(a)
295 except AttributeError:
295 except AttributeError:
296 odict[o] = [odict[o],a]
296 odict[o] = [odict[o],a]
297 except KeyError:
297 except KeyError:
298 if list_all:
298 if list_all:
299 odict[o] = [a]
299 odict[o] = [a]
300 else:
300 else:
301 odict[o] = a
301 odict[o] = a
302
302
303 # Prepare opts,args for return
303 # Prepare opts,args for return
304 opts = Struct(odict)
304 opts = Struct(odict)
305 if mode == 'string':
305 if mode == 'string':
306 args = ' '.join(args)
306 args = ' '.join(args)
307
307
308 return opts,args
308 return opts,args
309
309
310 #......................................................................
310 #......................................................................
311 # And now the actual magic functions
311 # And now the actual magic functions
312
312
313 # Functions for IPython shell work (vars,funcs, config, etc)
313 # Functions for IPython shell work (vars,funcs, config, etc)
314 def magic_lsmagic(self, parameter_s = ''):
314 def magic_lsmagic(self, parameter_s = ''):
315 """List currently available magic functions."""
315 """List currently available magic functions."""
316 mesc = ESC_MAGIC
316 mesc = ESC_MAGIC
317 print 'Available magic functions:\n'+mesc+\
317 print 'Available magic functions:\n'+mesc+\
318 (' '+mesc).join(self.lsmagic())
318 (' '+mesc).join(self.lsmagic())
319 print '\n' + Magic.auto_status[self.shell.automagic]
319 print '\n' + Magic.auto_status[self.shell.automagic]
320 return None
320 return None
321
321
322 def magic_magic(self, parameter_s = ''):
322 def magic_magic(self, parameter_s = ''):
323 """Print information about the magic function system.
323 """Print information about the magic function system.
324
324
325 Supported formats: -latex, -brief, -rest
325 Supported formats: -latex, -brief, -rest
326 """
326 """
327
327
328 mode = ''
328 mode = ''
329 try:
329 try:
330 if parameter_s.split()[0] == '-latex':
330 if parameter_s.split()[0] == '-latex':
331 mode = 'latex'
331 mode = 'latex'
332 if parameter_s.split()[0] == '-brief':
332 if parameter_s.split()[0] == '-brief':
333 mode = 'brief'
333 mode = 'brief'
334 if parameter_s.split()[0] == '-rest':
334 if parameter_s.split()[0] == '-rest':
335 mode = 'rest'
335 mode = 'rest'
336 rest_docs = []
336 rest_docs = []
337 except:
337 except:
338 pass
338 pass
339
339
340 magic_docs = []
340 magic_docs = []
341 for fname in self.lsmagic():
341 for fname in self.lsmagic():
342 mname = 'magic_' + fname
342 mname = 'magic_' + fname
343 for space in (Magic,self,self.__class__):
343 for space in (Magic,self,self.__class__):
344 try:
344 try:
345 fn = space.__dict__[mname]
345 fn = space.__dict__[mname]
346 except KeyError:
346 except KeyError:
347 pass
347 pass
348 else:
348 else:
349 break
349 break
350 if mode == 'brief':
350 if mode == 'brief':
351 # only first line
351 # only first line
352 if fn.__doc__:
352 if fn.__doc__:
353 fndoc = fn.__doc__.split('\n',1)[0]
353 fndoc = fn.__doc__.split('\n',1)[0]
354 else:
354 else:
355 fndoc = 'No documentation'
355 fndoc = 'No documentation'
356 else:
356 else:
357 if fn.__doc__:
357 if fn.__doc__:
358 fndoc = fn.__doc__.rstrip()
358 fndoc = fn.__doc__.rstrip()
359 else:
359 else:
360 fndoc = 'No documentation'
360 fndoc = 'No documentation'
361
361
362
362
363 if mode == 'rest':
363 if mode == 'rest':
364 rest_docs.append('**%s%s**::\n\n\t%s\n\n' %(ESC_MAGIC,
364 rest_docs.append('**%s%s**::\n\n\t%s\n\n' %(ESC_MAGIC,
365 fname,fndoc))
365 fname,fndoc))
366
366
367 else:
367 else:
368 magic_docs.append('%s%s:\n\t%s\n' %(ESC_MAGIC,
368 magic_docs.append('%s%s:\n\t%s\n' %(ESC_MAGIC,
369 fname,fndoc))
369 fname,fndoc))
370
370
371 magic_docs = ''.join(magic_docs)
371 magic_docs = ''.join(magic_docs)
372
372
373 if mode == 'rest':
373 if mode == 'rest':
374 return "".join(rest_docs)
374 return "".join(rest_docs)
375
375
376 if mode == 'latex':
376 if mode == 'latex':
377 print self.format_latex(magic_docs)
377 print self.format_latex(magic_docs)
378 return
378 return
379 else:
379 else:
380 magic_docs = format_screen(magic_docs)
380 magic_docs = format_screen(magic_docs)
381 if mode == 'brief':
381 if mode == 'brief':
382 return magic_docs
382 return magic_docs
383
383
384 outmsg = """
384 outmsg = """
385 IPython's 'magic' functions
385 IPython's 'magic' functions
386 ===========================
386 ===========================
387
387
388 The magic function system provides a series of functions which allow you to
388 The magic function system provides a series of functions which allow you to
389 control the behavior of IPython itself, plus a lot of system-type
389 control the behavior of IPython itself, plus a lot of system-type
390 features. All these functions are prefixed with a % character, but parameters
390 features. All these functions are prefixed with a % character, but parameters
391 are given without parentheses or quotes.
391 are given without parentheses or quotes.
392
392
393 NOTE: If you have 'automagic' enabled (via the command line option or with the
393 NOTE: If you have 'automagic' enabled (via the command line option or with the
394 %automagic function), you don't need to type in the % explicitly. By default,
394 %automagic function), you don't need to type in the % explicitly. By default,
395 IPython ships with automagic on, so you should only rarely need the % escape.
395 IPython ships with automagic on, so you should only rarely need the % escape.
396
396
397 Example: typing '%cd mydir' (without the quotes) changes you working directory
397 Example: typing '%cd mydir' (without the quotes) changes you working directory
398 to 'mydir', if it exists.
398 to 'mydir', if it exists.
399
399
400 For a list of the available magic functions, use %lsmagic. For a description
400 For a list of the available magic functions, use %lsmagic. For a description
401 of any of them, type %magic_name?, e.g. '%cd?'.
401 of any of them, type %magic_name?, e.g. '%cd?'.
402
402
403 Currently the magic system has the following functions:\n"""
403 Currently the magic system has the following functions:\n"""
404
404
405 mesc = ESC_MAGIC
405 mesc = ESC_MAGIC
406 outmsg = ("%s\n%s\n\nSummary of magic functions (from %slsmagic):"
406 outmsg = ("%s\n%s\n\nSummary of magic functions (from %slsmagic):"
407 "\n\n%s%s\n\n%s" % (outmsg,
407 "\n\n%s%s\n\n%s" % (outmsg,
408 magic_docs,mesc,mesc,
408 magic_docs,mesc,mesc,
409 (' '+mesc).join(self.lsmagic()),
409 (' '+mesc).join(self.lsmagic()),
410 Magic.auto_status[self.shell.automagic] ) )
410 Magic.auto_status[self.shell.automagic] ) )
411 page.page(outmsg)
411 page.page(outmsg)
412
412
413 def magic_automagic(self, parameter_s = ''):
413 def magic_automagic(self, parameter_s = ''):
414 """Make magic functions callable without having to type the initial %.
414 """Make magic functions callable without having to type the initial %.
415
415
416 Without argumentsl toggles on/off (when off, you must call it as
416 Without argumentsl toggles on/off (when off, you must call it as
417 %automagic, of course). With arguments it sets the value, and you can
417 %automagic, of course). With arguments it sets the value, and you can
418 use any of (case insensitive):
418 use any of (case insensitive):
419
419
420 - on,1,True: to activate
420 - on,1,True: to activate
421
421
422 - off,0,False: to deactivate.
422 - off,0,False: to deactivate.
423
423
424 Note that magic functions have lowest priority, so if there's a
424 Note that magic functions have lowest priority, so if there's a
425 variable whose name collides with that of a magic fn, automagic won't
425 variable whose name collides with that of a magic fn, automagic won't
426 work for that function (you get the variable instead). However, if you
426 work for that function (you get the variable instead). However, if you
427 delete the variable (del var), the previously shadowed magic function
427 delete the variable (del var), the previously shadowed magic function
428 becomes visible to automagic again."""
428 becomes visible to automagic again."""
429
429
430 arg = parameter_s.lower()
430 arg = parameter_s.lower()
431 if parameter_s in ('on','1','true'):
431 if parameter_s in ('on','1','true'):
432 self.shell.automagic = True
432 self.shell.automagic = True
433 elif parameter_s in ('off','0','false'):
433 elif parameter_s in ('off','0','false'):
434 self.shell.automagic = False
434 self.shell.automagic = False
435 else:
435 else:
436 self.shell.automagic = not self.shell.automagic
436 self.shell.automagic = not self.shell.automagic
437 print '\n' + Magic.auto_status[self.shell.automagic]
437 print '\n' + Magic.auto_status[self.shell.automagic]
438
438
439 @skip_doctest
439 @skip_doctest
440 def magic_autocall(self, parameter_s = ''):
440 def magic_autocall(self, parameter_s = ''):
441 """Make functions callable without having to type parentheses.
441 """Make functions callable without having to type parentheses.
442
442
443 Usage:
443 Usage:
444
444
445 %autocall [mode]
445 %autocall [mode]
446
446
447 The mode can be one of: 0->Off, 1->Smart, 2->Full. If not given, the
447 The mode can be one of: 0->Off, 1->Smart, 2->Full. If not given, the
448 value is toggled on and off (remembering the previous state).
448 value is toggled on and off (remembering the previous state).
449
449
450 In more detail, these values mean:
450 In more detail, these values mean:
451
451
452 0 -> fully disabled
452 0 -> fully disabled
453
453
454 1 -> active, but do not apply if there are no arguments on the line.
454 1 -> active, but do not apply if there are no arguments on the line.
455
455
456 In this mode, you get::
456 In this mode, you get::
457
457
458 In [1]: callable
458 In [1]: callable
459 Out[1]: <built-in function callable>
459 Out[1]: <built-in function callable>
460
460
461 In [2]: callable 'hello'
461 In [2]: callable 'hello'
462 ------> callable('hello')
462 ------> callable('hello')
463 Out[2]: False
463 Out[2]: False
464
464
465 2 -> Active always. Even if no arguments are present, the callable
465 2 -> Active always. Even if no arguments are present, the callable
466 object is called::
466 object is called::
467
467
468 In [2]: float
468 In [2]: float
469 ------> float()
469 ------> float()
470 Out[2]: 0.0
470 Out[2]: 0.0
471
471
472 Note that even with autocall off, you can still use '/' at the start of
472 Note that even with autocall off, you can still use '/' at the start of
473 a line to treat the first argument on the command line as a function
473 a line to treat the first argument on the command line as a function
474 and add parentheses to it::
474 and add parentheses to it::
475
475
476 In [8]: /str 43
476 In [8]: /str 43
477 ------> str(43)
477 ------> str(43)
478 Out[8]: '43'
478 Out[8]: '43'
479
479
480 # all-random (note for auto-testing)
480 # all-random (note for auto-testing)
481 """
481 """
482
482
483 if parameter_s:
483 if parameter_s:
484 arg = int(parameter_s)
484 arg = int(parameter_s)
485 else:
485 else:
486 arg = 'toggle'
486 arg = 'toggle'
487
487
488 if not arg in (0,1,2,'toggle'):
488 if not arg in (0,1,2,'toggle'):
489 error('Valid modes: (0->Off, 1->Smart, 2->Full')
489 error('Valid modes: (0->Off, 1->Smart, 2->Full')
490 return
490 return
491
491
492 if arg in (0,1,2):
492 if arg in (0,1,2):
493 self.shell.autocall = arg
493 self.shell.autocall = arg
494 else: # toggle
494 else: # toggle
495 if self.shell.autocall:
495 if self.shell.autocall:
496 self._magic_state.autocall_save = self.shell.autocall
496 self._magic_state.autocall_save = self.shell.autocall
497 self.shell.autocall = 0
497 self.shell.autocall = 0
498 else:
498 else:
499 try:
499 try:
500 self.shell.autocall = self._magic_state.autocall_save
500 self.shell.autocall = self._magic_state.autocall_save
501 except AttributeError:
501 except AttributeError:
502 self.shell.autocall = self._magic_state.autocall_save = 1
502 self.shell.autocall = self._magic_state.autocall_save = 1
503
503
504 print "Automatic calling is:",['OFF','Smart','Full'][self.shell.autocall]
504 print "Automatic calling is:",['OFF','Smart','Full'][self.shell.autocall]
505
505
506
506
507 def magic_page(self, parameter_s=''):
507 def magic_page(self, parameter_s=''):
508 """Pretty print the object and display it through a pager.
508 """Pretty print the object and display it through a pager.
509
509
510 %page [options] OBJECT
510 %page [options] OBJECT
511
511
512 If no object is given, use _ (last output).
512 If no object is given, use _ (last output).
513
513
514 Options:
514 Options:
515
515
516 -r: page str(object), don't pretty-print it."""
516 -r: page str(object), don't pretty-print it."""
517
517
518 # After a function contributed by Olivier Aubert, slightly modified.
518 # After a function contributed by Olivier Aubert, slightly modified.
519
519
520 # Process options/args
520 # Process options/args
521 opts,args = self.parse_options(parameter_s,'r')
521 opts,args = self.parse_options(parameter_s,'r')
522 raw = 'r' in opts
522 raw = 'r' in opts
523
523
524 oname = args and args or '_'
524 oname = args and args or '_'
525 info = self._ofind(oname)
525 info = self._ofind(oname)
526 if info['found']:
526 if info['found']:
527 txt = (raw and str or pformat)( info['obj'] )
527 txt = (raw and str or pformat)( info['obj'] )
528 page.page(txt)
528 page.page(txt)
529 else:
529 else:
530 print 'Object `%s` not found' % oname
530 print 'Object `%s` not found' % oname
531
531
532 def magic_profile(self, parameter_s=''):
532 def magic_profile(self, parameter_s=''):
533 """Print your currently active IPython profile."""
533 """Print your currently active IPython profile."""
534 from IPython.core.application import BaseIPythonApplication
534 from IPython.core.application import BaseIPythonApplication
535 if BaseIPythonApplication.initialized():
535 if BaseIPythonApplication.initialized():
536 print BaseIPythonApplication.instance().profile
536 print BaseIPythonApplication.instance().profile
537 else:
537 else:
538 error("profile is an application-level value, but you don't appear to be in an IPython application")
538 error("profile is an application-level value, but you don't appear to be in an IPython application")
539
539
540 def magic_pinfo(self, parameter_s='', namespaces=None):
540 def magic_pinfo(self, parameter_s='', namespaces=None):
541 """Provide detailed information about an object.
541 """Provide detailed information about an object.
542
542
543 '%pinfo object' is just a synonym for object? or ?object."""
543 '%pinfo object' is just a synonym for object? or ?object."""
544
544
545 #print 'pinfo par: <%s>' % parameter_s # dbg
545 #print 'pinfo par: <%s>' % parameter_s # dbg
546
546
547
547
548 # detail_level: 0 -> obj? , 1 -> obj??
548 # detail_level: 0 -> obj? , 1 -> obj??
549 detail_level = 0
549 detail_level = 0
550 # We need to detect if we got called as 'pinfo pinfo foo', which can
550 # We need to detect if we got called as 'pinfo pinfo foo', which can
551 # happen if the user types 'pinfo foo?' at the cmd line.
551 # happen if the user types 'pinfo foo?' at the cmd line.
552 pinfo,qmark1,oname,qmark2 = \
552 pinfo,qmark1,oname,qmark2 = \
553 re.match('(pinfo )?(\?*)(.*?)(\??$)',parameter_s).groups()
553 re.match('(pinfo )?(\?*)(.*?)(\??$)',parameter_s).groups()
554 if pinfo or qmark1 or qmark2:
554 if pinfo or qmark1 or qmark2:
555 detail_level = 1
555 detail_level = 1
556 if "*" in oname:
556 if "*" in oname:
557 self.magic_psearch(oname)
557 self.magic_psearch(oname)
558 else:
558 else:
559 self.shell._inspect('pinfo', oname, detail_level=detail_level,
559 self.shell._inspect('pinfo', oname, detail_level=detail_level,
560 namespaces=namespaces)
560 namespaces=namespaces)
561
561
562 def magic_pinfo2(self, parameter_s='', namespaces=None):
562 def magic_pinfo2(self, parameter_s='', namespaces=None):
563 """Provide extra detailed information about an object.
563 """Provide extra detailed information about an object.
564
564
565 '%pinfo2 object' is just a synonym for object?? or ??object."""
565 '%pinfo2 object' is just a synonym for object?? or ??object."""
566 self.shell._inspect('pinfo', parameter_s, detail_level=1,
566 self.shell._inspect('pinfo', parameter_s, detail_level=1,
567 namespaces=namespaces)
567 namespaces=namespaces)
568
568
569 @skip_doctest
569 @skip_doctest
570 def magic_pdef(self, parameter_s='', namespaces=None):
570 def magic_pdef(self, parameter_s='', namespaces=None):
571 """Print the definition header for any callable object.
571 """Print the definition header for any callable object.
572
572
573 If the object is a class, print the constructor information.
573 If the object is a class, print the constructor information.
574
574
575 Examples
575 Examples
576 --------
576 --------
577 ::
577 ::
578
578
579 In [3]: %pdef urllib.urlopen
579 In [3]: %pdef urllib.urlopen
580 urllib.urlopen(url, data=None, proxies=None)
580 urllib.urlopen(url, data=None, proxies=None)
581 """
581 """
582 self._inspect('pdef',parameter_s, namespaces)
582 self._inspect('pdef',parameter_s, namespaces)
583
583
584 def magic_pdoc(self, parameter_s='', namespaces=None):
584 def magic_pdoc(self, parameter_s='', namespaces=None):
585 """Print the docstring for an object.
585 """Print the docstring for an object.
586
586
587 If the given object is a class, it will print both the class and the
587 If the given object is a class, it will print both the class and the
588 constructor docstrings."""
588 constructor docstrings."""
589 self._inspect('pdoc',parameter_s, namespaces)
589 self._inspect('pdoc',parameter_s, namespaces)
590
590
591 def magic_psource(self, parameter_s='', namespaces=None):
591 def magic_psource(self, parameter_s='', namespaces=None):
592 """Print (or run through pager) the source code for an object."""
592 """Print (or run through pager) the source code for an object."""
593 self._inspect('psource',parameter_s, namespaces)
593 self._inspect('psource',parameter_s, namespaces)
594
594
595 def magic_pfile(self, parameter_s=''):
595 def magic_pfile(self, parameter_s=''):
596 """Print (or run through pager) the file where an object is defined.
596 """Print (or run through pager) the file where an object is defined.
597
597
598 The file opens at the line where the object definition begins. IPython
598 The file opens at the line where the object definition begins. IPython
599 will honor the environment variable PAGER if set, and otherwise will
599 will honor the environment variable PAGER if set, and otherwise will
600 do its best to print the file in a convenient form.
600 do its best to print the file in a convenient form.
601
601
602 If the given argument is not an object currently defined, IPython will
602 If the given argument is not an object currently defined, IPython will
603 try to interpret it as a filename (automatically adding a .py extension
603 try to interpret it as a filename (automatically adding a .py extension
604 if needed). You can thus use %pfile as a syntax highlighting code
604 if needed). You can thus use %pfile as a syntax highlighting code
605 viewer."""
605 viewer."""
606
606
607 # first interpret argument as an object name
607 # first interpret argument as an object name
608 out = self._inspect('pfile',parameter_s)
608 out = self._inspect('pfile',parameter_s)
609 # if not, try the input as a filename
609 # if not, try the input as a filename
610 if out == 'not found':
610 if out == 'not found':
611 try:
611 try:
612 filename = get_py_filename(parameter_s)
612 filename = get_py_filename(parameter_s)
613 except IOError,msg:
613 except IOError,msg:
614 print msg
614 print msg
615 return
615 return
616 page.page(self.shell.inspector.format(open(filename).read()))
616 page.page(self.shell.inspector.format(open(filename).read()))
617
617
618 def magic_psearch(self, parameter_s=''):
618 def magic_psearch(self, parameter_s=''):
619 """Search for object in namespaces by wildcard.
619 """Search for object in namespaces by wildcard.
620
620
621 %psearch [options] PATTERN [OBJECT TYPE]
621 %psearch [options] PATTERN [OBJECT TYPE]
622
622
623 Note: ? can be used as a synonym for %psearch, at the beginning or at
623 Note: ? can be used as a synonym for %psearch, at the beginning or at
624 the end: both a*? and ?a* are equivalent to '%psearch a*'. Still, the
624 the end: both a*? and ?a* are equivalent to '%psearch a*'. Still, the
625 rest of the command line must be unchanged (options come first), so
625 rest of the command line must be unchanged (options come first), so
626 for example the following forms are equivalent
626 for example the following forms are equivalent
627
627
628 %psearch -i a* function
628 %psearch -i a* function
629 -i a* function?
629 -i a* function?
630 ?-i a* function
630 ?-i a* function
631
631
632 Arguments:
632 Arguments:
633
633
634 PATTERN
634 PATTERN
635
635
636 where PATTERN is a string containing * as a wildcard similar to its
636 where PATTERN is a string containing * as a wildcard similar to its
637 use in a shell. The pattern is matched in all namespaces on the
637 use in a shell. The pattern is matched in all namespaces on the
638 search path. By default objects starting with a single _ are not
638 search path. By default objects starting with a single _ are not
639 matched, many IPython generated objects have a single
639 matched, many IPython generated objects have a single
640 underscore. The default is case insensitive matching. Matching is
640 underscore. The default is case insensitive matching. Matching is
641 also done on the attributes of objects and not only on the objects
641 also done on the attributes of objects and not only on the objects
642 in a module.
642 in a module.
643
643
644 [OBJECT TYPE]
644 [OBJECT TYPE]
645
645
646 Is the name of a python type from the types module. The name is
646 Is the name of a python type from the types module. The name is
647 given in lowercase without the ending type, ex. StringType is
647 given in lowercase without the ending type, ex. StringType is
648 written string. By adding a type here only objects matching the
648 written string. By adding a type here only objects matching the
649 given type are matched. Using all here makes the pattern match all
649 given type are matched. Using all here makes the pattern match all
650 types (this is the default).
650 types (this is the default).
651
651
652 Options:
652 Options:
653
653
654 -a: makes the pattern match even objects whose names start with a
654 -a: makes the pattern match even objects whose names start with a
655 single underscore. These names are normally omitted from the
655 single underscore. These names are normally omitted from the
656 search.
656 search.
657
657
658 -i/-c: make the pattern case insensitive/sensitive. If neither of
658 -i/-c: make the pattern case insensitive/sensitive. If neither of
659 these options are given, the default is read from your configuration
659 these options are given, the default is read from your configuration
660 file, with the option ``InteractiveShell.wildcards_case_sensitive``.
660 file, with the option ``InteractiveShell.wildcards_case_sensitive``.
661 If this option is not specified in your configuration file, IPython's
661 If this option is not specified in your configuration file, IPython's
662 internal default is to do a case sensitive search.
662 internal default is to do a case sensitive search.
663
663
664 -e/-s NAMESPACE: exclude/search a given namespace. The pattern you
664 -e/-s NAMESPACE: exclude/search a given namespace. The pattern you
665 specify can be searched in any of the following namespaces:
665 specify can be searched in any of the following namespaces:
666 'builtin', 'user', 'user_global','internal', 'alias', where
666 'builtin', 'user', 'user_global','internal', 'alias', where
667 'builtin' and 'user' are the search defaults. Note that you should
667 'builtin' and 'user' are the search defaults. Note that you should
668 not use quotes when specifying namespaces.
668 not use quotes when specifying namespaces.
669
669
670 'Builtin' contains the python module builtin, 'user' contains all
670 'Builtin' contains the python module builtin, 'user' contains all
671 user data, 'alias' only contain the shell aliases and no python
671 user data, 'alias' only contain the shell aliases and no python
672 objects, 'internal' contains objects used by IPython. The
672 objects, 'internal' contains objects used by IPython. The
673 'user_global' namespace is only used by embedded IPython instances,
673 'user_global' namespace is only used by embedded IPython instances,
674 and it contains module-level globals. You can add namespaces to the
674 and it contains module-level globals. You can add namespaces to the
675 search with -s or exclude them with -e (these options can be given
675 search with -s or exclude them with -e (these options can be given
676 more than once).
676 more than once).
677
677
678 Examples
678 Examples
679 --------
679 --------
680 ::
680 ::
681
681
682 %psearch a* -> objects beginning with an a
682 %psearch a* -> objects beginning with an a
683 %psearch -e builtin a* -> objects NOT in the builtin space starting in a
683 %psearch -e builtin a* -> objects NOT in the builtin space starting in a
684 %psearch a* function -> all functions beginning with an a
684 %psearch a* function -> all functions beginning with an a
685 %psearch re.e* -> objects beginning with an e in module re
685 %psearch re.e* -> objects beginning with an e in module re
686 %psearch r*.e* -> objects that start with e in modules starting in r
686 %psearch r*.e* -> objects that start with e in modules starting in r
687 %psearch r*.* string -> all strings in modules beginning with r
687 %psearch r*.* string -> all strings in modules beginning with r
688
688
689 Case sensitive search::
689 Case sensitive search::
690
690
691 %psearch -c a* list all object beginning with lower case a
691 %psearch -c a* list all object beginning with lower case a
692
692
693 Show objects beginning with a single _::
693 Show objects beginning with a single _::
694
694
695 %psearch -a _* list objects beginning with a single underscore"""
695 %psearch -a _* list objects beginning with a single underscore"""
696 try:
696 try:
697 parameter_s.encode('ascii')
697 parameter_s.encode('ascii')
698 except UnicodeEncodeError:
698 except UnicodeEncodeError:
699 print 'Python identifiers can only contain ascii characters.'
699 print 'Python identifiers can only contain ascii characters.'
700 return
700 return
701
701
702 # default namespaces to be searched
702 # default namespaces to be searched
703 def_search = ['user_local', 'user_global', 'builtin']
703 def_search = ['user_local', 'user_global', 'builtin']
704
704
705 # Process options/args
705 # Process options/args
706 opts,args = self.parse_options(parameter_s,'cias:e:',list_all=True)
706 opts,args = self.parse_options(parameter_s,'cias:e:',list_all=True)
707 opt = opts.get
707 opt = opts.get
708 shell = self.shell
708 shell = self.shell
709 psearch = shell.inspector.psearch
709 psearch = shell.inspector.psearch
710
710
711 # select case options
711 # select case options
712 if opts.has_key('i'):
712 if opts.has_key('i'):
713 ignore_case = True
713 ignore_case = True
714 elif opts.has_key('c'):
714 elif opts.has_key('c'):
715 ignore_case = False
715 ignore_case = False
716 else:
716 else:
717 ignore_case = not shell.wildcards_case_sensitive
717 ignore_case = not shell.wildcards_case_sensitive
718
718
719 # Build list of namespaces to search from user options
719 # Build list of namespaces to search from user options
720 def_search.extend(opt('s',[]))
720 def_search.extend(opt('s',[]))
721 ns_exclude = ns_exclude=opt('e',[])
721 ns_exclude = ns_exclude=opt('e',[])
722 ns_search = [nm for nm in def_search if nm not in ns_exclude]
722 ns_search = [nm for nm in def_search if nm not in ns_exclude]
723
723
724 # Call the actual search
724 # Call the actual search
725 try:
725 try:
726 psearch(args,shell.ns_table,ns_search,
726 psearch(args,shell.ns_table,ns_search,
727 show_all=opt('a'),ignore_case=ignore_case)
727 show_all=opt('a'),ignore_case=ignore_case)
728 except:
728 except:
729 shell.showtraceback()
729 shell.showtraceback()
730
730
731 @skip_doctest
731 @skip_doctest
732 def magic_who_ls(self, parameter_s=''):
732 def magic_who_ls(self, parameter_s=''):
733 """Return a sorted list of all interactive variables.
733 """Return a sorted list of all interactive variables.
734
734
735 If arguments are given, only variables of types matching these
735 If arguments are given, only variables of types matching these
736 arguments are returned.
736 arguments are returned.
737
737
738 Examples
738 Examples
739 --------
739 --------
740
740
741 Define two variables and list them with who_ls::
741 Define two variables and list them with who_ls::
742
742
743 In [1]: alpha = 123
743 In [1]: alpha = 123
744
744
745 In [2]: beta = 'test'
745 In [2]: beta = 'test'
746
746
747 In [3]: %who_ls
747 In [3]: %who_ls
748 Out[3]: ['alpha', 'beta']
748 Out[3]: ['alpha', 'beta']
749
749
750 In [4]: %who_ls int
750 In [4]: %who_ls int
751 Out[4]: ['alpha']
751 Out[4]: ['alpha']
752
752
753 In [5]: %who_ls str
753 In [5]: %who_ls str
754 Out[5]: ['beta']
754 Out[5]: ['beta']
755 """
755 """
756
756
757 user_ns = self.shell.user_ns
757 user_ns = self.shell.user_ns
758 user_ns_hidden = self.shell.user_ns_hidden
758 user_ns_hidden = self.shell.user_ns_hidden
759 out = [ i for i in user_ns
759 out = [ i for i in user_ns
760 if not i.startswith('_') \
760 if not i.startswith('_') \
761 and not i in user_ns_hidden ]
761 and not i in user_ns_hidden ]
762
762
763 typelist = parameter_s.split()
763 typelist = parameter_s.split()
764 if typelist:
764 if typelist:
765 typeset = set(typelist)
765 typeset = set(typelist)
766 out = [i for i in out if type(user_ns[i]).__name__ in typeset]
766 out = [i for i in out if type(user_ns[i]).__name__ in typeset]
767
767
768 out.sort()
768 out.sort()
769 return out
769 return out
770
770
771 @skip_doctest
771 @skip_doctest
772 def magic_who(self, parameter_s=''):
772 def magic_who(self, parameter_s=''):
773 """Print all interactive variables, with some minimal formatting.
773 """Print all interactive variables, with some minimal formatting.
774
774
775 If any arguments are given, only variables whose type matches one of
775 If any arguments are given, only variables whose type matches one of
776 these are printed. For example::
776 these are printed. For example::
777
777
778 %who function str
778 %who function str
779
779
780 will only list functions and strings, excluding all other types of
780 will only list functions and strings, excluding all other types of
781 variables. To find the proper type names, simply use type(var) at a
781 variables. To find the proper type names, simply use type(var) at a
782 command line to see how python prints type names. For example:
782 command line to see how python prints type names. For example:
783
783
784 ::
784 ::
785
785
786 In [1]: type('hello')\\
786 In [1]: type('hello')\\
787 Out[1]: <type 'str'>
787 Out[1]: <type 'str'>
788
788
789 indicates that the type name for strings is 'str'.
789 indicates that the type name for strings is 'str'.
790
790
791 ``%who`` always excludes executed names loaded through your configuration
791 ``%who`` always excludes executed names loaded through your configuration
792 file and things which are internal to IPython.
792 file and things which are internal to IPython.
793
793
794 This is deliberate, as typically you may load many modules and the
794 This is deliberate, as typically you may load many modules and the
795 purpose of %who is to show you only what you've manually defined.
795 purpose of %who is to show you only what you've manually defined.
796
796
797 Examples
797 Examples
798 --------
798 --------
799
799
800 Define two variables and list them with who::
800 Define two variables and list them with who::
801
801
802 In [1]: alpha = 123
802 In [1]: alpha = 123
803
803
804 In [2]: beta = 'test'
804 In [2]: beta = 'test'
805
805
806 In [3]: %who
806 In [3]: %who
807 alpha beta
807 alpha beta
808
808
809 In [4]: %who int
809 In [4]: %who int
810 alpha
810 alpha
811
811
812 In [5]: %who str
812 In [5]: %who str
813 beta
813 beta
814 """
814 """
815
815
816 varlist = self.magic_who_ls(parameter_s)
816 varlist = self.magic_who_ls(parameter_s)
817 if not varlist:
817 if not varlist:
818 if parameter_s:
818 if parameter_s:
819 print 'No variables match your requested type.'
819 print 'No variables match your requested type.'
820 else:
820 else:
821 print 'Interactive namespace is empty.'
821 print 'Interactive namespace is empty.'
822 return
822 return
823
823
824 # if we have variables, move on...
824 # if we have variables, move on...
825 count = 0
825 count = 0
826 for i in varlist:
826 for i in varlist:
827 print i+'\t',
827 print i+'\t',
828 count += 1
828 count += 1
829 if count > 8:
829 if count > 8:
830 count = 0
830 count = 0
831 print
831 print
832 print
832 print
833
833
834 @skip_doctest
834 @skip_doctest
835 def magic_whos(self, parameter_s=''):
835 def magic_whos(self, parameter_s=''):
836 """Like %who, but gives some extra information about each variable.
836 """Like %who, but gives some extra information about each variable.
837
837
838 The same type filtering of %who can be applied here.
838 The same type filtering of %who can be applied here.
839
839
840 For all variables, the type is printed. Additionally it prints:
840 For all variables, the type is printed. Additionally it prints:
841
841
842 - For {},[],(): their length.
842 - For {},[],(): their length.
843
843
844 - For numpy arrays, a summary with shape, number of
844 - For numpy arrays, a summary with shape, number of
845 elements, typecode and size in memory.
845 elements, typecode and size in memory.
846
846
847 - Everything else: a string representation, snipping their middle if
847 - Everything else: a string representation, snipping their middle if
848 too long.
848 too long.
849
849
850 Examples
850 Examples
851 --------
851 --------
852
852
853 Define two variables and list them with whos::
853 Define two variables and list them with whos::
854
854
855 In [1]: alpha = 123
855 In [1]: alpha = 123
856
856
857 In [2]: beta = 'test'
857 In [2]: beta = 'test'
858
858
859 In [3]: %whos
859 In [3]: %whos
860 Variable Type Data/Info
860 Variable Type Data/Info
861 --------------------------------
861 --------------------------------
862 alpha int 123
862 alpha int 123
863 beta str test
863 beta str test
864 """
864 """
865
865
866 varnames = self.magic_who_ls(parameter_s)
866 varnames = self.magic_who_ls(parameter_s)
867 if not varnames:
867 if not varnames:
868 if parameter_s:
868 if parameter_s:
869 print 'No variables match your requested type.'
869 print 'No variables match your requested type.'
870 else:
870 else:
871 print 'Interactive namespace is empty.'
871 print 'Interactive namespace is empty.'
872 return
872 return
873
873
874 # if we have variables, move on...
874 # if we have variables, move on...
875
875
876 # for these types, show len() instead of data:
876 # for these types, show len() instead of data:
877 seq_types = ['dict', 'list', 'tuple']
877 seq_types = ['dict', 'list', 'tuple']
878
878
879 # for numpy arrays, display summary info
879 # for numpy arrays, display summary info
880 ndarray_type = None
880 ndarray_type = None
881 if 'numpy' in sys.modules:
881 if 'numpy' in sys.modules:
882 try:
882 try:
883 from numpy import ndarray
883 from numpy import ndarray
884 except ImportError:
884 except ImportError:
885 pass
885 pass
886 else:
886 else:
887 ndarray_type = ndarray.__name__
887 ndarray_type = ndarray.__name__
888
888
889 # Find all variable names and types so we can figure out column sizes
889 # Find all variable names and types so we can figure out column sizes
890 def get_vars(i):
890 def get_vars(i):
891 return self.shell.user_ns[i]
891 return self.shell.user_ns[i]
892
892
893 # some types are well known and can be shorter
893 # some types are well known and can be shorter
894 abbrevs = {'IPython.core.macro.Macro' : 'Macro'}
894 abbrevs = {'IPython.core.macro.Macro' : 'Macro'}
895 def type_name(v):
895 def type_name(v):
896 tn = type(v).__name__
896 tn = type(v).__name__
897 return abbrevs.get(tn,tn)
897 return abbrevs.get(tn,tn)
898
898
899 varlist = map(get_vars,varnames)
899 varlist = map(get_vars,varnames)
900
900
901 typelist = []
901 typelist = []
902 for vv in varlist:
902 for vv in varlist:
903 tt = type_name(vv)
903 tt = type_name(vv)
904
904
905 if tt=='instance':
905 if tt=='instance':
906 typelist.append( abbrevs.get(str(vv.__class__),
906 typelist.append( abbrevs.get(str(vv.__class__),
907 str(vv.__class__)))
907 str(vv.__class__)))
908 else:
908 else:
909 typelist.append(tt)
909 typelist.append(tt)
910
910
911 # column labels and # of spaces as separator
911 # column labels and # of spaces as separator
912 varlabel = 'Variable'
912 varlabel = 'Variable'
913 typelabel = 'Type'
913 typelabel = 'Type'
914 datalabel = 'Data/Info'
914 datalabel = 'Data/Info'
915 colsep = 3
915 colsep = 3
916 # variable format strings
916 # variable format strings
917 vformat = "{0:<{varwidth}}{1:<{typewidth}}"
917 vformat = "{0:<{varwidth}}{1:<{typewidth}}"
918 aformat = "%s: %s elems, type `%s`, %s bytes"
918 aformat = "%s: %s elems, type `%s`, %s bytes"
919 # find the size of the columns to format the output nicely
919 # find the size of the columns to format the output nicely
920 varwidth = max(max(map(len,varnames)), len(varlabel)) + colsep
920 varwidth = max(max(map(len,varnames)), len(varlabel)) + colsep
921 typewidth = max(max(map(len,typelist)), len(typelabel)) + colsep
921 typewidth = max(max(map(len,typelist)), len(typelabel)) + colsep
922 # table header
922 # table header
923 print varlabel.ljust(varwidth) + typelabel.ljust(typewidth) + \
923 print varlabel.ljust(varwidth) + typelabel.ljust(typewidth) + \
924 ' '+datalabel+'\n' + '-'*(varwidth+typewidth+len(datalabel)+1)
924 ' '+datalabel+'\n' + '-'*(varwidth+typewidth+len(datalabel)+1)
925 # and the table itself
925 # and the table itself
926 kb = 1024
926 kb = 1024
927 Mb = 1048576 # kb**2
927 Mb = 1048576 # kb**2
928 for vname,var,vtype in zip(varnames,varlist,typelist):
928 for vname,var,vtype in zip(varnames,varlist,typelist):
929 print vformat.format(vname, vtype, varwidth=varwidth, typewidth=typewidth),
929 print vformat.format(vname, vtype, varwidth=varwidth, typewidth=typewidth),
930 if vtype in seq_types:
930 if vtype in seq_types:
931 print "n="+str(len(var))
931 print "n="+str(len(var))
932 elif vtype == ndarray_type:
932 elif vtype == ndarray_type:
933 vshape = str(var.shape).replace(',','').replace(' ','x')[1:-1]
933 vshape = str(var.shape).replace(',','').replace(' ','x')[1:-1]
934 if vtype==ndarray_type:
934 if vtype==ndarray_type:
935 # numpy
935 # numpy
936 vsize = var.size
936 vsize = var.size
937 vbytes = vsize*var.itemsize
937 vbytes = vsize*var.itemsize
938 vdtype = var.dtype
938 vdtype = var.dtype
939
939
940 if vbytes < 100000:
940 if vbytes < 100000:
941 print aformat % (vshape,vsize,vdtype,vbytes)
941 print aformat % (vshape,vsize,vdtype,vbytes)
942 else:
942 else:
943 print aformat % (vshape,vsize,vdtype,vbytes),
943 print aformat % (vshape,vsize,vdtype,vbytes),
944 if vbytes < Mb:
944 if vbytes < Mb:
945 print '(%s kb)' % (vbytes/kb,)
945 print '(%s kb)' % (vbytes/kb,)
946 else:
946 else:
947 print '(%s Mb)' % (vbytes/Mb,)
947 print '(%s Mb)' % (vbytes/Mb,)
948 else:
948 else:
949 try:
949 try:
950 vstr = str(var)
950 vstr = str(var)
951 except UnicodeEncodeError:
951 except UnicodeEncodeError:
952 vstr = unicode(var).encode(sys.getdefaultencoding(),
952 vstr = unicode(var).encode(py3compat.getdefaultencoding(),
953 'backslashreplace')
953 'backslashreplace')
954 vstr = vstr.replace('\n','\\n')
954 vstr = vstr.replace('\n','\\n')
955 if len(vstr) < 50:
955 if len(vstr) < 50:
956 print vstr
956 print vstr
957 else:
957 else:
958 print vstr[:25] + "<...>" + vstr[-25:]
958 print vstr[:25] + "<...>" + vstr[-25:]
959
959
960 def magic_reset(self, parameter_s=''):
960 def magic_reset(self, parameter_s=''):
961 """Resets the namespace by removing all names defined by the user, if
961 """Resets the namespace by removing all names defined by the user, if
962 called without arguments, or by removing some types of objects, such
962 called without arguments, or by removing some types of objects, such
963 as everything currently in IPython's In[] and Out[] containers (see
963 as everything currently in IPython's In[] and Out[] containers (see
964 the parameters for details).
964 the parameters for details).
965
965
966 Parameters
966 Parameters
967 ----------
967 ----------
968 -f : force reset without asking for confirmation.
968 -f : force reset without asking for confirmation.
969
969
970 -s : 'Soft' reset: Only clears your namespace, leaving history intact.
970 -s : 'Soft' reset: Only clears your namespace, leaving history intact.
971 References to objects may be kept. By default (without this option),
971 References to objects may be kept. By default (without this option),
972 we do a 'hard' reset, giving you a new session and removing all
972 we do a 'hard' reset, giving you a new session and removing all
973 references to objects from the current session.
973 references to objects from the current session.
974
974
975 in : reset input history
975 in : reset input history
976
976
977 out : reset output history
977 out : reset output history
978
978
979 dhist : reset directory history
979 dhist : reset directory history
980
980
981 array : reset only variables that are NumPy arrays
981 array : reset only variables that are NumPy arrays
982
982
983 See Also
983 See Also
984 --------
984 --------
985 magic_reset_selective : invoked as ``%reset_selective``
985 magic_reset_selective : invoked as ``%reset_selective``
986
986
987 Examples
987 Examples
988 --------
988 --------
989 ::
989 ::
990
990
991 In [6]: a = 1
991 In [6]: a = 1
992
992
993 In [7]: a
993 In [7]: a
994 Out[7]: 1
994 Out[7]: 1
995
995
996 In [8]: 'a' in _ip.user_ns
996 In [8]: 'a' in _ip.user_ns
997 Out[8]: True
997 Out[8]: True
998
998
999 In [9]: %reset -f
999 In [9]: %reset -f
1000
1000
1001 In [1]: 'a' in _ip.user_ns
1001 In [1]: 'a' in _ip.user_ns
1002 Out[1]: False
1002 Out[1]: False
1003
1003
1004 In [2]: %reset -f in
1004 In [2]: %reset -f in
1005 Flushing input history
1005 Flushing input history
1006
1006
1007 In [3]: %reset -f dhist in
1007 In [3]: %reset -f dhist in
1008 Flushing directory history
1008 Flushing directory history
1009 Flushing input history
1009 Flushing input history
1010
1010
1011 Notes
1011 Notes
1012 -----
1012 -----
1013 Calling this magic from clients that do not implement standard input,
1013 Calling this magic from clients that do not implement standard input,
1014 such as the ipython notebook interface, will reset the namespace
1014 such as the ipython notebook interface, will reset the namespace
1015 without confirmation.
1015 without confirmation.
1016 """
1016 """
1017 opts, args = self.parse_options(parameter_s,'sf', mode='list')
1017 opts, args = self.parse_options(parameter_s,'sf', mode='list')
1018 if 'f' in opts:
1018 if 'f' in opts:
1019 ans = True
1019 ans = True
1020 else:
1020 else:
1021 try:
1021 try:
1022 ans = self.shell.ask_yes_no(
1022 ans = self.shell.ask_yes_no(
1023 "Once deleted, variables cannot be recovered. Proceed (y/[n])? ", default='n')
1023 "Once deleted, variables cannot be recovered. Proceed (y/[n])? ", default='n')
1024 except StdinNotImplementedError:
1024 except StdinNotImplementedError:
1025 ans = True
1025 ans = True
1026 if not ans:
1026 if not ans:
1027 print 'Nothing done.'
1027 print 'Nothing done.'
1028 return
1028 return
1029
1029
1030 if 's' in opts: # Soft reset
1030 if 's' in opts: # Soft reset
1031 user_ns = self.shell.user_ns
1031 user_ns = self.shell.user_ns
1032 for i in self.magic_who_ls():
1032 for i in self.magic_who_ls():
1033 del(user_ns[i])
1033 del(user_ns[i])
1034 elif len(args) == 0: # Hard reset
1034 elif len(args) == 0: # Hard reset
1035 self.shell.reset(new_session = False)
1035 self.shell.reset(new_session = False)
1036
1036
1037 # reset in/out/dhist/array: previously extensinions/clearcmd.py
1037 # reset in/out/dhist/array: previously extensinions/clearcmd.py
1038 ip = self.shell
1038 ip = self.shell
1039 user_ns = self.user_ns # local lookup, heavily used
1039 user_ns = self.user_ns # local lookup, heavily used
1040
1040
1041 for target in args:
1041 for target in args:
1042 target = target.lower() # make matches case insensitive
1042 target = target.lower() # make matches case insensitive
1043 if target == 'out':
1043 if target == 'out':
1044 print "Flushing output cache (%d entries)" % len(user_ns['_oh'])
1044 print "Flushing output cache (%d entries)" % len(user_ns['_oh'])
1045 self.displayhook.flush()
1045 self.displayhook.flush()
1046
1046
1047 elif target == 'in':
1047 elif target == 'in':
1048 print "Flushing input history"
1048 print "Flushing input history"
1049 pc = self.displayhook.prompt_count + 1
1049 pc = self.displayhook.prompt_count + 1
1050 for n in range(1, pc):
1050 for n in range(1, pc):
1051 key = '_i'+repr(n)
1051 key = '_i'+repr(n)
1052 user_ns.pop(key,None)
1052 user_ns.pop(key,None)
1053 user_ns.update(dict(_i=u'',_ii=u'',_iii=u''))
1053 user_ns.update(dict(_i=u'',_ii=u'',_iii=u''))
1054 hm = ip.history_manager
1054 hm = ip.history_manager
1055 # don't delete these, as %save and %macro depending on the length
1055 # don't delete these, as %save and %macro depending on the length
1056 # of these lists to be preserved
1056 # of these lists to be preserved
1057 hm.input_hist_parsed[:] = [''] * pc
1057 hm.input_hist_parsed[:] = [''] * pc
1058 hm.input_hist_raw[:] = [''] * pc
1058 hm.input_hist_raw[:] = [''] * pc
1059 # hm has internal machinery for _i,_ii,_iii, clear it out
1059 # hm has internal machinery for _i,_ii,_iii, clear it out
1060 hm._i = hm._ii = hm._iii = hm._i00 = u''
1060 hm._i = hm._ii = hm._iii = hm._i00 = u''
1061
1061
1062 elif target == 'array':
1062 elif target == 'array':
1063 # Support cleaning up numpy arrays
1063 # Support cleaning up numpy arrays
1064 try:
1064 try:
1065 from numpy import ndarray
1065 from numpy import ndarray
1066 # This must be done with items and not iteritems because we're
1066 # This must be done with items and not iteritems because we're
1067 # going to modify the dict in-place.
1067 # going to modify the dict in-place.
1068 for x,val in user_ns.items():
1068 for x,val in user_ns.items():
1069 if isinstance(val,ndarray):
1069 if isinstance(val,ndarray):
1070 del user_ns[x]
1070 del user_ns[x]
1071 except ImportError:
1071 except ImportError:
1072 print "reset array only works if Numpy is available."
1072 print "reset array only works if Numpy is available."
1073
1073
1074 elif target == 'dhist':
1074 elif target == 'dhist':
1075 print "Flushing directory history"
1075 print "Flushing directory history"
1076 del user_ns['_dh'][:]
1076 del user_ns['_dh'][:]
1077
1077
1078 else:
1078 else:
1079 print "Don't know how to reset ",
1079 print "Don't know how to reset ",
1080 print target + ", please run `%reset?` for details"
1080 print target + ", please run `%reset?` for details"
1081
1081
1082 gc.collect()
1082 gc.collect()
1083
1083
1084 def magic_reset_selective(self, parameter_s=''):
1084 def magic_reset_selective(self, parameter_s=''):
1085 """Resets the namespace by removing names defined by the user.
1085 """Resets the namespace by removing names defined by the user.
1086
1086
1087 Input/Output history are left around in case you need them.
1087 Input/Output history are left around in case you need them.
1088
1088
1089 %reset_selective [-f] regex
1089 %reset_selective [-f] regex
1090
1090
1091 No action is taken if regex is not included
1091 No action is taken if regex is not included
1092
1092
1093 Options
1093 Options
1094 -f : force reset without asking for confirmation.
1094 -f : force reset without asking for confirmation.
1095
1095
1096 See Also
1096 See Also
1097 --------
1097 --------
1098 magic_reset : invoked as ``%reset``
1098 magic_reset : invoked as ``%reset``
1099
1099
1100 Examples
1100 Examples
1101 --------
1101 --------
1102
1102
1103 We first fully reset the namespace so your output looks identical to
1103 We first fully reset the namespace so your output looks identical to
1104 this example for pedagogical reasons; in practice you do not need a
1104 this example for pedagogical reasons; in practice you do not need a
1105 full reset::
1105 full reset::
1106
1106
1107 In [1]: %reset -f
1107 In [1]: %reset -f
1108
1108
1109 Now, with a clean namespace we can make a few variables and use
1109 Now, with a clean namespace we can make a few variables and use
1110 ``%reset_selective`` to only delete names that match our regexp::
1110 ``%reset_selective`` to only delete names that match our regexp::
1111
1111
1112 In [2]: a=1; b=2; c=3; b1m=4; b2m=5; b3m=6; b4m=7; b2s=8
1112 In [2]: a=1; b=2; c=3; b1m=4; b2m=5; b3m=6; b4m=7; b2s=8
1113
1113
1114 In [3]: who_ls
1114 In [3]: who_ls
1115 Out[3]: ['a', 'b', 'b1m', 'b2m', 'b2s', 'b3m', 'b4m', 'c']
1115 Out[3]: ['a', 'b', 'b1m', 'b2m', 'b2s', 'b3m', 'b4m', 'c']
1116
1116
1117 In [4]: %reset_selective -f b[2-3]m
1117 In [4]: %reset_selective -f b[2-3]m
1118
1118
1119 In [5]: who_ls
1119 In [5]: who_ls
1120 Out[5]: ['a', 'b', 'b1m', 'b2s', 'b4m', 'c']
1120 Out[5]: ['a', 'b', 'b1m', 'b2s', 'b4m', 'c']
1121
1121
1122 In [6]: %reset_selective -f d
1122 In [6]: %reset_selective -f d
1123
1123
1124 In [7]: who_ls
1124 In [7]: who_ls
1125 Out[7]: ['a', 'b', 'b1m', 'b2s', 'b4m', 'c']
1125 Out[7]: ['a', 'b', 'b1m', 'b2s', 'b4m', 'c']
1126
1126
1127 In [8]: %reset_selective -f c
1127 In [8]: %reset_selective -f c
1128
1128
1129 In [9]: who_ls
1129 In [9]: who_ls
1130 Out[9]: ['a', 'b', 'b1m', 'b2s', 'b4m']
1130 Out[9]: ['a', 'b', 'b1m', 'b2s', 'b4m']
1131
1131
1132 In [10]: %reset_selective -f b
1132 In [10]: %reset_selective -f b
1133
1133
1134 In [11]: who_ls
1134 In [11]: who_ls
1135 Out[11]: ['a']
1135 Out[11]: ['a']
1136
1136
1137 Notes
1137 Notes
1138 -----
1138 -----
1139 Calling this magic from clients that do not implement standard input,
1139 Calling this magic from clients that do not implement standard input,
1140 such as the ipython notebook interface, will reset the namespace
1140 such as the ipython notebook interface, will reset the namespace
1141 without confirmation.
1141 without confirmation.
1142 """
1142 """
1143
1143
1144 opts, regex = self.parse_options(parameter_s,'f')
1144 opts, regex = self.parse_options(parameter_s,'f')
1145
1145
1146 if opts.has_key('f'):
1146 if opts.has_key('f'):
1147 ans = True
1147 ans = True
1148 else:
1148 else:
1149 try:
1149 try:
1150 ans = self.shell.ask_yes_no(
1150 ans = self.shell.ask_yes_no(
1151 "Once deleted, variables cannot be recovered. Proceed (y/[n])? ",
1151 "Once deleted, variables cannot be recovered. Proceed (y/[n])? ",
1152 default='n')
1152 default='n')
1153 except StdinNotImplementedError:
1153 except StdinNotImplementedError:
1154 ans = True
1154 ans = True
1155 if not ans:
1155 if not ans:
1156 print 'Nothing done.'
1156 print 'Nothing done.'
1157 return
1157 return
1158 user_ns = self.shell.user_ns
1158 user_ns = self.shell.user_ns
1159 if not regex:
1159 if not regex:
1160 print 'No regex pattern specified. Nothing done.'
1160 print 'No regex pattern specified. Nothing done.'
1161 return
1161 return
1162 else:
1162 else:
1163 try:
1163 try:
1164 m = re.compile(regex)
1164 m = re.compile(regex)
1165 except TypeError:
1165 except TypeError:
1166 raise TypeError('regex must be a string or compiled pattern')
1166 raise TypeError('regex must be a string or compiled pattern')
1167 for i in self.magic_who_ls():
1167 for i in self.magic_who_ls():
1168 if m.search(i):
1168 if m.search(i):
1169 del(user_ns[i])
1169 del(user_ns[i])
1170
1170
1171 def magic_xdel(self, parameter_s=''):
1171 def magic_xdel(self, parameter_s=''):
1172 """Delete a variable, trying to clear it from anywhere that
1172 """Delete a variable, trying to clear it from anywhere that
1173 IPython's machinery has references to it. By default, this uses
1173 IPython's machinery has references to it. By default, this uses
1174 the identity of the named object in the user namespace to remove
1174 the identity of the named object in the user namespace to remove
1175 references held under other names. The object is also removed
1175 references held under other names. The object is also removed
1176 from the output history.
1176 from the output history.
1177
1177
1178 Options
1178 Options
1179 -n : Delete the specified name from all namespaces, without
1179 -n : Delete the specified name from all namespaces, without
1180 checking their identity.
1180 checking their identity.
1181 """
1181 """
1182 opts, varname = self.parse_options(parameter_s,'n')
1182 opts, varname = self.parse_options(parameter_s,'n')
1183 try:
1183 try:
1184 self.shell.del_var(varname, ('n' in opts))
1184 self.shell.del_var(varname, ('n' in opts))
1185 except (NameError, ValueError) as e:
1185 except (NameError, ValueError) as e:
1186 print type(e).__name__ +": "+ str(e)
1186 print type(e).__name__ +": "+ str(e)
1187
1187
1188 def magic_logstart(self,parameter_s=''):
1188 def magic_logstart(self,parameter_s=''):
1189 """Start logging anywhere in a session.
1189 """Start logging anywhere in a session.
1190
1190
1191 %logstart [-o|-r|-t] [log_name [log_mode]]
1191 %logstart [-o|-r|-t] [log_name [log_mode]]
1192
1192
1193 If no name is given, it defaults to a file named 'ipython_log.py' in your
1193 If no name is given, it defaults to a file named 'ipython_log.py' in your
1194 current directory, in 'rotate' mode (see below).
1194 current directory, in 'rotate' mode (see below).
1195
1195
1196 '%logstart name' saves to file 'name' in 'backup' mode. It saves your
1196 '%logstart name' saves to file 'name' in 'backup' mode. It saves your
1197 history up to that point and then continues logging.
1197 history up to that point and then continues logging.
1198
1198
1199 %logstart takes a second optional parameter: logging mode. This can be one
1199 %logstart takes a second optional parameter: logging mode. This can be one
1200 of (note that the modes are given unquoted):\\
1200 of (note that the modes are given unquoted):\\
1201 append: well, that says it.\\
1201 append: well, that says it.\\
1202 backup: rename (if exists) to name~ and start name.\\
1202 backup: rename (if exists) to name~ and start name.\\
1203 global: single logfile in your home dir, appended to.\\
1203 global: single logfile in your home dir, appended to.\\
1204 over : overwrite existing log.\\
1204 over : overwrite existing log.\\
1205 rotate: create rotating logs name.1~, name.2~, etc.
1205 rotate: create rotating logs name.1~, name.2~, etc.
1206
1206
1207 Options:
1207 Options:
1208
1208
1209 -o: log also IPython's output. In this mode, all commands which
1209 -o: log also IPython's output. In this mode, all commands which
1210 generate an Out[NN] prompt are recorded to the logfile, right after
1210 generate an Out[NN] prompt are recorded to the logfile, right after
1211 their corresponding input line. The output lines are always
1211 their corresponding input line. The output lines are always
1212 prepended with a '#[Out]# ' marker, so that the log remains valid
1212 prepended with a '#[Out]# ' marker, so that the log remains valid
1213 Python code.
1213 Python code.
1214
1214
1215 Since this marker is always the same, filtering only the output from
1215 Since this marker is always the same, filtering only the output from
1216 a log is very easy, using for example a simple awk call::
1216 a log is very easy, using for example a simple awk call::
1217
1217
1218 awk -F'#\\[Out\\]# ' '{if($2) {print $2}}' ipython_log.py
1218 awk -F'#\\[Out\\]# ' '{if($2) {print $2}}' ipython_log.py
1219
1219
1220 -r: log 'raw' input. Normally, IPython's logs contain the processed
1220 -r: log 'raw' input. Normally, IPython's logs contain the processed
1221 input, so that user lines are logged in their final form, converted
1221 input, so that user lines are logged in their final form, converted
1222 into valid Python. For example, %Exit is logged as
1222 into valid Python. For example, %Exit is logged as
1223 _ip.magic("Exit"). If the -r flag is given, all input is logged
1223 _ip.magic("Exit"). If the -r flag is given, all input is logged
1224 exactly as typed, with no transformations applied.
1224 exactly as typed, with no transformations applied.
1225
1225
1226 -t: put timestamps before each input line logged (these are put in
1226 -t: put timestamps before each input line logged (these are put in
1227 comments)."""
1227 comments)."""
1228
1228
1229 opts,par = self.parse_options(parameter_s,'ort')
1229 opts,par = self.parse_options(parameter_s,'ort')
1230 log_output = 'o' in opts
1230 log_output = 'o' in opts
1231 log_raw_input = 'r' in opts
1231 log_raw_input = 'r' in opts
1232 timestamp = 't' in opts
1232 timestamp = 't' in opts
1233
1233
1234 logger = self.shell.logger
1234 logger = self.shell.logger
1235
1235
1236 # if no args are given, the defaults set in the logger constructor by
1236 # if no args are given, the defaults set in the logger constructor by
1237 # ipython remain valid
1237 # ipython remain valid
1238 if par:
1238 if par:
1239 try:
1239 try:
1240 logfname,logmode = par.split()
1240 logfname,logmode = par.split()
1241 except:
1241 except:
1242 logfname = par
1242 logfname = par
1243 logmode = 'backup'
1243 logmode = 'backup'
1244 else:
1244 else:
1245 logfname = logger.logfname
1245 logfname = logger.logfname
1246 logmode = logger.logmode
1246 logmode = logger.logmode
1247 # put logfname into rc struct as if it had been called on the command
1247 # put logfname into rc struct as if it had been called on the command
1248 # line, so it ends up saved in the log header Save it in case we need
1248 # line, so it ends up saved in the log header Save it in case we need
1249 # to restore it...
1249 # to restore it...
1250 old_logfile = self.shell.logfile
1250 old_logfile = self.shell.logfile
1251 if logfname:
1251 if logfname:
1252 logfname = os.path.expanduser(logfname)
1252 logfname = os.path.expanduser(logfname)
1253 self.shell.logfile = logfname
1253 self.shell.logfile = logfname
1254
1254
1255 loghead = '# IPython log file\n\n'
1255 loghead = '# IPython log file\n\n'
1256 try:
1256 try:
1257 started = logger.logstart(logfname,loghead,logmode,
1257 started = logger.logstart(logfname,loghead,logmode,
1258 log_output,timestamp,log_raw_input)
1258 log_output,timestamp,log_raw_input)
1259 except:
1259 except:
1260 self.shell.logfile = old_logfile
1260 self.shell.logfile = old_logfile
1261 warn("Couldn't start log: %s" % sys.exc_info()[1])
1261 warn("Couldn't start log: %s" % sys.exc_info()[1])
1262 else:
1262 else:
1263 # log input history up to this point, optionally interleaving
1263 # log input history up to this point, optionally interleaving
1264 # output if requested
1264 # output if requested
1265
1265
1266 if timestamp:
1266 if timestamp:
1267 # disable timestamping for the previous history, since we've
1267 # disable timestamping for the previous history, since we've
1268 # lost those already (no time machine here).
1268 # lost those already (no time machine here).
1269 logger.timestamp = False
1269 logger.timestamp = False
1270
1270
1271 if log_raw_input:
1271 if log_raw_input:
1272 input_hist = self.shell.history_manager.input_hist_raw
1272 input_hist = self.shell.history_manager.input_hist_raw
1273 else:
1273 else:
1274 input_hist = self.shell.history_manager.input_hist_parsed
1274 input_hist = self.shell.history_manager.input_hist_parsed
1275
1275
1276 if log_output:
1276 if log_output:
1277 log_write = logger.log_write
1277 log_write = logger.log_write
1278 output_hist = self.shell.history_manager.output_hist
1278 output_hist = self.shell.history_manager.output_hist
1279 for n in range(1,len(input_hist)-1):
1279 for n in range(1,len(input_hist)-1):
1280 log_write(input_hist[n].rstrip() + '\n')
1280 log_write(input_hist[n].rstrip() + '\n')
1281 if n in output_hist:
1281 if n in output_hist:
1282 log_write(repr(output_hist[n]),'output')
1282 log_write(repr(output_hist[n]),'output')
1283 else:
1283 else:
1284 logger.log_write('\n'.join(input_hist[1:]))
1284 logger.log_write('\n'.join(input_hist[1:]))
1285 logger.log_write('\n')
1285 logger.log_write('\n')
1286 if timestamp:
1286 if timestamp:
1287 # re-enable timestamping
1287 # re-enable timestamping
1288 logger.timestamp = True
1288 logger.timestamp = True
1289
1289
1290 print ('Activating auto-logging. '
1290 print ('Activating auto-logging. '
1291 'Current session state plus future input saved.')
1291 'Current session state plus future input saved.')
1292 logger.logstate()
1292 logger.logstate()
1293
1293
1294 def magic_logstop(self,parameter_s=''):
1294 def magic_logstop(self,parameter_s=''):
1295 """Fully stop logging and close log file.
1295 """Fully stop logging and close log file.
1296
1296
1297 In order to start logging again, a new %logstart call needs to be made,
1297 In order to start logging again, a new %logstart call needs to be made,
1298 possibly (though not necessarily) with a new filename, mode and other
1298 possibly (though not necessarily) with a new filename, mode and other
1299 options."""
1299 options."""
1300 self.logger.logstop()
1300 self.logger.logstop()
1301
1301
1302 def magic_logoff(self,parameter_s=''):
1302 def magic_logoff(self,parameter_s=''):
1303 """Temporarily stop logging.
1303 """Temporarily stop logging.
1304
1304
1305 You must have previously started logging."""
1305 You must have previously started logging."""
1306 self.shell.logger.switch_log(0)
1306 self.shell.logger.switch_log(0)
1307
1307
1308 def magic_logon(self,parameter_s=''):
1308 def magic_logon(self,parameter_s=''):
1309 """Restart logging.
1309 """Restart logging.
1310
1310
1311 This function is for restarting logging which you've temporarily
1311 This function is for restarting logging which you've temporarily
1312 stopped with %logoff. For starting logging for the first time, you
1312 stopped with %logoff. For starting logging for the first time, you
1313 must use the %logstart function, which allows you to specify an
1313 must use the %logstart function, which allows you to specify an
1314 optional log filename."""
1314 optional log filename."""
1315
1315
1316 self.shell.logger.switch_log(1)
1316 self.shell.logger.switch_log(1)
1317
1317
1318 def magic_logstate(self,parameter_s=''):
1318 def magic_logstate(self,parameter_s=''):
1319 """Print the status of the logging system."""
1319 """Print the status of the logging system."""
1320
1320
1321 self.shell.logger.logstate()
1321 self.shell.logger.logstate()
1322
1322
1323 def magic_pdb(self, parameter_s=''):
1323 def magic_pdb(self, parameter_s=''):
1324 """Control the automatic calling of the pdb interactive debugger.
1324 """Control the automatic calling of the pdb interactive debugger.
1325
1325
1326 Call as '%pdb on', '%pdb 1', '%pdb off' or '%pdb 0'. If called without
1326 Call as '%pdb on', '%pdb 1', '%pdb off' or '%pdb 0'. If called without
1327 argument it works as a toggle.
1327 argument it works as a toggle.
1328
1328
1329 When an exception is triggered, IPython can optionally call the
1329 When an exception is triggered, IPython can optionally call the
1330 interactive pdb debugger after the traceback printout. %pdb toggles
1330 interactive pdb debugger after the traceback printout. %pdb toggles
1331 this feature on and off.
1331 this feature on and off.
1332
1332
1333 The initial state of this feature is set in your configuration
1333 The initial state of this feature is set in your configuration
1334 file (the option is ``InteractiveShell.pdb``).
1334 file (the option is ``InteractiveShell.pdb``).
1335
1335
1336 If you want to just activate the debugger AFTER an exception has fired,
1336 If you want to just activate the debugger AFTER an exception has fired,
1337 without having to type '%pdb on' and rerunning your code, you can use
1337 without having to type '%pdb on' and rerunning your code, you can use
1338 the %debug magic."""
1338 the %debug magic."""
1339
1339
1340 par = parameter_s.strip().lower()
1340 par = parameter_s.strip().lower()
1341
1341
1342 if par:
1342 if par:
1343 try:
1343 try:
1344 new_pdb = {'off':0,'0':0,'on':1,'1':1}[par]
1344 new_pdb = {'off':0,'0':0,'on':1,'1':1}[par]
1345 except KeyError:
1345 except KeyError:
1346 print ('Incorrect argument. Use on/1, off/0, '
1346 print ('Incorrect argument. Use on/1, off/0, '
1347 'or nothing for a toggle.')
1347 'or nothing for a toggle.')
1348 return
1348 return
1349 else:
1349 else:
1350 # toggle
1350 # toggle
1351 new_pdb = not self.shell.call_pdb
1351 new_pdb = not self.shell.call_pdb
1352
1352
1353 # set on the shell
1353 # set on the shell
1354 self.shell.call_pdb = new_pdb
1354 self.shell.call_pdb = new_pdb
1355 print 'Automatic pdb calling has been turned',on_off(new_pdb)
1355 print 'Automatic pdb calling has been turned',on_off(new_pdb)
1356
1356
1357 def magic_debug(self, parameter_s=''):
1357 def magic_debug(self, parameter_s=''):
1358 """Activate the interactive debugger in post-mortem mode.
1358 """Activate the interactive debugger in post-mortem mode.
1359
1359
1360 If an exception has just occurred, this lets you inspect its stack
1360 If an exception has just occurred, this lets you inspect its stack
1361 frames interactively. Note that this will always work only on the last
1361 frames interactively. Note that this will always work only on the last
1362 traceback that occurred, so you must call this quickly after an
1362 traceback that occurred, so you must call this quickly after an
1363 exception that you wish to inspect has fired, because if another one
1363 exception that you wish to inspect has fired, because if another one
1364 occurs, it clobbers the previous one.
1364 occurs, it clobbers the previous one.
1365
1365
1366 If you want IPython to automatically do this on every exception, see
1366 If you want IPython to automatically do this on every exception, see
1367 the %pdb magic for more details.
1367 the %pdb magic for more details.
1368 """
1368 """
1369 self.shell.debugger(force=True)
1369 self.shell.debugger(force=True)
1370
1370
1371 @skip_doctest
1371 @skip_doctest
1372 def magic_prun(self, parameter_s ='',user_mode=1,
1372 def magic_prun(self, parameter_s ='',user_mode=1,
1373 opts=None,arg_lst=None,prog_ns=None):
1373 opts=None,arg_lst=None,prog_ns=None):
1374
1374
1375 """Run a statement through the python code profiler.
1375 """Run a statement through the python code profiler.
1376
1376
1377 Usage:
1377 Usage:
1378 %prun [options] statement
1378 %prun [options] statement
1379
1379
1380 The given statement (which doesn't require quote marks) is run via the
1380 The given statement (which doesn't require quote marks) is run via the
1381 python profiler in a manner similar to the profile.run() function.
1381 python profiler in a manner similar to the profile.run() function.
1382 Namespaces are internally managed to work correctly; profile.run
1382 Namespaces are internally managed to work correctly; profile.run
1383 cannot be used in IPython because it makes certain assumptions about
1383 cannot be used in IPython because it makes certain assumptions about
1384 namespaces which do not hold under IPython.
1384 namespaces which do not hold under IPython.
1385
1385
1386 Options:
1386 Options:
1387
1387
1388 -l <limit>: you can place restrictions on what or how much of the
1388 -l <limit>: you can place restrictions on what or how much of the
1389 profile gets printed. The limit value can be:
1389 profile gets printed. The limit value can be:
1390
1390
1391 * A string: only information for function names containing this string
1391 * A string: only information for function names containing this string
1392 is printed.
1392 is printed.
1393
1393
1394 * An integer: only these many lines are printed.
1394 * An integer: only these many lines are printed.
1395
1395
1396 * A float (between 0 and 1): this fraction of the report is printed
1396 * A float (between 0 and 1): this fraction of the report is printed
1397 (for example, use a limit of 0.4 to see the topmost 40% only).
1397 (for example, use a limit of 0.4 to see the topmost 40% only).
1398
1398
1399 You can combine several limits with repeated use of the option. For
1399 You can combine several limits with repeated use of the option. For
1400 example, '-l __init__ -l 5' will print only the topmost 5 lines of
1400 example, '-l __init__ -l 5' will print only the topmost 5 lines of
1401 information about class constructors.
1401 information about class constructors.
1402
1402
1403 -r: return the pstats.Stats object generated by the profiling. This
1403 -r: return the pstats.Stats object generated by the profiling. This
1404 object has all the information about the profile in it, and you can
1404 object has all the information about the profile in it, and you can
1405 later use it for further analysis or in other functions.
1405 later use it for further analysis or in other functions.
1406
1406
1407 -s <key>: sort profile by given key. You can provide more than one key
1407 -s <key>: sort profile by given key. You can provide more than one key
1408 by using the option several times: '-s key1 -s key2 -s key3...'. The
1408 by using the option several times: '-s key1 -s key2 -s key3...'. The
1409 default sorting key is 'time'.
1409 default sorting key is 'time'.
1410
1410
1411 The following is copied verbatim from the profile documentation
1411 The following is copied verbatim from the profile documentation
1412 referenced below:
1412 referenced below:
1413
1413
1414 When more than one key is provided, additional keys are used as
1414 When more than one key is provided, additional keys are used as
1415 secondary criteria when the there is equality in all keys selected
1415 secondary criteria when the there is equality in all keys selected
1416 before them.
1416 before them.
1417
1417
1418 Abbreviations can be used for any key names, as long as the
1418 Abbreviations can be used for any key names, as long as the
1419 abbreviation is unambiguous. The following are the keys currently
1419 abbreviation is unambiguous. The following are the keys currently
1420 defined:
1420 defined:
1421
1421
1422 Valid Arg Meaning
1422 Valid Arg Meaning
1423 "calls" call count
1423 "calls" call count
1424 "cumulative" cumulative time
1424 "cumulative" cumulative time
1425 "file" file name
1425 "file" file name
1426 "module" file name
1426 "module" file name
1427 "pcalls" primitive call count
1427 "pcalls" primitive call count
1428 "line" line number
1428 "line" line number
1429 "name" function name
1429 "name" function name
1430 "nfl" name/file/line
1430 "nfl" name/file/line
1431 "stdname" standard name
1431 "stdname" standard name
1432 "time" internal time
1432 "time" internal time
1433
1433
1434 Note that all sorts on statistics are in descending order (placing
1434 Note that all sorts on statistics are in descending order (placing
1435 most time consuming items first), where as name, file, and line number
1435 most time consuming items first), where as name, file, and line number
1436 searches are in ascending order (i.e., alphabetical). The subtle
1436 searches are in ascending order (i.e., alphabetical). The subtle
1437 distinction between "nfl" and "stdname" is that the standard name is a
1437 distinction between "nfl" and "stdname" is that the standard name is a
1438 sort of the name as printed, which means that the embedded line
1438 sort of the name as printed, which means that the embedded line
1439 numbers get compared in an odd way. For example, lines 3, 20, and 40
1439 numbers get compared in an odd way. For example, lines 3, 20, and 40
1440 would (if the file names were the same) appear in the string order
1440 would (if the file names were the same) appear in the string order
1441 "20" "3" and "40". In contrast, "nfl" does a numeric compare of the
1441 "20" "3" and "40". In contrast, "nfl" does a numeric compare of the
1442 line numbers. In fact, sort_stats("nfl") is the same as
1442 line numbers. In fact, sort_stats("nfl") is the same as
1443 sort_stats("name", "file", "line").
1443 sort_stats("name", "file", "line").
1444
1444
1445 -T <filename>: save profile results as shown on screen to a text
1445 -T <filename>: save profile results as shown on screen to a text
1446 file. The profile is still shown on screen.
1446 file. The profile is still shown on screen.
1447
1447
1448 -D <filename>: save (via dump_stats) profile statistics to given
1448 -D <filename>: save (via dump_stats) profile statistics to given
1449 filename. This data is in a format understood by the pstats module, and
1449 filename. This data is in a format understood by the pstats module, and
1450 is generated by a call to the dump_stats() method of profile
1450 is generated by a call to the dump_stats() method of profile
1451 objects. The profile is still shown on screen.
1451 objects. The profile is still shown on screen.
1452
1452
1453 -q: suppress output to the pager. Best used with -T and/or -D above.
1453 -q: suppress output to the pager. Best used with -T and/or -D above.
1454
1454
1455 If you want to run complete programs under the profiler's control, use
1455 If you want to run complete programs under the profiler's control, use
1456 '%run -p [prof_opts] filename.py [args to program]' where prof_opts
1456 '%run -p [prof_opts] filename.py [args to program]' where prof_opts
1457 contains profiler specific options as described here.
1457 contains profiler specific options as described here.
1458
1458
1459 You can read the complete documentation for the profile module with::
1459 You can read the complete documentation for the profile module with::
1460
1460
1461 In [1]: import profile; profile.help()
1461 In [1]: import profile; profile.help()
1462 """
1462 """
1463
1463
1464 opts_def = Struct(D=[''],l=[],s=['time'],T=[''])
1464 opts_def = Struct(D=[''],l=[],s=['time'],T=[''])
1465
1465
1466 if user_mode: # regular user call
1466 if user_mode: # regular user call
1467 opts,arg_str = self.parse_options(parameter_s,'D:l:rs:T:q',
1467 opts,arg_str = self.parse_options(parameter_s,'D:l:rs:T:q',
1468 list_all=1, posix=False)
1468 list_all=1, posix=False)
1469 namespace = self.shell.user_ns
1469 namespace = self.shell.user_ns
1470 else: # called to run a program by %run -p
1470 else: # called to run a program by %run -p
1471 try:
1471 try:
1472 filename = get_py_filename(arg_lst[0])
1472 filename = get_py_filename(arg_lst[0])
1473 except IOError as e:
1473 except IOError as e:
1474 try:
1474 try:
1475 msg = str(e)
1475 msg = str(e)
1476 except UnicodeError:
1476 except UnicodeError:
1477 msg = e.message
1477 msg = e.message
1478 error(msg)
1478 error(msg)
1479 return
1479 return
1480
1480
1481 arg_str = 'execfile(filename,prog_ns)'
1481 arg_str = 'execfile(filename,prog_ns)'
1482 namespace = {
1482 namespace = {
1483 'execfile': self.shell.safe_execfile,
1483 'execfile': self.shell.safe_execfile,
1484 'prog_ns': prog_ns,
1484 'prog_ns': prog_ns,
1485 'filename': filename
1485 'filename': filename
1486 }
1486 }
1487
1487
1488 opts.merge(opts_def)
1488 opts.merge(opts_def)
1489
1489
1490 prof = profile.Profile()
1490 prof = profile.Profile()
1491 try:
1491 try:
1492 prof = prof.runctx(arg_str,namespace,namespace)
1492 prof = prof.runctx(arg_str,namespace,namespace)
1493 sys_exit = ''
1493 sys_exit = ''
1494 except SystemExit:
1494 except SystemExit:
1495 sys_exit = """*** SystemExit exception caught in code being profiled."""
1495 sys_exit = """*** SystemExit exception caught in code being profiled."""
1496
1496
1497 stats = pstats.Stats(prof).strip_dirs().sort_stats(*opts.s)
1497 stats = pstats.Stats(prof).strip_dirs().sort_stats(*opts.s)
1498
1498
1499 lims = opts.l
1499 lims = opts.l
1500 if lims:
1500 if lims:
1501 lims = [] # rebuild lims with ints/floats/strings
1501 lims = [] # rebuild lims with ints/floats/strings
1502 for lim in opts.l:
1502 for lim in opts.l:
1503 try:
1503 try:
1504 lims.append(int(lim))
1504 lims.append(int(lim))
1505 except ValueError:
1505 except ValueError:
1506 try:
1506 try:
1507 lims.append(float(lim))
1507 lims.append(float(lim))
1508 except ValueError:
1508 except ValueError:
1509 lims.append(lim)
1509 lims.append(lim)
1510
1510
1511 # Trap output.
1511 # Trap output.
1512 stdout_trap = StringIO()
1512 stdout_trap = StringIO()
1513
1513
1514 if hasattr(stats,'stream'):
1514 if hasattr(stats,'stream'):
1515 # In newer versions of python, the stats object has a 'stream'
1515 # In newer versions of python, the stats object has a 'stream'
1516 # attribute to write into.
1516 # attribute to write into.
1517 stats.stream = stdout_trap
1517 stats.stream = stdout_trap
1518 stats.print_stats(*lims)
1518 stats.print_stats(*lims)
1519 else:
1519 else:
1520 # For older versions, we manually redirect stdout during printing
1520 # For older versions, we manually redirect stdout during printing
1521 sys_stdout = sys.stdout
1521 sys_stdout = sys.stdout
1522 try:
1522 try:
1523 sys.stdout = stdout_trap
1523 sys.stdout = stdout_trap
1524 stats.print_stats(*lims)
1524 stats.print_stats(*lims)
1525 finally:
1525 finally:
1526 sys.stdout = sys_stdout
1526 sys.stdout = sys_stdout
1527
1527
1528 output = stdout_trap.getvalue()
1528 output = stdout_trap.getvalue()
1529 output = output.rstrip()
1529 output = output.rstrip()
1530
1530
1531 if 'q' not in opts:
1531 if 'q' not in opts:
1532 page.page(output)
1532 page.page(output)
1533 print sys_exit,
1533 print sys_exit,
1534
1534
1535 dump_file = opts.D[0]
1535 dump_file = opts.D[0]
1536 text_file = opts.T[0]
1536 text_file = opts.T[0]
1537 if dump_file:
1537 if dump_file:
1538 dump_file = unquote_filename(dump_file)
1538 dump_file = unquote_filename(dump_file)
1539 prof.dump_stats(dump_file)
1539 prof.dump_stats(dump_file)
1540 print '\n*** Profile stats marshalled to file',\
1540 print '\n*** Profile stats marshalled to file',\
1541 `dump_file`+'.',sys_exit
1541 `dump_file`+'.',sys_exit
1542 if text_file:
1542 if text_file:
1543 text_file = unquote_filename(text_file)
1543 text_file = unquote_filename(text_file)
1544 pfile = open(text_file,'w')
1544 pfile = open(text_file,'w')
1545 pfile.write(output)
1545 pfile.write(output)
1546 pfile.close()
1546 pfile.close()
1547 print '\n*** Profile printout saved to text file',\
1547 print '\n*** Profile printout saved to text file',\
1548 `text_file`+'.',sys_exit
1548 `text_file`+'.',sys_exit
1549
1549
1550 if opts.has_key('r'):
1550 if opts.has_key('r'):
1551 return stats
1551 return stats
1552 else:
1552 else:
1553 return None
1553 return None
1554
1554
1555 @skip_doctest
1555 @skip_doctest
1556 def magic_run(self, parameter_s ='', runner=None,
1556 def magic_run(self, parameter_s ='', runner=None,
1557 file_finder=get_py_filename):
1557 file_finder=get_py_filename):
1558 """Run the named file inside IPython as a program.
1558 """Run the named file inside IPython as a program.
1559
1559
1560 Usage:\\
1560 Usage:\\
1561 %run [-n -i -t [-N<N>] -d [-b<N>] -p [profile options]] file [args]
1561 %run [-n -i -t [-N<N>] -d [-b<N>] -p [profile options]] file [args]
1562
1562
1563 Parameters after the filename are passed as command-line arguments to
1563 Parameters after the filename are passed as command-line arguments to
1564 the program (put in sys.argv). Then, control returns to IPython's
1564 the program (put in sys.argv). Then, control returns to IPython's
1565 prompt.
1565 prompt.
1566
1566
1567 This is similar to running at a system prompt:\\
1567 This is similar to running at a system prompt:\\
1568 $ python file args\\
1568 $ python file args\\
1569 but with the advantage of giving you IPython's tracebacks, and of
1569 but with the advantage of giving you IPython's tracebacks, and of
1570 loading all variables into your interactive namespace for further use
1570 loading all variables into your interactive namespace for further use
1571 (unless -p is used, see below).
1571 (unless -p is used, see below).
1572
1572
1573 The file is executed in a namespace initially consisting only of
1573 The file is executed in a namespace initially consisting only of
1574 __name__=='__main__' and sys.argv constructed as indicated. It thus
1574 __name__=='__main__' and sys.argv constructed as indicated. It thus
1575 sees its environment as if it were being run as a stand-alone program
1575 sees its environment as if it were being run as a stand-alone program
1576 (except for sharing global objects such as previously imported
1576 (except for sharing global objects such as previously imported
1577 modules). But after execution, the IPython interactive namespace gets
1577 modules). But after execution, the IPython interactive namespace gets
1578 updated with all variables defined in the program (except for __name__
1578 updated with all variables defined in the program (except for __name__
1579 and sys.argv). This allows for very convenient loading of code for
1579 and sys.argv). This allows for very convenient loading of code for
1580 interactive work, while giving each program a 'clean sheet' to run in.
1580 interactive work, while giving each program a 'clean sheet' to run in.
1581
1581
1582 Options:
1582 Options:
1583
1583
1584 -n: __name__ is NOT set to '__main__', but to the running file's name
1584 -n: __name__ is NOT set to '__main__', but to the running file's name
1585 without extension (as python does under import). This allows running
1585 without extension (as python does under import). This allows running
1586 scripts and reloading the definitions in them without calling code
1586 scripts and reloading the definitions in them without calling code
1587 protected by an ' if __name__ == "__main__" ' clause.
1587 protected by an ' if __name__ == "__main__" ' clause.
1588
1588
1589 -i: run the file in IPython's namespace instead of an empty one. This
1589 -i: run the file in IPython's namespace instead of an empty one. This
1590 is useful if you are experimenting with code written in a text editor
1590 is useful if you are experimenting with code written in a text editor
1591 which depends on variables defined interactively.
1591 which depends on variables defined interactively.
1592
1592
1593 -e: ignore sys.exit() calls or SystemExit exceptions in the script
1593 -e: ignore sys.exit() calls or SystemExit exceptions in the script
1594 being run. This is particularly useful if IPython is being used to
1594 being run. This is particularly useful if IPython is being used to
1595 run unittests, which always exit with a sys.exit() call. In such
1595 run unittests, which always exit with a sys.exit() call. In such
1596 cases you are interested in the output of the test results, not in
1596 cases you are interested in the output of the test results, not in
1597 seeing a traceback of the unittest module.
1597 seeing a traceback of the unittest module.
1598
1598
1599 -t: print timing information at the end of the run. IPython will give
1599 -t: print timing information at the end of the run. IPython will give
1600 you an estimated CPU time consumption for your script, which under
1600 you an estimated CPU time consumption for your script, which under
1601 Unix uses the resource module to avoid the wraparound problems of
1601 Unix uses the resource module to avoid the wraparound problems of
1602 time.clock(). Under Unix, an estimate of time spent on system tasks
1602 time.clock(). Under Unix, an estimate of time spent on system tasks
1603 is also given (for Windows platforms this is reported as 0.0).
1603 is also given (for Windows platforms this is reported as 0.0).
1604
1604
1605 If -t is given, an additional -N<N> option can be given, where <N>
1605 If -t is given, an additional -N<N> option can be given, where <N>
1606 must be an integer indicating how many times you want the script to
1606 must be an integer indicating how many times you want the script to
1607 run. The final timing report will include total and per run results.
1607 run. The final timing report will include total and per run results.
1608
1608
1609 For example (testing the script uniq_stable.py)::
1609 For example (testing the script uniq_stable.py)::
1610
1610
1611 In [1]: run -t uniq_stable
1611 In [1]: run -t uniq_stable
1612
1612
1613 IPython CPU timings (estimated):\\
1613 IPython CPU timings (estimated):\\
1614 User : 0.19597 s.\\
1614 User : 0.19597 s.\\
1615 System: 0.0 s.\\
1615 System: 0.0 s.\\
1616
1616
1617 In [2]: run -t -N5 uniq_stable
1617 In [2]: run -t -N5 uniq_stable
1618
1618
1619 IPython CPU timings (estimated):\\
1619 IPython CPU timings (estimated):\\
1620 Total runs performed: 5\\
1620 Total runs performed: 5\\
1621 Times : Total Per run\\
1621 Times : Total Per run\\
1622 User : 0.910862 s, 0.1821724 s.\\
1622 User : 0.910862 s, 0.1821724 s.\\
1623 System: 0.0 s, 0.0 s.
1623 System: 0.0 s, 0.0 s.
1624
1624
1625 -d: run your program under the control of pdb, the Python debugger.
1625 -d: run your program under the control of pdb, the Python debugger.
1626 This allows you to execute your program step by step, watch variables,
1626 This allows you to execute your program step by step, watch variables,
1627 etc. Internally, what IPython does is similar to calling:
1627 etc. Internally, what IPython does is similar to calling:
1628
1628
1629 pdb.run('execfile("YOURFILENAME")')
1629 pdb.run('execfile("YOURFILENAME")')
1630
1630
1631 with a breakpoint set on line 1 of your file. You can change the line
1631 with a breakpoint set on line 1 of your file. You can change the line
1632 number for this automatic breakpoint to be <N> by using the -bN option
1632 number for this automatic breakpoint to be <N> by using the -bN option
1633 (where N must be an integer). For example::
1633 (where N must be an integer). For example::
1634
1634
1635 %run -d -b40 myscript
1635 %run -d -b40 myscript
1636
1636
1637 will set the first breakpoint at line 40 in myscript.py. Note that
1637 will set the first breakpoint at line 40 in myscript.py. Note that
1638 the first breakpoint must be set on a line which actually does
1638 the first breakpoint must be set on a line which actually does
1639 something (not a comment or docstring) for it to stop execution.
1639 something (not a comment or docstring) for it to stop execution.
1640
1640
1641 When the pdb debugger starts, you will see a (Pdb) prompt. You must
1641 When the pdb debugger starts, you will see a (Pdb) prompt. You must
1642 first enter 'c' (without quotes) to start execution up to the first
1642 first enter 'c' (without quotes) to start execution up to the first
1643 breakpoint.
1643 breakpoint.
1644
1644
1645 Entering 'help' gives information about the use of the debugger. You
1645 Entering 'help' gives information about the use of the debugger. You
1646 can easily see pdb's full documentation with "import pdb;pdb.help()"
1646 can easily see pdb's full documentation with "import pdb;pdb.help()"
1647 at a prompt.
1647 at a prompt.
1648
1648
1649 -p: run program under the control of the Python profiler module (which
1649 -p: run program under the control of the Python profiler module (which
1650 prints a detailed report of execution times, function calls, etc).
1650 prints a detailed report of execution times, function calls, etc).
1651
1651
1652 You can pass other options after -p which affect the behavior of the
1652 You can pass other options after -p which affect the behavior of the
1653 profiler itself. See the docs for %prun for details.
1653 profiler itself. See the docs for %prun for details.
1654
1654
1655 In this mode, the program's variables do NOT propagate back to the
1655 In this mode, the program's variables do NOT propagate back to the
1656 IPython interactive namespace (because they remain in the namespace
1656 IPython interactive namespace (because they remain in the namespace
1657 where the profiler executes them).
1657 where the profiler executes them).
1658
1658
1659 Internally this triggers a call to %prun, see its documentation for
1659 Internally this triggers a call to %prun, see its documentation for
1660 details on the options available specifically for profiling.
1660 details on the options available specifically for profiling.
1661
1661
1662 There is one special usage for which the text above doesn't apply:
1662 There is one special usage for which the text above doesn't apply:
1663 if the filename ends with .ipy, the file is run as ipython script,
1663 if the filename ends with .ipy, the file is run as ipython script,
1664 just as if the commands were written on IPython prompt.
1664 just as if the commands were written on IPython prompt.
1665
1665
1666 -m: specify module name to load instead of script path. Similar to
1666 -m: specify module name to load instead of script path. Similar to
1667 the -m option for the python interpreter. Use this option last if you
1667 the -m option for the python interpreter. Use this option last if you
1668 want to combine with other %run options. Unlike the python interpreter
1668 want to combine with other %run options. Unlike the python interpreter
1669 only source modules are allowed no .pyc or .pyo files.
1669 only source modules are allowed no .pyc or .pyo files.
1670 For example::
1670 For example::
1671
1671
1672 %run -m example
1672 %run -m example
1673
1673
1674 will run the example module.
1674 will run the example module.
1675
1675
1676 """
1676 """
1677
1677
1678 # get arguments and set sys.argv for program to be run.
1678 # get arguments and set sys.argv for program to be run.
1679 opts, arg_lst = self.parse_options(parameter_s, 'nidtN:b:pD:l:rs:T:em:',
1679 opts, arg_lst = self.parse_options(parameter_s, 'nidtN:b:pD:l:rs:T:em:',
1680 mode='list', list_all=1)
1680 mode='list', list_all=1)
1681 if "m" in opts:
1681 if "m" in opts:
1682 modulename = opts["m"][0]
1682 modulename = opts["m"][0]
1683 modpath = find_mod(modulename)
1683 modpath = find_mod(modulename)
1684 if modpath is None:
1684 if modpath is None:
1685 warn('%r is not a valid modulename on sys.path'%modulename)
1685 warn('%r is not a valid modulename on sys.path'%modulename)
1686 return
1686 return
1687 arg_lst = [modpath] + arg_lst
1687 arg_lst = [modpath] + arg_lst
1688 try:
1688 try:
1689 filename = file_finder(arg_lst[0])
1689 filename = file_finder(arg_lst[0])
1690 except IndexError:
1690 except IndexError:
1691 warn('you must provide at least a filename.')
1691 warn('you must provide at least a filename.')
1692 print '\n%run:\n', oinspect.getdoc(self.magic_run)
1692 print '\n%run:\n', oinspect.getdoc(self.magic_run)
1693 return
1693 return
1694 except IOError as e:
1694 except IOError as e:
1695 try:
1695 try:
1696 msg = str(e)
1696 msg = str(e)
1697 except UnicodeError:
1697 except UnicodeError:
1698 msg = e.message
1698 msg = e.message
1699 error(msg)
1699 error(msg)
1700 return
1700 return
1701
1701
1702 if filename.lower().endswith('.ipy'):
1702 if filename.lower().endswith('.ipy'):
1703 self.shell.safe_execfile_ipy(filename)
1703 self.shell.safe_execfile_ipy(filename)
1704 return
1704 return
1705
1705
1706 # Control the response to exit() calls made by the script being run
1706 # Control the response to exit() calls made by the script being run
1707 exit_ignore = 'e' in opts
1707 exit_ignore = 'e' in opts
1708
1708
1709 # Make sure that the running script gets a proper sys.argv as if it
1709 # Make sure that the running script gets a proper sys.argv as if it
1710 # were run from a system shell.
1710 # were run from a system shell.
1711 save_argv = sys.argv # save it for later restoring
1711 save_argv = sys.argv # save it for later restoring
1712
1712
1713 # simulate shell expansion on arguments, at least tilde expansion
1713 # simulate shell expansion on arguments, at least tilde expansion
1714 args = [ os.path.expanduser(a) for a in arg_lst[1:] ]
1714 args = [ os.path.expanduser(a) for a in arg_lst[1:] ]
1715
1715
1716 sys.argv = [filename] + args # put in the proper filename
1716 sys.argv = [filename] + args # put in the proper filename
1717 # protect sys.argv from potential unicode strings on Python 2:
1717 # protect sys.argv from potential unicode strings on Python 2:
1718 if not py3compat.PY3:
1718 if not py3compat.PY3:
1719 sys.argv = [ py3compat.cast_bytes(a) for a in sys.argv ]
1719 sys.argv = [ py3compat.cast_bytes(a) for a in sys.argv ]
1720
1720
1721 if 'i' in opts:
1721 if 'i' in opts:
1722 # Run in user's interactive namespace
1722 # Run in user's interactive namespace
1723 prog_ns = self.shell.user_ns
1723 prog_ns = self.shell.user_ns
1724 __name__save = self.shell.user_ns['__name__']
1724 __name__save = self.shell.user_ns['__name__']
1725 prog_ns['__name__'] = '__main__'
1725 prog_ns['__name__'] = '__main__'
1726 main_mod = self.shell.new_main_mod(prog_ns)
1726 main_mod = self.shell.new_main_mod(prog_ns)
1727 else:
1727 else:
1728 # Run in a fresh, empty namespace
1728 # Run in a fresh, empty namespace
1729 if 'n' in opts:
1729 if 'n' in opts:
1730 name = os.path.splitext(os.path.basename(filename))[0]
1730 name = os.path.splitext(os.path.basename(filename))[0]
1731 else:
1731 else:
1732 name = '__main__'
1732 name = '__main__'
1733
1733
1734 main_mod = self.shell.new_main_mod()
1734 main_mod = self.shell.new_main_mod()
1735 prog_ns = main_mod.__dict__
1735 prog_ns = main_mod.__dict__
1736 prog_ns['__name__'] = name
1736 prog_ns['__name__'] = name
1737
1737
1738 # Since '%run foo' emulates 'python foo.py' at the cmd line, we must
1738 # Since '%run foo' emulates 'python foo.py' at the cmd line, we must
1739 # set the __file__ global in the script's namespace
1739 # set the __file__ global in the script's namespace
1740 prog_ns['__file__'] = filename
1740 prog_ns['__file__'] = filename
1741
1741
1742 # pickle fix. See interactiveshell for an explanation. But we need to make sure
1742 # pickle fix. See interactiveshell for an explanation. But we need to make sure
1743 # that, if we overwrite __main__, we replace it at the end
1743 # that, if we overwrite __main__, we replace it at the end
1744 main_mod_name = prog_ns['__name__']
1744 main_mod_name = prog_ns['__name__']
1745
1745
1746 if main_mod_name == '__main__':
1746 if main_mod_name == '__main__':
1747 restore_main = sys.modules['__main__']
1747 restore_main = sys.modules['__main__']
1748 else:
1748 else:
1749 restore_main = False
1749 restore_main = False
1750
1750
1751 # This needs to be undone at the end to prevent holding references to
1751 # This needs to be undone at the end to prevent holding references to
1752 # every single object ever created.
1752 # every single object ever created.
1753 sys.modules[main_mod_name] = main_mod
1753 sys.modules[main_mod_name] = main_mod
1754
1754
1755 try:
1755 try:
1756 stats = None
1756 stats = None
1757 with self.readline_no_record:
1757 with self.readline_no_record:
1758 if 'p' in opts:
1758 if 'p' in opts:
1759 stats = self.magic_prun('', 0, opts, arg_lst, prog_ns)
1759 stats = self.magic_prun('', 0, opts, arg_lst, prog_ns)
1760 else:
1760 else:
1761 if 'd' in opts:
1761 if 'd' in opts:
1762 deb = debugger.Pdb(self.shell.colors)
1762 deb = debugger.Pdb(self.shell.colors)
1763 # reset Breakpoint state, which is moronically kept
1763 # reset Breakpoint state, which is moronically kept
1764 # in a class
1764 # in a class
1765 bdb.Breakpoint.next = 1
1765 bdb.Breakpoint.next = 1
1766 bdb.Breakpoint.bplist = {}
1766 bdb.Breakpoint.bplist = {}
1767 bdb.Breakpoint.bpbynumber = [None]
1767 bdb.Breakpoint.bpbynumber = [None]
1768 # Set an initial breakpoint to stop execution
1768 # Set an initial breakpoint to stop execution
1769 maxtries = 10
1769 maxtries = 10
1770 bp = int(opts.get('b', [1])[0])
1770 bp = int(opts.get('b', [1])[0])
1771 checkline = deb.checkline(filename, bp)
1771 checkline = deb.checkline(filename, bp)
1772 if not checkline:
1772 if not checkline:
1773 for bp in range(bp + 1, bp + maxtries + 1):
1773 for bp in range(bp + 1, bp + maxtries + 1):
1774 if deb.checkline(filename, bp):
1774 if deb.checkline(filename, bp):
1775 break
1775 break
1776 else:
1776 else:
1777 msg = ("\nI failed to find a valid line to set "
1777 msg = ("\nI failed to find a valid line to set "
1778 "a breakpoint\n"
1778 "a breakpoint\n"
1779 "after trying up to line: %s.\n"
1779 "after trying up to line: %s.\n"
1780 "Please set a valid breakpoint manually "
1780 "Please set a valid breakpoint manually "
1781 "with the -b option." % bp)
1781 "with the -b option." % bp)
1782 error(msg)
1782 error(msg)
1783 return
1783 return
1784 # if we find a good linenumber, set the breakpoint
1784 # if we find a good linenumber, set the breakpoint
1785 deb.do_break('%s:%s' % (filename, bp))
1785 deb.do_break('%s:%s' % (filename, bp))
1786 # Start file run
1786 # Start file run
1787 print "NOTE: Enter 'c' at the",
1787 print "NOTE: Enter 'c' at the",
1788 print "%s prompt to start your script." % deb.prompt
1788 print "%s prompt to start your script." % deb.prompt
1789 ns = {'execfile': py3compat.execfile, 'prog_ns': prog_ns}
1789 ns = {'execfile': py3compat.execfile, 'prog_ns': prog_ns}
1790 try:
1790 try:
1791 deb.run('execfile("%s", prog_ns)' % filename, ns)
1791 deb.run('execfile("%s", prog_ns)' % filename, ns)
1792
1792
1793 except:
1793 except:
1794 etype, value, tb = sys.exc_info()
1794 etype, value, tb = sys.exc_info()
1795 # Skip three frames in the traceback: the %run one,
1795 # Skip three frames in the traceback: the %run one,
1796 # one inside bdb.py, and the command-line typed by the
1796 # one inside bdb.py, and the command-line typed by the
1797 # user (run by exec in pdb itself).
1797 # user (run by exec in pdb itself).
1798 self.shell.InteractiveTB(etype, value, tb, tb_offset=3)
1798 self.shell.InteractiveTB(etype, value, tb, tb_offset=3)
1799 else:
1799 else:
1800 if runner is None:
1800 if runner is None:
1801 runner = self.shell.safe_execfile
1801 runner = self.shell.safe_execfile
1802 if 't' in opts:
1802 if 't' in opts:
1803 # timed execution
1803 # timed execution
1804 try:
1804 try:
1805 nruns = int(opts['N'][0])
1805 nruns = int(opts['N'][0])
1806 if nruns < 1:
1806 if nruns < 1:
1807 error('Number of runs must be >=1')
1807 error('Number of runs must be >=1')
1808 return
1808 return
1809 except (KeyError):
1809 except (KeyError):
1810 nruns = 1
1810 nruns = 1
1811 twall0 = time.time()
1811 twall0 = time.time()
1812 if nruns == 1:
1812 if nruns == 1:
1813 t0 = clock2()
1813 t0 = clock2()
1814 runner(filename, prog_ns, prog_ns,
1814 runner(filename, prog_ns, prog_ns,
1815 exit_ignore=exit_ignore)
1815 exit_ignore=exit_ignore)
1816 t1 = clock2()
1816 t1 = clock2()
1817 t_usr = t1[0] - t0[0]
1817 t_usr = t1[0] - t0[0]
1818 t_sys = t1[1] - t0[1]
1818 t_sys = t1[1] - t0[1]
1819 print "\nIPython CPU timings (estimated):"
1819 print "\nIPython CPU timings (estimated):"
1820 print " User : %10.2f s." % t_usr
1820 print " User : %10.2f s." % t_usr
1821 print " System : %10.2f s." % t_sys
1821 print " System : %10.2f s." % t_sys
1822 else:
1822 else:
1823 runs = range(nruns)
1823 runs = range(nruns)
1824 t0 = clock2()
1824 t0 = clock2()
1825 for nr in runs:
1825 for nr in runs:
1826 runner(filename, prog_ns, prog_ns,
1826 runner(filename, prog_ns, prog_ns,
1827 exit_ignore=exit_ignore)
1827 exit_ignore=exit_ignore)
1828 t1 = clock2()
1828 t1 = clock2()
1829 t_usr = t1[0] - t0[0]
1829 t_usr = t1[0] - t0[0]
1830 t_sys = t1[1] - t0[1]
1830 t_sys = t1[1] - t0[1]
1831 print "\nIPython CPU timings (estimated):"
1831 print "\nIPython CPU timings (estimated):"
1832 print "Total runs performed:", nruns
1832 print "Total runs performed:", nruns
1833 print " Times : %10.2f %10.2f" % ('Total', 'Per run')
1833 print " Times : %10.2f %10.2f" % ('Total', 'Per run')
1834 print " User : %10.2f s, %10.2f s." % (t_usr, t_usr / nruns)
1834 print " User : %10.2f s, %10.2f s." % (t_usr, t_usr / nruns)
1835 print " System : %10.2f s, %10.2f s." % (t_sys, t_sys / nruns)
1835 print " System : %10.2f s, %10.2f s." % (t_sys, t_sys / nruns)
1836 twall1 = time.time()
1836 twall1 = time.time()
1837 print "Wall time: %10.2f s." % (twall1 - twall0)
1837 print "Wall time: %10.2f s." % (twall1 - twall0)
1838
1838
1839 else:
1839 else:
1840 # regular execution
1840 # regular execution
1841 runner(filename, prog_ns, prog_ns, exit_ignore=exit_ignore)
1841 runner(filename, prog_ns, prog_ns, exit_ignore=exit_ignore)
1842
1842
1843 if 'i' in opts:
1843 if 'i' in opts:
1844 self.shell.user_ns['__name__'] = __name__save
1844 self.shell.user_ns['__name__'] = __name__save
1845 else:
1845 else:
1846 # The shell MUST hold a reference to prog_ns so after %run
1846 # The shell MUST hold a reference to prog_ns so after %run
1847 # exits, the python deletion mechanism doesn't zero it out
1847 # exits, the python deletion mechanism doesn't zero it out
1848 # (leaving dangling references).
1848 # (leaving dangling references).
1849 self.shell.cache_main_mod(prog_ns, filename)
1849 self.shell.cache_main_mod(prog_ns, filename)
1850 # update IPython interactive namespace
1850 # update IPython interactive namespace
1851
1851
1852 # Some forms of read errors on the file may mean the
1852 # Some forms of read errors on the file may mean the
1853 # __name__ key was never set; using pop we don't have to
1853 # __name__ key was never set; using pop we don't have to
1854 # worry about a possible KeyError.
1854 # worry about a possible KeyError.
1855 prog_ns.pop('__name__', None)
1855 prog_ns.pop('__name__', None)
1856
1856
1857 self.shell.user_ns.update(prog_ns)
1857 self.shell.user_ns.update(prog_ns)
1858 finally:
1858 finally:
1859 # It's a bit of a mystery why, but __builtins__ can change from
1859 # It's a bit of a mystery why, but __builtins__ can change from
1860 # being a module to becoming a dict missing some key data after
1860 # being a module to becoming a dict missing some key data after
1861 # %run. As best I can see, this is NOT something IPython is doing
1861 # %run. As best I can see, this is NOT something IPython is doing
1862 # at all, and similar problems have been reported before:
1862 # at all, and similar problems have been reported before:
1863 # http://coding.derkeiler.com/Archive/Python/comp.lang.python/2004-10/0188.html
1863 # http://coding.derkeiler.com/Archive/Python/comp.lang.python/2004-10/0188.html
1864 # Since this seems to be done by the interpreter itself, the best
1864 # Since this seems to be done by the interpreter itself, the best
1865 # we can do is to at least restore __builtins__ for the user on
1865 # we can do is to at least restore __builtins__ for the user on
1866 # exit.
1866 # exit.
1867 self.shell.user_ns['__builtins__'] = builtin_mod
1867 self.shell.user_ns['__builtins__'] = builtin_mod
1868
1868
1869 # Ensure key global structures are restored
1869 # Ensure key global structures are restored
1870 sys.argv = save_argv
1870 sys.argv = save_argv
1871 if restore_main:
1871 if restore_main:
1872 sys.modules['__main__'] = restore_main
1872 sys.modules['__main__'] = restore_main
1873 else:
1873 else:
1874 # Remove from sys.modules the reference to main_mod we'd
1874 # Remove from sys.modules the reference to main_mod we'd
1875 # added. Otherwise it will trap references to objects
1875 # added. Otherwise it will trap references to objects
1876 # contained therein.
1876 # contained therein.
1877 del sys.modules[main_mod_name]
1877 del sys.modules[main_mod_name]
1878
1878
1879 return stats
1879 return stats
1880
1880
1881 @skip_doctest
1881 @skip_doctest
1882 def magic_timeit(self, parameter_s =''):
1882 def magic_timeit(self, parameter_s =''):
1883 """Time execution of a Python statement or expression
1883 """Time execution of a Python statement or expression
1884
1884
1885 Usage:\\
1885 Usage:\\
1886 %timeit [-n<N> -r<R> [-t|-c]] statement
1886 %timeit [-n<N> -r<R> [-t|-c]] statement
1887
1887
1888 Time execution of a Python statement or expression using the timeit
1888 Time execution of a Python statement or expression using the timeit
1889 module.
1889 module.
1890
1890
1891 Options:
1891 Options:
1892 -n<N>: execute the given statement <N> times in a loop. If this value
1892 -n<N>: execute the given statement <N> times in a loop. If this value
1893 is not given, a fitting value is chosen.
1893 is not given, a fitting value is chosen.
1894
1894
1895 -r<R>: repeat the loop iteration <R> times and take the best result.
1895 -r<R>: repeat the loop iteration <R> times and take the best result.
1896 Default: 3
1896 Default: 3
1897
1897
1898 -t: use time.time to measure the time, which is the default on Unix.
1898 -t: use time.time to measure the time, which is the default on Unix.
1899 This function measures wall time.
1899 This function measures wall time.
1900
1900
1901 -c: use time.clock to measure the time, which is the default on
1901 -c: use time.clock to measure the time, which is the default on
1902 Windows and measures wall time. On Unix, resource.getrusage is used
1902 Windows and measures wall time. On Unix, resource.getrusage is used
1903 instead and returns the CPU user time.
1903 instead and returns the CPU user time.
1904
1904
1905 -p<P>: use a precision of <P> digits to display the timing result.
1905 -p<P>: use a precision of <P> digits to display the timing result.
1906 Default: 3
1906 Default: 3
1907
1907
1908
1908
1909 Examples
1909 Examples
1910 --------
1910 --------
1911 ::
1911 ::
1912
1912
1913 In [1]: %timeit pass
1913 In [1]: %timeit pass
1914 10000000 loops, best of 3: 53.3 ns per loop
1914 10000000 loops, best of 3: 53.3 ns per loop
1915
1915
1916 In [2]: u = None
1916 In [2]: u = None
1917
1917
1918 In [3]: %timeit u is None
1918 In [3]: %timeit u is None
1919 10000000 loops, best of 3: 184 ns per loop
1919 10000000 loops, best of 3: 184 ns per loop
1920
1920
1921 In [4]: %timeit -r 4 u == None
1921 In [4]: %timeit -r 4 u == None
1922 1000000 loops, best of 4: 242 ns per loop
1922 1000000 loops, best of 4: 242 ns per loop
1923
1923
1924 In [5]: import time
1924 In [5]: import time
1925
1925
1926 In [6]: %timeit -n1 time.sleep(2)
1926 In [6]: %timeit -n1 time.sleep(2)
1927 1 loops, best of 3: 2 s per loop
1927 1 loops, best of 3: 2 s per loop
1928
1928
1929
1929
1930 The times reported by %timeit will be slightly higher than those
1930 The times reported by %timeit will be slightly higher than those
1931 reported by the timeit.py script when variables are accessed. This is
1931 reported by the timeit.py script when variables are accessed. This is
1932 due to the fact that %timeit executes the statement in the namespace
1932 due to the fact that %timeit executes the statement in the namespace
1933 of the shell, compared with timeit.py, which uses a single setup
1933 of the shell, compared with timeit.py, which uses a single setup
1934 statement to import function or create variables. Generally, the bias
1934 statement to import function or create variables. Generally, the bias
1935 does not matter as long as results from timeit.py are not mixed with
1935 does not matter as long as results from timeit.py are not mixed with
1936 those from %timeit."""
1936 those from %timeit."""
1937
1937
1938 import timeit
1938 import timeit
1939 import math
1939 import math
1940
1940
1941 # XXX: Unfortunately the unicode 'micro' symbol can cause problems in
1941 # XXX: Unfortunately the unicode 'micro' symbol can cause problems in
1942 # certain terminals. Until we figure out a robust way of
1942 # certain terminals. Until we figure out a robust way of
1943 # auto-detecting if the terminal can deal with it, use plain 'us' for
1943 # auto-detecting if the terminal can deal with it, use plain 'us' for
1944 # microseconds. I am really NOT happy about disabling the proper
1944 # microseconds. I am really NOT happy about disabling the proper
1945 # 'micro' prefix, but crashing is worse... If anyone knows what the
1945 # 'micro' prefix, but crashing is worse... If anyone knows what the
1946 # right solution for this is, I'm all ears...
1946 # right solution for this is, I'm all ears...
1947 #
1947 #
1948 # Note: using
1948 # Note: using
1949 #
1949 #
1950 # s = u'\xb5'
1950 # s = u'\xb5'
1951 # s.encode(sys.getdefaultencoding())
1951 # s.encode(sys.getdefaultencoding())
1952 #
1952 #
1953 # is not sufficient, as I've seen terminals where that fails but
1953 # is not sufficient, as I've seen terminals where that fails but
1954 # print s
1954 # print s
1955 #
1955 #
1956 # succeeds
1956 # succeeds
1957 #
1957 #
1958 # See bug: https://bugs.launchpad.net/ipython/+bug/348466
1958 # See bug: https://bugs.launchpad.net/ipython/+bug/348466
1959
1959
1960 #units = [u"s", u"ms",u'\xb5',"ns"]
1960 #units = [u"s", u"ms",u'\xb5',"ns"]
1961 units = [u"s", u"ms",u'us',"ns"]
1961 units = [u"s", u"ms",u'us',"ns"]
1962
1962
1963 scaling = [1, 1e3, 1e6, 1e9]
1963 scaling = [1, 1e3, 1e6, 1e9]
1964
1964
1965 opts, stmt = self.parse_options(parameter_s,'n:r:tcp:',
1965 opts, stmt = self.parse_options(parameter_s,'n:r:tcp:',
1966 posix=False, strict=False)
1966 posix=False, strict=False)
1967 if stmt == "":
1967 if stmt == "":
1968 return
1968 return
1969 timefunc = timeit.default_timer
1969 timefunc = timeit.default_timer
1970 number = int(getattr(opts, "n", 0))
1970 number = int(getattr(opts, "n", 0))
1971 repeat = int(getattr(opts, "r", timeit.default_repeat))
1971 repeat = int(getattr(opts, "r", timeit.default_repeat))
1972 precision = int(getattr(opts, "p", 3))
1972 precision = int(getattr(opts, "p", 3))
1973 if hasattr(opts, "t"):
1973 if hasattr(opts, "t"):
1974 timefunc = time.time
1974 timefunc = time.time
1975 if hasattr(opts, "c"):
1975 if hasattr(opts, "c"):
1976 timefunc = clock
1976 timefunc = clock
1977
1977
1978 timer = timeit.Timer(timer=timefunc)
1978 timer = timeit.Timer(timer=timefunc)
1979 # this code has tight coupling to the inner workings of timeit.Timer,
1979 # this code has tight coupling to the inner workings of timeit.Timer,
1980 # but is there a better way to achieve that the code stmt has access
1980 # but is there a better way to achieve that the code stmt has access
1981 # to the shell namespace?
1981 # to the shell namespace?
1982
1982
1983 src = timeit.template % {'stmt': timeit.reindent(stmt, 8),
1983 src = timeit.template % {'stmt': timeit.reindent(stmt, 8),
1984 'setup': "pass"}
1984 'setup': "pass"}
1985 # Track compilation time so it can be reported if too long
1985 # Track compilation time so it can be reported if too long
1986 # Minimum time above which compilation time will be reported
1986 # Minimum time above which compilation time will be reported
1987 tc_min = 0.1
1987 tc_min = 0.1
1988
1988
1989 t0 = clock()
1989 t0 = clock()
1990 code = compile(src, "<magic-timeit>", "exec")
1990 code = compile(src, "<magic-timeit>", "exec")
1991 tc = clock()-t0
1991 tc = clock()-t0
1992
1992
1993 ns = {}
1993 ns = {}
1994 exec code in self.shell.user_ns, ns
1994 exec code in self.shell.user_ns, ns
1995 timer.inner = ns["inner"]
1995 timer.inner = ns["inner"]
1996
1996
1997 if number == 0:
1997 if number == 0:
1998 # determine number so that 0.2 <= total time < 2.0
1998 # determine number so that 0.2 <= total time < 2.0
1999 number = 1
1999 number = 1
2000 for i in range(1, 10):
2000 for i in range(1, 10):
2001 if timer.timeit(number) >= 0.2:
2001 if timer.timeit(number) >= 0.2:
2002 break
2002 break
2003 number *= 10
2003 number *= 10
2004
2004
2005 best = min(timer.repeat(repeat, number)) / number
2005 best = min(timer.repeat(repeat, number)) / number
2006
2006
2007 if best > 0.0 and best < 1000.0:
2007 if best > 0.0 and best < 1000.0:
2008 order = min(-int(math.floor(math.log10(best)) // 3), 3)
2008 order = min(-int(math.floor(math.log10(best)) // 3), 3)
2009 elif best >= 1000.0:
2009 elif best >= 1000.0:
2010 order = 0
2010 order = 0
2011 else:
2011 else:
2012 order = 3
2012 order = 3
2013 print u"%d loops, best of %d: %.*g %s per loop" % (number, repeat,
2013 print u"%d loops, best of %d: %.*g %s per loop" % (number, repeat,
2014 precision,
2014 precision,
2015 best * scaling[order],
2015 best * scaling[order],
2016 units[order])
2016 units[order])
2017 if tc > tc_min:
2017 if tc > tc_min:
2018 print "Compiler time: %.2f s" % tc
2018 print "Compiler time: %.2f s" % tc
2019
2019
2020 @skip_doctest
2020 @skip_doctest
2021 @needs_local_scope
2021 @needs_local_scope
2022 def magic_time(self,parameter_s = ''):
2022 def magic_time(self,parameter_s = ''):
2023 """Time execution of a Python statement or expression.
2023 """Time execution of a Python statement or expression.
2024
2024
2025 The CPU and wall clock times are printed, and the value of the
2025 The CPU and wall clock times are printed, and the value of the
2026 expression (if any) is returned. Note that under Win32, system time
2026 expression (if any) is returned. Note that under Win32, system time
2027 is always reported as 0, since it can not be measured.
2027 is always reported as 0, since it can not be measured.
2028
2028
2029 This function provides very basic timing functionality. In Python
2029 This function provides very basic timing functionality. In Python
2030 2.3, the timeit module offers more control and sophistication, so this
2030 2.3, the timeit module offers more control and sophistication, so this
2031 could be rewritten to use it (patches welcome).
2031 could be rewritten to use it (patches welcome).
2032
2032
2033 Examples
2033 Examples
2034 --------
2034 --------
2035 ::
2035 ::
2036
2036
2037 In [1]: time 2**128
2037 In [1]: time 2**128
2038 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2038 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2039 Wall time: 0.00
2039 Wall time: 0.00
2040 Out[1]: 340282366920938463463374607431768211456L
2040 Out[1]: 340282366920938463463374607431768211456L
2041
2041
2042 In [2]: n = 1000000
2042 In [2]: n = 1000000
2043
2043
2044 In [3]: time sum(range(n))
2044 In [3]: time sum(range(n))
2045 CPU times: user 1.20 s, sys: 0.05 s, total: 1.25 s
2045 CPU times: user 1.20 s, sys: 0.05 s, total: 1.25 s
2046 Wall time: 1.37
2046 Wall time: 1.37
2047 Out[3]: 499999500000L
2047 Out[3]: 499999500000L
2048
2048
2049 In [4]: time print 'hello world'
2049 In [4]: time print 'hello world'
2050 hello world
2050 hello world
2051 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2051 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2052 Wall time: 0.00
2052 Wall time: 0.00
2053
2053
2054 Note that the time needed by Python to compile the given expression
2054 Note that the time needed by Python to compile the given expression
2055 will be reported if it is more than 0.1s. In this example, the
2055 will be reported if it is more than 0.1s. In this example, the
2056 actual exponentiation is done by Python at compilation time, so while
2056 actual exponentiation is done by Python at compilation time, so while
2057 the expression can take a noticeable amount of time to compute, that
2057 the expression can take a noticeable amount of time to compute, that
2058 time is purely due to the compilation:
2058 time is purely due to the compilation:
2059
2059
2060 In [5]: time 3**9999;
2060 In [5]: time 3**9999;
2061 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2061 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2062 Wall time: 0.00 s
2062 Wall time: 0.00 s
2063
2063
2064 In [6]: time 3**999999;
2064 In [6]: time 3**999999;
2065 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2065 CPU times: user 0.00 s, sys: 0.00 s, total: 0.00 s
2066 Wall time: 0.00 s
2066 Wall time: 0.00 s
2067 Compiler : 0.78 s
2067 Compiler : 0.78 s
2068 """
2068 """
2069
2069
2070 # fail immediately if the given expression can't be compiled
2070 # fail immediately if the given expression can't be compiled
2071
2071
2072 expr = self.shell.prefilter(parameter_s,False)
2072 expr = self.shell.prefilter(parameter_s,False)
2073
2073
2074 # Minimum time above which compilation time will be reported
2074 # Minimum time above which compilation time will be reported
2075 tc_min = 0.1
2075 tc_min = 0.1
2076
2076
2077 try:
2077 try:
2078 mode = 'eval'
2078 mode = 'eval'
2079 t0 = clock()
2079 t0 = clock()
2080 code = compile(expr,'<timed eval>',mode)
2080 code = compile(expr,'<timed eval>',mode)
2081 tc = clock()-t0
2081 tc = clock()-t0
2082 except SyntaxError:
2082 except SyntaxError:
2083 mode = 'exec'
2083 mode = 'exec'
2084 t0 = clock()
2084 t0 = clock()
2085 code = compile(expr,'<timed exec>',mode)
2085 code = compile(expr,'<timed exec>',mode)
2086 tc = clock()-t0
2086 tc = clock()-t0
2087 # skew measurement as little as possible
2087 # skew measurement as little as possible
2088 glob = self.shell.user_ns
2088 glob = self.shell.user_ns
2089 locs = self._magic_locals
2089 locs = self._magic_locals
2090 clk = clock2
2090 clk = clock2
2091 wtime = time.time
2091 wtime = time.time
2092 # time execution
2092 # time execution
2093 wall_st = wtime()
2093 wall_st = wtime()
2094 if mode=='eval':
2094 if mode=='eval':
2095 st = clk()
2095 st = clk()
2096 out = eval(code, glob, locs)
2096 out = eval(code, glob, locs)
2097 end = clk()
2097 end = clk()
2098 else:
2098 else:
2099 st = clk()
2099 st = clk()
2100 exec code in glob, locs
2100 exec code in glob, locs
2101 end = clk()
2101 end = clk()
2102 out = None
2102 out = None
2103 wall_end = wtime()
2103 wall_end = wtime()
2104 # Compute actual times and report
2104 # Compute actual times and report
2105 wall_time = wall_end-wall_st
2105 wall_time = wall_end-wall_st
2106 cpu_user = end[0]-st[0]
2106 cpu_user = end[0]-st[0]
2107 cpu_sys = end[1]-st[1]
2107 cpu_sys = end[1]-st[1]
2108 cpu_tot = cpu_user+cpu_sys
2108 cpu_tot = cpu_user+cpu_sys
2109 print "CPU times: user %.2f s, sys: %.2f s, total: %.2f s" % \
2109 print "CPU times: user %.2f s, sys: %.2f s, total: %.2f s" % \
2110 (cpu_user,cpu_sys,cpu_tot)
2110 (cpu_user,cpu_sys,cpu_tot)
2111 print "Wall time: %.2f s" % wall_time
2111 print "Wall time: %.2f s" % wall_time
2112 if tc > tc_min:
2112 if tc > tc_min:
2113 print "Compiler : %.2f s" % tc
2113 print "Compiler : %.2f s" % tc
2114 return out
2114 return out
2115
2115
2116 @skip_doctest
2116 @skip_doctest
2117 def magic_macro(self,parameter_s = ''):
2117 def magic_macro(self,parameter_s = ''):
2118 """Define a macro for future re-execution. It accepts ranges of history,
2118 """Define a macro for future re-execution. It accepts ranges of history,
2119 filenames or string objects.
2119 filenames or string objects.
2120
2120
2121 Usage:\\
2121 Usage:\\
2122 %macro [options] name n1-n2 n3-n4 ... n5 .. n6 ...
2122 %macro [options] name n1-n2 n3-n4 ... n5 .. n6 ...
2123
2123
2124 Options:
2124 Options:
2125
2125
2126 -r: use 'raw' input. By default, the 'processed' history is used,
2126 -r: use 'raw' input. By default, the 'processed' history is used,
2127 so that magics are loaded in their transformed version to valid
2127 so that magics are loaded in their transformed version to valid
2128 Python. If this option is given, the raw input as typed as the
2128 Python. If this option is given, the raw input as typed as the
2129 command line is used instead.
2129 command line is used instead.
2130
2130
2131 This will define a global variable called `name` which is a string
2131 This will define a global variable called `name` which is a string
2132 made of joining the slices and lines you specify (n1,n2,... numbers
2132 made of joining the slices and lines you specify (n1,n2,... numbers
2133 above) from your input history into a single string. This variable
2133 above) from your input history into a single string. This variable
2134 acts like an automatic function which re-executes those lines as if
2134 acts like an automatic function which re-executes those lines as if
2135 you had typed them. You just type 'name' at the prompt and the code
2135 you had typed them. You just type 'name' at the prompt and the code
2136 executes.
2136 executes.
2137
2137
2138 The syntax for indicating input ranges is described in %history.
2138 The syntax for indicating input ranges is described in %history.
2139
2139
2140 Note: as a 'hidden' feature, you can also use traditional python slice
2140 Note: as a 'hidden' feature, you can also use traditional python slice
2141 notation, where N:M means numbers N through M-1.
2141 notation, where N:M means numbers N through M-1.
2142
2142
2143 For example, if your history contains (%hist prints it)::
2143 For example, if your history contains (%hist prints it)::
2144
2144
2145 44: x=1
2145 44: x=1
2146 45: y=3
2146 45: y=3
2147 46: z=x+y
2147 46: z=x+y
2148 47: print x
2148 47: print x
2149 48: a=5
2149 48: a=5
2150 49: print 'x',x,'y',y
2150 49: print 'x',x,'y',y
2151
2151
2152 you can create a macro with lines 44 through 47 (included) and line 49
2152 you can create a macro with lines 44 through 47 (included) and line 49
2153 called my_macro with::
2153 called my_macro with::
2154
2154
2155 In [55]: %macro my_macro 44-47 49
2155 In [55]: %macro my_macro 44-47 49
2156
2156
2157 Now, typing `my_macro` (without quotes) will re-execute all this code
2157 Now, typing `my_macro` (without quotes) will re-execute all this code
2158 in one pass.
2158 in one pass.
2159
2159
2160 You don't need to give the line-numbers in order, and any given line
2160 You don't need to give the line-numbers in order, and any given line
2161 number can appear multiple times. You can assemble macros with any
2161 number can appear multiple times. You can assemble macros with any
2162 lines from your input history in any order.
2162 lines from your input history in any order.
2163
2163
2164 The macro is a simple object which holds its value in an attribute,
2164 The macro is a simple object which holds its value in an attribute,
2165 but IPython's display system checks for macros and executes them as
2165 but IPython's display system checks for macros and executes them as
2166 code instead of printing them when you type their name.
2166 code instead of printing them when you type their name.
2167
2167
2168 You can view a macro's contents by explicitly printing it with::
2168 You can view a macro's contents by explicitly printing it with::
2169
2169
2170 print macro_name
2170 print macro_name
2171
2171
2172 """
2172 """
2173 opts,args = self.parse_options(parameter_s,'r',mode='list')
2173 opts,args = self.parse_options(parameter_s,'r',mode='list')
2174 if not args: # List existing macros
2174 if not args: # List existing macros
2175 return sorted(k for k,v in self.shell.user_ns.iteritems() if\
2175 return sorted(k for k,v in self.shell.user_ns.iteritems() if\
2176 isinstance(v, Macro))
2176 isinstance(v, Macro))
2177 if len(args) == 1:
2177 if len(args) == 1:
2178 raise UsageError(
2178 raise UsageError(
2179 "%macro insufficient args; usage '%macro name n1-n2 n3-4...")
2179 "%macro insufficient args; usage '%macro name n1-n2 n3-4...")
2180 name, codefrom = args[0], " ".join(args[1:])
2180 name, codefrom = args[0], " ".join(args[1:])
2181
2181
2182 #print 'rng',ranges # dbg
2182 #print 'rng',ranges # dbg
2183 try:
2183 try:
2184 lines = self.shell.find_user_code(codefrom, 'r' in opts)
2184 lines = self.shell.find_user_code(codefrom, 'r' in opts)
2185 except (ValueError, TypeError) as e:
2185 except (ValueError, TypeError) as e:
2186 print e.args[0]
2186 print e.args[0]
2187 return
2187 return
2188 macro = Macro(lines)
2188 macro = Macro(lines)
2189 self.shell.define_macro(name, macro)
2189 self.shell.define_macro(name, macro)
2190 print 'Macro `%s` created. To execute, type its name (without quotes).' % name
2190 print 'Macro `%s` created. To execute, type its name (without quotes).' % name
2191 print '=== Macro contents: ==='
2191 print '=== Macro contents: ==='
2192 print macro,
2192 print macro,
2193
2193
2194 def magic_save(self,parameter_s = ''):
2194 def magic_save(self,parameter_s = ''):
2195 """Save a set of lines or a macro to a given filename.
2195 """Save a set of lines or a macro to a given filename.
2196
2196
2197 Usage:\\
2197 Usage:\\
2198 %save [options] filename n1-n2 n3-n4 ... n5 .. n6 ...
2198 %save [options] filename n1-n2 n3-n4 ... n5 .. n6 ...
2199
2199
2200 Options:
2200 Options:
2201
2201
2202 -r: use 'raw' input. By default, the 'processed' history is used,
2202 -r: use 'raw' input. By default, the 'processed' history is used,
2203 so that magics are loaded in their transformed version to valid
2203 so that magics are loaded in their transformed version to valid
2204 Python. If this option is given, the raw input as typed as the
2204 Python. If this option is given, the raw input as typed as the
2205 command line is used instead.
2205 command line is used instead.
2206
2206
2207 This function uses the same syntax as %history for input ranges,
2207 This function uses the same syntax as %history for input ranges,
2208 then saves the lines to the filename you specify.
2208 then saves the lines to the filename you specify.
2209
2209
2210 It adds a '.py' extension to the file if you don't do so yourself, and
2210 It adds a '.py' extension to the file if you don't do so yourself, and
2211 it asks for confirmation before overwriting existing files."""
2211 it asks for confirmation before overwriting existing files."""
2212
2212
2213 opts,args = self.parse_options(parameter_s,'r',mode='list')
2213 opts,args = self.parse_options(parameter_s,'r',mode='list')
2214 fname, codefrom = unquote_filename(args[0]), " ".join(args[1:])
2214 fname, codefrom = unquote_filename(args[0]), " ".join(args[1:])
2215 if not fname.endswith('.py'):
2215 if not fname.endswith('.py'):
2216 fname += '.py'
2216 fname += '.py'
2217 if os.path.isfile(fname):
2217 if os.path.isfile(fname):
2218 ans = raw_input('File `%s` exists. Overwrite (y/[N])? ' % fname)
2218 ans = raw_input('File `%s` exists. Overwrite (y/[N])? ' % fname)
2219 if ans.lower() not in ['y','yes']:
2219 if ans.lower() not in ['y','yes']:
2220 print 'Operation cancelled.'
2220 print 'Operation cancelled.'
2221 return
2221 return
2222 try:
2222 try:
2223 cmds = self.shell.find_user_code(codefrom, 'r' in opts)
2223 cmds = self.shell.find_user_code(codefrom, 'r' in opts)
2224 except (TypeError, ValueError) as e:
2224 except (TypeError, ValueError) as e:
2225 print e.args[0]
2225 print e.args[0]
2226 return
2226 return
2227 with io.open(fname,'w', encoding="utf-8") as f:
2227 with io.open(fname,'w', encoding="utf-8") as f:
2228 f.write(u"# coding: utf-8\n")
2228 f.write(u"# coding: utf-8\n")
2229 f.write(py3compat.cast_unicode(cmds))
2229 f.write(py3compat.cast_unicode(cmds))
2230 print 'The following commands were written to file `%s`:' % fname
2230 print 'The following commands were written to file `%s`:' % fname
2231 print cmds
2231 print cmds
2232
2232
2233 def magic_pastebin(self, parameter_s = ''):
2233 def magic_pastebin(self, parameter_s = ''):
2234 """Upload code to the 'Lodge it' paste bin, returning the URL."""
2234 """Upload code to the 'Lodge it' paste bin, returning the URL."""
2235 try:
2235 try:
2236 code = self.shell.find_user_code(parameter_s)
2236 code = self.shell.find_user_code(parameter_s)
2237 except (ValueError, TypeError) as e:
2237 except (ValueError, TypeError) as e:
2238 print e.args[0]
2238 print e.args[0]
2239 return
2239 return
2240 pbserver = ServerProxy('http://paste.pocoo.org/xmlrpc/')
2240 pbserver = ServerProxy('http://paste.pocoo.org/xmlrpc/')
2241 id = pbserver.pastes.newPaste("python", code)
2241 id = pbserver.pastes.newPaste("python", code)
2242 return "http://paste.pocoo.org/show/" + id
2242 return "http://paste.pocoo.org/show/" + id
2243
2243
2244 def magic_loadpy(self, arg_s):
2244 def magic_loadpy(self, arg_s):
2245 """Load a .py python script into the GUI console.
2245 """Load a .py python script into the GUI console.
2246
2246
2247 This magic command can either take a local filename or a url::
2247 This magic command can either take a local filename or a url::
2248
2248
2249 %loadpy myscript.py
2249 %loadpy myscript.py
2250 %loadpy http://www.example.com/myscript.py
2250 %loadpy http://www.example.com/myscript.py
2251 """
2251 """
2252 arg_s = unquote_filename(arg_s)
2252 arg_s = unquote_filename(arg_s)
2253 remote_url = arg_s.startswith(('http://', 'https://'))
2253 remote_url = arg_s.startswith(('http://', 'https://'))
2254 local_url = not remote_url
2254 local_url = not remote_url
2255 if local_url and not arg_s.endswith('.py'):
2255 if local_url and not arg_s.endswith('.py'):
2256 # Local files must be .py; for remote URLs it's possible that the
2256 # Local files must be .py; for remote URLs it's possible that the
2257 # fetch URL doesn't have a .py in it (many servers have an opaque
2257 # fetch URL doesn't have a .py in it (many servers have an opaque
2258 # URL, such as scipy-central.org).
2258 # URL, such as scipy-central.org).
2259 raise ValueError('%%loadpy only works with .py files: %s' % arg_s)
2259 raise ValueError('%%loadpy only works with .py files: %s' % arg_s)
2260
2260
2261 # openpy takes care of finding the source encoding (per PEP 263)
2261 # openpy takes care of finding the source encoding (per PEP 263)
2262 if remote_url:
2262 if remote_url:
2263 contents = openpy.read_py_url(arg_s, skip_encoding_cookie=True)
2263 contents = openpy.read_py_url(arg_s, skip_encoding_cookie=True)
2264 else:
2264 else:
2265 contents = openpy.read_py_file(arg_s, skip_encoding_cookie=True)
2265 contents = openpy.read_py_file(arg_s, skip_encoding_cookie=True)
2266
2266
2267 self.set_next_input(contents)
2267 self.set_next_input(contents)
2268
2268
2269 def _find_edit_target(self, args, opts, last_call):
2269 def _find_edit_target(self, args, opts, last_call):
2270 """Utility method used by magic_edit to find what to edit."""
2270 """Utility method used by magic_edit to find what to edit."""
2271
2271
2272 def make_filename(arg):
2272 def make_filename(arg):
2273 "Make a filename from the given args"
2273 "Make a filename from the given args"
2274 arg = unquote_filename(arg)
2274 arg = unquote_filename(arg)
2275 try:
2275 try:
2276 filename = get_py_filename(arg)
2276 filename = get_py_filename(arg)
2277 except IOError:
2277 except IOError:
2278 # If it ends with .py but doesn't already exist, assume we want
2278 # If it ends with .py but doesn't already exist, assume we want
2279 # a new file.
2279 # a new file.
2280 if arg.endswith('.py'):
2280 if arg.endswith('.py'):
2281 filename = arg
2281 filename = arg
2282 else:
2282 else:
2283 filename = None
2283 filename = None
2284 return filename
2284 return filename
2285
2285
2286 # Set a few locals from the options for convenience:
2286 # Set a few locals from the options for convenience:
2287 opts_prev = 'p' in opts
2287 opts_prev = 'p' in opts
2288 opts_raw = 'r' in opts
2288 opts_raw = 'r' in opts
2289
2289
2290 # custom exceptions
2290 # custom exceptions
2291 class DataIsObject(Exception): pass
2291 class DataIsObject(Exception): pass
2292
2292
2293 # Default line number value
2293 # Default line number value
2294 lineno = opts.get('n',None)
2294 lineno = opts.get('n',None)
2295
2295
2296 if opts_prev:
2296 if opts_prev:
2297 args = '_%s' % last_call[0]
2297 args = '_%s' % last_call[0]
2298 if not self.shell.user_ns.has_key(args):
2298 if not self.shell.user_ns.has_key(args):
2299 args = last_call[1]
2299 args = last_call[1]
2300
2300
2301 # use last_call to remember the state of the previous call, but don't
2301 # use last_call to remember the state of the previous call, but don't
2302 # let it be clobbered by successive '-p' calls.
2302 # let it be clobbered by successive '-p' calls.
2303 try:
2303 try:
2304 last_call[0] = self.shell.displayhook.prompt_count
2304 last_call[0] = self.shell.displayhook.prompt_count
2305 if not opts_prev:
2305 if not opts_prev:
2306 last_call[1] = args
2306 last_call[1] = args
2307 except:
2307 except:
2308 pass
2308 pass
2309
2309
2310 # by default this is done with temp files, except when the given
2310 # by default this is done with temp files, except when the given
2311 # arg is a filename
2311 # arg is a filename
2312 use_temp = True
2312 use_temp = True
2313
2313
2314 data = ''
2314 data = ''
2315
2315
2316 # First, see if the arguments should be a filename.
2316 # First, see if the arguments should be a filename.
2317 filename = make_filename(args)
2317 filename = make_filename(args)
2318 if filename:
2318 if filename:
2319 use_temp = False
2319 use_temp = False
2320 elif args:
2320 elif args:
2321 # Mode where user specifies ranges of lines, like in %macro.
2321 # Mode where user specifies ranges of lines, like in %macro.
2322 data = self.extract_input_lines(args, opts_raw)
2322 data = self.extract_input_lines(args, opts_raw)
2323 if not data:
2323 if not data:
2324 try:
2324 try:
2325 # Load the parameter given as a variable. If not a string,
2325 # Load the parameter given as a variable. If not a string,
2326 # process it as an object instead (below)
2326 # process it as an object instead (below)
2327
2327
2328 #print '*** args',args,'type',type(args) # dbg
2328 #print '*** args',args,'type',type(args) # dbg
2329 data = eval(args, self.shell.user_ns)
2329 data = eval(args, self.shell.user_ns)
2330 if not isinstance(data, basestring):
2330 if not isinstance(data, basestring):
2331 raise DataIsObject
2331 raise DataIsObject
2332
2332
2333 except (NameError,SyntaxError):
2333 except (NameError,SyntaxError):
2334 # given argument is not a variable, try as a filename
2334 # given argument is not a variable, try as a filename
2335 filename = make_filename(args)
2335 filename = make_filename(args)
2336 if filename is None:
2336 if filename is None:
2337 warn("Argument given (%s) can't be found as a variable "
2337 warn("Argument given (%s) can't be found as a variable "
2338 "or as a filename." % args)
2338 "or as a filename." % args)
2339 return
2339 return
2340 use_temp = False
2340 use_temp = False
2341
2341
2342 except DataIsObject:
2342 except DataIsObject:
2343 # macros have a special edit function
2343 # macros have a special edit function
2344 if isinstance(data, Macro):
2344 if isinstance(data, Macro):
2345 raise MacroToEdit(data)
2345 raise MacroToEdit(data)
2346
2346
2347 # For objects, try to edit the file where they are defined
2347 # For objects, try to edit the file where they are defined
2348 try:
2348 try:
2349 filename = inspect.getabsfile(data)
2349 filename = inspect.getabsfile(data)
2350 if 'fakemodule' in filename.lower() and inspect.isclass(data):
2350 if 'fakemodule' in filename.lower() and inspect.isclass(data):
2351 # class created by %edit? Try to find source
2351 # class created by %edit? Try to find source
2352 # by looking for method definitions instead, the
2352 # by looking for method definitions instead, the
2353 # __module__ in those classes is FakeModule.
2353 # __module__ in those classes is FakeModule.
2354 attrs = [getattr(data, aname) for aname in dir(data)]
2354 attrs = [getattr(data, aname) for aname in dir(data)]
2355 for attr in attrs:
2355 for attr in attrs:
2356 if not inspect.ismethod(attr):
2356 if not inspect.ismethod(attr):
2357 continue
2357 continue
2358 filename = inspect.getabsfile(attr)
2358 filename = inspect.getabsfile(attr)
2359 if filename and 'fakemodule' not in filename.lower():
2359 if filename and 'fakemodule' not in filename.lower():
2360 # change the attribute to be the edit target instead
2360 # change the attribute to be the edit target instead
2361 data = attr
2361 data = attr
2362 break
2362 break
2363
2363
2364 datafile = 1
2364 datafile = 1
2365 except TypeError:
2365 except TypeError:
2366 filename = make_filename(args)
2366 filename = make_filename(args)
2367 datafile = 1
2367 datafile = 1
2368 warn('Could not find file where `%s` is defined.\n'
2368 warn('Could not find file where `%s` is defined.\n'
2369 'Opening a file named `%s`' % (args,filename))
2369 'Opening a file named `%s`' % (args,filename))
2370 # Now, make sure we can actually read the source (if it was in
2370 # Now, make sure we can actually read the source (if it was in
2371 # a temp file it's gone by now).
2371 # a temp file it's gone by now).
2372 if datafile:
2372 if datafile:
2373 try:
2373 try:
2374 if lineno is None:
2374 if lineno is None:
2375 lineno = inspect.getsourcelines(data)[1]
2375 lineno = inspect.getsourcelines(data)[1]
2376 except IOError:
2376 except IOError:
2377 filename = make_filename(args)
2377 filename = make_filename(args)
2378 if filename is None:
2378 if filename is None:
2379 warn('The file `%s` where `%s` was defined cannot '
2379 warn('The file `%s` where `%s` was defined cannot '
2380 'be read.' % (filename,data))
2380 'be read.' % (filename,data))
2381 return
2381 return
2382 use_temp = False
2382 use_temp = False
2383
2383
2384 if use_temp:
2384 if use_temp:
2385 filename = self.shell.mktempfile(data)
2385 filename = self.shell.mktempfile(data)
2386 print 'IPython will make a temporary file named:',filename
2386 print 'IPython will make a temporary file named:',filename
2387
2387
2388 return filename, lineno, use_temp
2388 return filename, lineno, use_temp
2389
2389
2390 def _edit_macro(self,mname,macro):
2390 def _edit_macro(self,mname,macro):
2391 """open an editor with the macro data in a file"""
2391 """open an editor with the macro data in a file"""
2392 filename = self.shell.mktempfile(macro.value)
2392 filename = self.shell.mktempfile(macro.value)
2393 self.shell.hooks.editor(filename)
2393 self.shell.hooks.editor(filename)
2394
2394
2395 # and make a new macro object, to replace the old one
2395 # and make a new macro object, to replace the old one
2396 mfile = open(filename)
2396 mfile = open(filename)
2397 mvalue = mfile.read()
2397 mvalue = mfile.read()
2398 mfile.close()
2398 mfile.close()
2399 self.shell.user_ns[mname] = Macro(mvalue)
2399 self.shell.user_ns[mname] = Macro(mvalue)
2400
2400
2401 def magic_ed(self,parameter_s=''):
2401 def magic_ed(self,parameter_s=''):
2402 """Alias to %edit."""
2402 """Alias to %edit."""
2403 return self.magic_edit(parameter_s)
2403 return self.magic_edit(parameter_s)
2404
2404
2405 @skip_doctest
2405 @skip_doctest
2406 def magic_edit(self,parameter_s='',last_call=['','']):
2406 def magic_edit(self,parameter_s='',last_call=['','']):
2407 """Bring up an editor and execute the resulting code.
2407 """Bring up an editor and execute the resulting code.
2408
2408
2409 Usage:
2409 Usage:
2410 %edit [options] [args]
2410 %edit [options] [args]
2411
2411
2412 %edit runs IPython's editor hook. The default version of this hook is
2412 %edit runs IPython's editor hook. The default version of this hook is
2413 set to call the editor specified by your $EDITOR environment variable.
2413 set to call the editor specified by your $EDITOR environment variable.
2414 If this isn't found, it will default to vi under Linux/Unix and to
2414 If this isn't found, it will default to vi under Linux/Unix and to
2415 notepad under Windows. See the end of this docstring for how to change
2415 notepad under Windows. See the end of this docstring for how to change
2416 the editor hook.
2416 the editor hook.
2417
2417
2418 You can also set the value of this editor via the
2418 You can also set the value of this editor via the
2419 ``TerminalInteractiveShell.editor`` option in your configuration file.
2419 ``TerminalInteractiveShell.editor`` option in your configuration file.
2420 This is useful if you wish to use a different editor from your typical
2420 This is useful if you wish to use a different editor from your typical
2421 default with IPython (and for Windows users who typically don't set
2421 default with IPython (and for Windows users who typically don't set
2422 environment variables).
2422 environment variables).
2423
2423
2424 This command allows you to conveniently edit multi-line code right in
2424 This command allows you to conveniently edit multi-line code right in
2425 your IPython session.
2425 your IPython session.
2426
2426
2427 If called without arguments, %edit opens up an empty editor with a
2427 If called without arguments, %edit opens up an empty editor with a
2428 temporary file and will execute the contents of this file when you
2428 temporary file and will execute the contents of this file when you
2429 close it (don't forget to save it!).
2429 close it (don't forget to save it!).
2430
2430
2431
2431
2432 Options:
2432 Options:
2433
2433
2434 -n <number>: open the editor at a specified line number. By default,
2434 -n <number>: open the editor at a specified line number. By default,
2435 the IPython editor hook uses the unix syntax 'editor +N filename', but
2435 the IPython editor hook uses the unix syntax 'editor +N filename', but
2436 you can configure this by providing your own modified hook if your
2436 you can configure this by providing your own modified hook if your
2437 favorite editor supports line-number specifications with a different
2437 favorite editor supports line-number specifications with a different
2438 syntax.
2438 syntax.
2439
2439
2440 -p: this will call the editor with the same data as the previous time
2440 -p: this will call the editor with the same data as the previous time
2441 it was used, regardless of how long ago (in your current session) it
2441 it was used, regardless of how long ago (in your current session) it
2442 was.
2442 was.
2443
2443
2444 -r: use 'raw' input. This option only applies to input taken from the
2444 -r: use 'raw' input. This option only applies to input taken from the
2445 user's history. By default, the 'processed' history is used, so that
2445 user's history. By default, the 'processed' history is used, so that
2446 magics are loaded in their transformed version to valid Python. If
2446 magics are loaded in their transformed version to valid Python. If
2447 this option is given, the raw input as typed as the command line is
2447 this option is given, the raw input as typed as the command line is
2448 used instead. When you exit the editor, it will be executed by
2448 used instead. When you exit the editor, it will be executed by
2449 IPython's own processor.
2449 IPython's own processor.
2450
2450
2451 -x: do not execute the edited code immediately upon exit. This is
2451 -x: do not execute the edited code immediately upon exit. This is
2452 mainly useful if you are editing programs which need to be called with
2452 mainly useful if you are editing programs which need to be called with
2453 command line arguments, which you can then do using %run.
2453 command line arguments, which you can then do using %run.
2454
2454
2455
2455
2456 Arguments:
2456 Arguments:
2457
2457
2458 If arguments are given, the following possibilities exist:
2458 If arguments are given, the following possibilities exist:
2459
2459
2460 - If the argument is a filename, IPython will load that into the
2460 - If the argument is a filename, IPython will load that into the
2461 editor. It will execute its contents with execfile() when you exit,
2461 editor. It will execute its contents with execfile() when you exit,
2462 loading any code in the file into your interactive namespace.
2462 loading any code in the file into your interactive namespace.
2463
2463
2464 - The arguments are ranges of input history, e.g. "7 ~1/4-6".
2464 - The arguments are ranges of input history, e.g. "7 ~1/4-6".
2465 The syntax is the same as in the %history magic.
2465 The syntax is the same as in the %history magic.
2466
2466
2467 - If the argument is a string variable, its contents are loaded
2467 - If the argument is a string variable, its contents are loaded
2468 into the editor. You can thus edit any string which contains
2468 into the editor. You can thus edit any string which contains
2469 python code (including the result of previous edits).
2469 python code (including the result of previous edits).
2470
2470
2471 - If the argument is the name of an object (other than a string),
2471 - If the argument is the name of an object (other than a string),
2472 IPython will try to locate the file where it was defined and open the
2472 IPython will try to locate the file where it was defined and open the
2473 editor at the point where it is defined. You can use `%edit function`
2473 editor at the point where it is defined. You can use `%edit function`
2474 to load an editor exactly at the point where 'function' is defined,
2474 to load an editor exactly at the point where 'function' is defined,
2475 edit it and have the file be executed automatically.
2475 edit it and have the file be executed automatically.
2476
2476
2477 - If the object is a macro (see %macro for details), this opens up your
2477 - If the object is a macro (see %macro for details), this opens up your
2478 specified editor with a temporary file containing the macro's data.
2478 specified editor with a temporary file containing the macro's data.
2479 Upon exit, the macro is reloaded with the contents of the file.
2479 Upon exit, the macro is reloaded with the contents of the file.
2480
2480
2481 Note: opening at an exact line is only supported under Unix, and some
2481 Note: opening at an exact line is only supported under Unix, and some
2482 editors (like kedit and gedit up to Gnome 2.8) do not understand the
2482 editors (like kedit and gedit up to Gnome 2.8) do not understand the
2483 '+NUMBER' parameter necessary for this feature. Good editors like
2483 '+NUMBER' parameter necessary for this feature. Good editors like
2484 (X)Emacs, vi, jed, pico and joe all do.
2484 (X)Emacs, vi, jed, pico and joe all do.
2485
2485
2486 After executing your code, %edit will return as output the code you
2486 After executing your code, %edit will return as output the code you
2487 typed in the editor (except when it was an existing file). This way
2487 typed in the editor (except when it was an existing file). This way
2488 you can reload the code in further invocations of %edit as a variable,
2488 you can reload the code in further invocations of %edit as a variable,
2489 via _<NUMBER> or Out[<NUMBER>], where <NUMBER> is the prompt number of
2489 via _<NUMBER> or Out[<NUMBER>], where <NUMBER> is the prompt number of
2490 the output.
2490 the output.
2491
2491
2492 Note that %edit is also available through the alias %ed.
2492 Note that %edit is also available through the alias %ed.
2493
2493
2494 This is an example of creating a simple function inside the editor and
2494 This is an example of creating a simple function inside the editor and
2495 then modifying it. First, start up the editor::
2495 then modifying it. First, start up the editor::
2496
2496
2497 In [1]: ed
2497 In [1]: ed
2498 Editing... done. Executing edited code...
2498 Editing... done. Executing edited code...
2499 Out[1]: 'def foo():\\n print "foo() was defined in an editing
2499 Out[1]: 'def foo():\\n print "foo() was defined in an editing
2500 session"\\n'
2500 session"\\n'
2501
2501
2502 We can then call the function foo()::
2502 We can then call the function foo()::
2503
2503
2504 In [2]: foo()
2504 In [2]: foo()
2505 foo() was defined in an editing session
2505 foo() was defined in an editing session
2506
2506
2507 Now we edit foo. IPython automatically loads the editor with the
2507 Now we edit foo. IPython automatically loads the editor with the
2508 (temporary) file where foo() was previously defined::
2508 (temporary) file where foo() was previously defined::
2509
2509
2510 In [3]: ed foo
2510 In [3]: ed foo
2511 Editing... done. Executing edited code...
2511 Editing... done. Executing edited code...
2512
2512
2513 And if we call foo() again we get the modified version::
2513 And if we call foo() again we get the modified version::
2514
2514
2515 In [4]: foo()
2515 In [4]: foo()
2516 foo() has now been changed!
2516 foo() has now been changed!
2517
2517
2518 Here is an example of how to edit a code snippet successive
2518 Here is an example of how to edit a code snippet successive
2519 times. First we call the editor::
2519 times. First we call the editor::
2520
2520
2521 In [5]: ed
2521 In [5]: ed
2522 Editing... done. Executing edited code...
2522 Editing... done. Executing edited code...
2523 hello
2523 hello
2524 Out[5]: "print 'hello'\\n"
2524 Out[5]: "print 'hello'\\n"
2525
2525
2526 Now we call it again with the previous output (stored in _)::
2526 Now we call it again with the previous output (stored in _)::
2527
2527
2528 In [6]: ed _
2528 In [6]: ed _
2529 Editing... done. Executing edited code...
2529 Editing... done. Executing edited code...
2530 hello world
2530 hello world
2531 Out[6]: "print 'hello world'\\n"
2531 Out[6]: "print 'hello world'\\n"
2532
2532
2533 Now we call it with the output #8 (stored in _8, also as Out[8])::
2533 Now we call it with the output #8 (stored in _8, also as Out[8])::
2534
2534
2535 In [7]: ed _8
2535 In [7]: ed _8
2536 Editing... done. Executing edited code...
2536 Editing... done. Executing edited code...
2537 hello again
2537 hello again
2538 Out[7]: "print 'hello again'\\n"
2538 Out[7]: "print 'hello again'\\n"
2539
2539
2540
2540
2541 Changing the default editor hook:
2541 Changing the default editor hook:
2542
2542
2543 If you wish to write your own editor hook, you can put it in a
2543 If you wish to write your own editor hook, you can put it in a
2544 configuration file which you load at startup time. The default hook
2544 configuration file which you load at startup time. The default hook
2545 is defined in the IPython.core.hooks module, and you can use that as a
2545 is defined in the IPython.core.hooks module, and you can use that as a
2546 starting example for further modifications. That file also has
2546 starting example for further modifications. That file also has
2547 general instructions on how to set a new hook for use once you've
2547 general instructions on how to set a new hook for use once you've
2548 defined it."""
2548 defined it."""
2549 opts,args = self.parse_options(parameter_s,'prxn:')
2549 opts,args = self.parse_options(parameter_s,'prxn:')
2550
2550
2551 try:
2551 try:
2552 filename, lineno, is_temp = self._find_edit_target(args, opts, last_call)
2552 filename, lineno, is_temp = self._find_edit_target(args, opts, last_call)
2553 except MacroToEdit as e:
2553 except MacroToEdit as e:
2554 self._edit_macro(args, e.args[0])
2554 self._edit_macro(args, e.args[0])
2555 return
2555 return
2556
2556
2557 # do actual editing here
2557 # do actual editing here
2558 print 'Editing...',
2558 print 'Editing...',
2559 sys.stdout.flush()
2559 sys.stdout.flush()
2560 try:
2560 try:
2561 # Quote filenames that may have spaces in them
2561 # Quote filenames that may have spaces in them
2562 if ' ' in filename:
2562 if ' ' in filename:
2563 filename = "'%s'" % filename
2563 filename = "'%s'" % filename
2564 self.shell.hooks.editor(filename,lineno)
2564 self.shell.hooks.editor(filename,lineno)
2565 except TryNext:
2565 except TryNext:
2566 warn('Could not open editor')
2566 warn('Could not open editor')
2567 return
2567 return
2568
2568
2569 # XXX TODO: should this be generalized for all string vars?
2569 # XXX TODO: should this be generalized for all string vars?
2570 # For now, this is special-cased to blocks created by cpaste
2570 # For now, this is special-cased to blocks created by cpaste
2571 if args.strip() == 'pasted_block':
2571 if args.strip() == 'pasted_block':
2572 self.shell.user_ns['pasted_block'] = file_read(filename)
2572 self.shell.user_ns['pasted_block'] = file_read(filename)
2573
2573
2574 if 'x' in opts: # -x prevents actual execution
2574 if 'x' in opts: # -x prevents actual execution
2575 print
2575 print
2576 else:
2576 else:
2577 print 'done. Executing edited code...'
2577 print 'done. Executing edited code...'
2578 if 'r' in opts: # Untranslated IPython code
2578 if 'r' in opts: # Untranslated IPython code
2579 self.shell.run_cell(file_read(filename),
2579 self.shell.run_cell(file_read(filename),
2580 store_history=False)
2580 store_history=False)
2581 else:
2581 else:
2582 self.shell.safe_execfile(filename,self.shell.user_ns,
2582 self.shell.safe_execfile(filename,self.shell.user_ns,
2583 self.shell.user_ns)
2583 self.shell.user_ns)
2584
2584
2585 if is_temp:
2585 if is_temp:
2586 try:
2586 try:
2587 return open(filename).read()
2587 return open(filename).read()
2588 except IOError,msg:
2588 except IOError,msg:
2589 if msg.filename == filename:
2589 if msg.filename == filename:
2590 warn('File not found. Did you forget to save?')
2590 warn('File not found. Did you forget to save?')
2591 return
2591 return
2592 else:
2592 else:
2593 self.shell.showtraceback()
2593 self.shell.showtraceback()
2594
2594
2595 def magic_xmode(self,parameter_s = ''):
2595 def magic_xmode(self,parameter_s = ''):
2596 """Switch modes for the exception handlers.
2596 """Switch modes for the exception handlers.
2597
2597
2598 Valid modes: Plain, Context and Verbose.
2598 Valid modes: Plain, Context and Verbose.
2599
2599
2600 If called without arguments, acts as a toggle."""
2600 If called without arguments, acts as a toggle."""
2601
2601
2602 def xmode_switch_err(name):
2602 def xmode_switch_err(name):
2603 warn('Error changing %s exception modes.\n%s' %
2603 warn('Error changing %s exception modes.\n%s' %
2604 (name,sys.exc_info()[1]))
2604 (name,sys.exc_info()[1]))
2605
2605
2606 shell = self.shell
2606 shell = self.shell
2607 new_mode = parameter_s.strip().capitalize()
2607 new_mode = parameter_s.strip().capitalize()
2608 try:
2608 try:
2609 shell.InteractiveTB.set_mode(mode=new_mode)
2609 shell.InteractiveTB.set_mode(mode=new_mode)
2610 print 'Exception reporting mode:',shell.InteractiveTB.mode
2610 print 'Exception reporting mode:',shell.InteractiveTB.mode
2611 except:
2611 except:
2612 xmode_switch_err('user')
2612 xmode_switch_err('user')
2613
2613
2614 def magic_colors(self,parameter_s = ''):
2614 def magic_colors(self,parameter_s = ''):
2615 """Switch color scheme for prompts, info system and exception handlers.
2615 """Switch color scheme for prompts, info system and exception handlers.
2616
2616
2617 Currently implemented schemes: NoColor, Linux, LightBG.
2617 Currently implemented schemes: NoColor, Linux, LightBG.
2618
2618
2619 Color scheme names are not case-sensitive.
2619 Color scheme names are not case-sensitive.
2620
2620
2621 Examples
2621 Examples
2622 --------
2622 --------
2623 To get a plain black and white terminal::
2623 To get a plain black and white terminal::
2624
2624
2625 %colors nocolor
2625 %colors nocolor
2626 """
2626 """
2627
2627
2628 def color_switch_err(name):
2628 def color_switch_err(name):
2629 warn('Error changing %s color schemes.\n%s' %
2629 warn('Error changing %s color schemes.\n%s' %
2630 (name,sys.exc_info()[1]))
2630 (name,sys.exc_info()[1]))
2631
2631
2632
2632
2633 new_scheme = parameter_s.strip()
2633 new_scheme = parameter_s.strip()
2634 if not new_scheme:
2634 if not new_scheme:
2635 raise UsageError(
2635 raise UsageError(
2636 "%colors: you must specify a color scheme. See '%colors?'")
2636 "%colors: you must specify a color scheme. See '%colors?'")
2637 return
2637 return
2638 # local shortcut
2638 # local shortcut
2639 shell = self.shell
2639 shell = self.shell
2640
2640
2641 import IPython.utils.rlineimpl as readline
2641 import IPython.utils.rlineimpl as readline
2642
2642
2643 if not shell.colors_force and \
2643 if not shell.colors_force and \
2644 not readline.have_readline and sys.platform == "win32":
2644 not readline.have_readline and sys.platform == "win32":
2645 msg = """\
2645 msg = """\
2646 Proper color support under MS Windows requires the pyreadline library.
2646 Proper color support under MS Windows requires the pyreadline library.
2647 You can find it at:
2647 You can find it at:
2648 http://ipython.org/pyreadline.html
2648 http://ipython.org/pyreadline.html
2649 Gary's readline needs the ctypes module, from:
2649 Gary's readline needs the ctypes module, from:
2650 http://starship.python.net/crew/theller/ctypes
2650 http://starship.python.net/crew/theller/ctypes
2651 (Note that ctypes is already part of Python versions 2.5 and newer).
2651 (Note that ctypes is already part of Python versions 2.5 and newer).
2652
2652
2653 Defaulting color scheme to 'NoColor'"""
2653 Defaulting color scheme to 'NoColor'"""
2654 new_scheme = 'NoColor'
2654 new_scheme = 'NoColor'
2655 warn(msg)
2655 warn(msg)
2656
2656
2657 # readline option is 0
2657 # readline option is 0
2658 if not shell.colors_force and not shell.has_readline:
2658 if not shell.colors_force and not shell.has_readline:
2659 new_scheme = 'NoColor'
2659 new_scheme = 'NoColor'
2660
2660
2661 # Set prompt colors
2661 # Set prompt colors
2662 try:
2662 try:
2663 shell.prompt_manager.color_scheme = new_scheme
2663 shell.prompt_manager.color_scheme = new_scheme
2664 except:
2664 except:
2665 color_switch_err('prompt')
2665 color_switch_err('prompt')
2666 else:
2666 else:
2667 shell.colors = \
2667 shell.colors = \
2668 shell.prompt_manager.color_scheme_table.active_scheme_name
2668 shell.prompt_manager.color_scheme_table.active_scheme_name
2669 # Set exception colors
2669 # Set exception colors
2670 try:
2670 try:
2671 shell.InteractiveTB.set_colors(scheme = new_scheme)
2671 shell.InteractiveTB.set_colors(scheme = new_scheme)
2672 shell.SyntaxTB.set_colors(scheme = new_scheme)
2672 shell.SyntaxTB.set_colors(scheme = new_scheme)
2673 except:
2673 except:
2674 color_switch_err('exception')
2674 color_switch_err('exception')
2675
2675
2676 # Set info (for 'object?') colors
2676 # Set info (for 'object?') colors
2677 if shell.color_info:
2677 if shell.color_info:
2678 try:
2678 try:
2679 shell.inspector.set_active_scheme(new_scheme)
2679 shell.inspector.set_active_scheme(new_scheme)
2680 except:
2680 except:
2681 color_switch_err('object inspector')
2681 color_switch_err('object inspector')
2682 else:
2682 else:
2683 shell.inspector.set_active_scheme('NoColor')
2683 shell.inspector.set_active_scheme('NoColor')
2684
2684
2685 def magic_pprint(self, parameter_s=''):
2685 def magic_pprint(self, parameter_s=''):
2686 """Toggle pretty printing on/off."""
2686 """Toggle pretty printing on/off."""
2687 ptformatter = self.shell.display_formatter.formatters['text/plain']
2687 ptformatter = self.shell.display_formatter.formatters['text/plain']
2688 ptformatter.pprint = bool(1 - ptformatter.pprint)
2688 ptformatter.pprint = bool(1 - ptformatter.pprint)
2689 print 'Pretty printing has been turned', \
2689 print 'Pretty printing has been turned', \
2690 ['OFF','ON'][ptformatter.pprint]
2690 ['OFF','ON'][ptformatter.pprint]
2691
2691
2692 #......................................................................
2692 #......................................................................
2693 # Functions to implement unix shell-type things
2693 # Functions to implement unix shell-type things
2694
2694
2695 @skip_doctest
2695 @skip_doctest
2696 def magic_alias(self, parameter_s = ''):
2696 def magic_alias(self, parameter_s = ''):
2697 """Define an alias for a system command.
2697 """Define an alias for a system command.
2698
2698
2699 '%alias alias_name cmd' defines 'alias_name' as an alias for 'cmd'
2699 '%alias alias_name cmd' defines 'alias_name' as an alias for 'cmd'
2700
2700
2701 Then, typing 'alias_name params' will execute the system command 'cmd
2701 Then, typing 'alias_name params' will execute the system command 'cmd
2702 params' (from your underlying operating system).
2702 params' (from your underlying operating system).
2703
2703
2704 Aliases have lower precedence than magic functions and Python normal
2704 Aliases have lower precedence than magic functions and Python normal
2705 variables, so if 'foo' is both a Python variable and an alias, the
2705 variables, so if 'foo' is both a Python variable and an alias, the
2706 alias can not be executed until 'del foo' removes the Python variable.
2706 alias can not be executed until 'del foo' removes the Python variable.
2707
2707
2708 You can use the %l specifier in an alias definition to represent the
2708 You can use the %l specifier in an alias definition to represent the
2709 whole line when the alias is called. For example::
2709 whole line when the alias is called. For example::
2710
2710
2711 In [2]: alias bracket echo "Input in brackets: <%l>"
2711 In [2]: alias bracket echo "Input in brackets: <%l>"
2712 In [3]: bracket hello world
2712 In [3]: bracket hello world
2713 Input in brackets: <hello world>
2713 Input in brackets: <hello world>
2714
2714
2715 You can also define aliases with parameters using %s specifiers (one
2715 You can also define aliases with parameters using %s specifiers (one
2716 per parameter)::
2716 per parameter)::
2717
2717
2718 In [1]: alias parts echo first %s second %s
2718 In [1]: alias parts echo first %s second %s
2719 In [2]: %parts A B
2719 In [2]: %parts A B
2720 first A second B
2720 first A second B
2721 In [3]: %parts A
2721 In [3]: %parts A
2722 Incorrect number of arguments: 2 expected.
2722 Incorrect number of arguments: 2 expected.
2723 parts is an alias to: 'echo first %s second %s'
2723 parts is an alias to: 'echo first %s second %s'
2724
2724
2725 Note that %l and %s are mutually exclusive. You can only use one or
2725 Note that %l and %s are mutually exclusive. You can only use one or
2726 the other in your aliases.
2726 the other in your aliases.
2727
2727
2728 Aliases expand Python variables just like system calls using ! or !!
2728 Aliases expand Python variables just like system calls using ! or !!
2729 do: all expressions prefixed with '$' get expanded. For details of
2729 do: all expressions prefixed with '$' get expanded. For details of
2730 the semantic rules, see PEP-215:
2730 the semantic rules, see PEP-215:
2731 http://www.python.org/peps/pep-0215.html. This is the library used by
2731 http://www.python.org/peps/pep-0215.html. This is the library used by
2732 IPython for variable expansion. If you want to access a true shell
2732 IPython for variable expansion. If you want to access a true shell
2733 variable, an extra $ is necessary to prevent its expansion by
2733 variable, an extra $ is necessary to prevent its expansion by
2734 IPython::
2734 IPython::
2735
2735
2736 In [6]: alias show echo
2736 In [6]: alias show echo
2737 In [7]: PATH='A Python string'
2737 In [7]: PATH='A Python string'
2738 In [8]: show $PATH
2738 In [8]: show $PATH
2739 A Python string
2739 A Python string
2740 In [9]: show $$PATH
2740 In [9]: show $$PATH
2741 /usr/local/lf9560/bin:/usr/local/intel/compiler70/ia32/bin:...
2741 /usr/local/lf9560/bin:/usr/local/intel/compiler70/ia32/bin:...
2742
2742
2743 You can use the alias facility to acess all of $PATH. See the %rehash
2743 You can use the alias facility to acess all of $PATH. See the %rehash
2744 and %rehashx functions, which automatically create aliases for the
2744 and %rehashx functions, which automatically create aliases for the
2745 contents of your $PATH.
2745 contents of your $PATH.
2746
2746
2747 If called with no parameters, %alias prints the current alias table."""
2747 If called with no parameters, %alias prints the current alias table."""
2748
2748
2749 par = parameter_s.strip()
2749 par = parameter_s.strip()
2750 if not par:
2750 if not par:
2751 stored = self.db.get('stored_aliases', {} )
2751 stored = self.db.get('stored_aliases', {} )
2752 aliases = sorted(self.shell.alias_manager.aliases)
2752 aliases = sorted(self.shell.alias_manager.aliases)
2753 # for k, v in stored:
2753 # for k, v in stored:
2754 # atab.append(k, v[0])
2754 # atab.append(k, v[0])
2755
2755
2756 print "Total number of aliases:", len(aliases)
2756 print "Total number of aliases:", len(aliases)
2757 sys.stdout.flush()
2757 sys.stdout.flush()
2758 return aliases
2758 return aliases
2759
2759
2760 # Now try to define a new one
2760 # Now try to define a new one
2761 try:
2761 try:
2762 alias,cmd = par.split(None, 1)
2762 alias,cmd = par.split(None, 1)
2763 except:
2763 except:
2764 print oinspect.getdoc(self.magic_alias)
2764 print oinspect.getdoc(self.magic_alias)
2765 else:
2765 else:
2766 self.shell.alias_manager.soft_define_alias(alias, cmd)
2766 self.shell.alias_manager.soft_define_alias(alias, cmd)
2767 # end magic_alias
2767 # end magic_alias
2768
2768
2769 def magic_unalias(self, parameter_s = ''):
2769 def magic_unalias(self, parameter_s = ''):
2770 """Remove an alias"""
2770 """Remove an alias"""
2771
2771
2772 aname = parameter_s.strip()
2772 aname = parameter_s.strip()
2773 self.shell.alias_manager.undefine_alias(aname)
2773 self.shell.alias_manager.undefine_alias(aname)
2774 stored = self.db.get('stored_aliases', {} )
2774 stored = self.db.get('stored_aliases', {} )
2775 if aname in stored:
2775 if aname in stored:
2776 print "Removing %stored alias",aname
2776 print "Removing %stored alias",aname
2777 del stored[aname]
2777 del stored[aname]
2778 self.db['stored_aliases'] = stored
2778 self.db['stored_aliases'] = stored
2779
2779
2780 def magic_rehashx(self, parameter_s = ''):
2780 def magic_rehashx(self, parameter_s = ''):
2781 """Update the alias table with all executable files in $PATH.
2781 """Update the alias table with all executable files in $PATH.
2782
2782
2783 This version explicitly checks that every entry in $PATH is a file
2783 This version explicitly checks that every entry in $PATH is a file
2784 with execute access (os.X_OK), so it is much slower than %rehash.
2784 with execute access (os.X_OK), so it is much slower than %rehash.
2785
2785
2786 Under Windows, it checks executability as a match against a
2786 Under Windows, it checks executability as a match against a
2787 '|'-separated string of extensions, stored in the IPython config
2787 '|'-separated string of extensions, stored in the IPython config
2788 variable win_exec_ext. This defaults to 'exe|com|bat'.
2788 variable win_exec_ext. This defaults to 'exe|com|bat'.
2789
2789
2790 This function also resets the root module cache of module completer,
2790 This function also resets the root module cache of module completer,
2791 used on slow filesystems.
2791 used on slow filesystems.
2792 """
2792 """
2793 from IPython.core.alias import InvalidAliasError
2793 from IPython.core.alias import InvalidAliasError
2794
2794
2795 # for the benefit of module completer in ipy_completers.py
2795 # for the benefit of module completer in ipy_completers.py
2796 del self.shell.db['rootmodules']
2796 del self.shell.db['rootmodules']
2797
2797
2798 path = [os.path.abspath(os.path.expanduser(p)) for p in
2798 path = [os.path.abspath(os.path.expanduser(p)) for p in
2799 os.environ.get('PATH','').split(os.pathsep)]
2799 os.environ.get('PATH','').split(os.pathsep)]
2800 path = filter(os.path.isdir,path)
2800 path = filter(os.path.isdir,path)
2801
2801
2802 syscmdlist = []
2802 syscmdlist = []
2803 # Now define isexec in a cross platform manner.
2803 # Now define isexec in a cross platform manner.
2804 if os.name == 'posix':
2804 if os.name == 'posix':
2805 isexec = lambda fname:os.path.isfile(fname) and \
2805 isexec = lambda fname:os.path.isfile(fname) and \
2806 os.access(fname,os.X_OK)
2806 os.access(fname,os.X_OK)
2807 else:
2807 else:
2808 try:
2808 try:
2809 winext = os.environ['pathext'].replace(';','|').replace('.','')
2809 winext = os.environ['pathext'].replace(';','|').replace('.','')
2810 except KeyError:
2810 except KeyError:
2811 winext = 'exe|com|bat|py'
2811 winext = 'exe|com|bat|py'
2812 if 'py' not in winext:
2812 if 'py' not in winext:
2813 winext += '|py'
2813 winext += '|py'
2814 execre = re.compile(r'(.*)\.(%s)$' % winext,re.IGNORECASE)
2814 execre = re.compile(r'(.*)\.(%s)$' % winext,re.IGNORECASE)
2815 isexec = lambda fname:os.path.isfile(fname) and execre.match(fname)
2815 isexec = lambda fname:os.path.isfile(fname) and execre.match(fname)
2816 savedir = os.getcwdu()
2816 savedir = os.getcwdu()
2817
2817
2818 # Now walk the paths looking for executables to alias.
2818 # Now walk the paths looking for executables to alias.
2819 try:
2819 try:
2820 # write the whole loop for posix/Windows so we don't have an if in
2820 # write the whole loop for posix/Windows so we don't have an if in
2821 # the innermost part
2821 # the innermost part
2822 if os.name == 'posix':
2822 if os.name == 'posix':
2823 for pdir in path:
2823 for pdir in path:
2824 os.chdir(pdir)
2824 os.chdir(pdir)
2825 for ff in os.listdir(pdir):
2825 for ff in os.listdir(pdir):
2826 if isexec(ff):
2826 if isexec(ff):
2827 try:
2827 try:
2828 # Removes dots from the name since ipython
2828 # Removes dots from the name since ipython
2829 # will assume names with dots to be python.
2829 # will assume names with dots to be python.
2830 self.shell.alias_manager.define_alias(
2830 self.shell.alias_manager.define_alias(
2831 ff.replace('.',''), ff)
2831 ff.replace('.',''), ff)
2832 except InvalidAliasError:
2832 except InvalidAliasError:
2833 pass
2833 pass
2834 else:
2834 else:
2835 syscmdlist.append(ff)
2835 syscmdlist.append(ff)
2836 else:
2836 else:
2837 no_alias = self.shell.alias_manager.no_alias
2837 no_alias = self.shell.alias_manager.no_alias
2838 for pdir in path:
2838 for pdir in path:
2839 os.chdir(pdir)
2839 os.chdir(pdir)
2840 for ff in os.listdir(pdir):
2840 for ff in os.listdir(pdir):
2841 base, ext = os.path.splitext(ff)
2841 base, ext = os.path.splitext(ff)
2842 if isexec(ff) and base.lower() not in no_alias:
2842 if isexec(ff) and base.lower() not in no_alias:
2843 if ext.lower() == '.exe':
2843 if ext.lower() == '.exe':
2844 ff = base
2844 ff = base
2845 try:
2845 try:
2846 # Removes dots from the name since ipython
2846 # Removes dots from the name since ipython
2847 # will assume names with dots to be python.
2847 # will assume names with dots to be python.
2848 self.shell.alias_manager.define_alias(
2848 self.shell.alias_manager.define_alias(
2849 base.lower().replace('.',''), ff)
2849 base.lower().replace('.',''), ff)
2850 except InvalidAliasError:
2850 except InvalidAliasError:
2851 pass
2851 pass
2852 syscmdlist.append(ff)
2852 syscmdlist.append(ff)
2853 self.shell.db['syscmdlist'] = syscmdlist
2853 self.shell.db['syscmdlist'] = syscmdlist
2854 finally:
2854 finally:
2855 os.chdir(savedir)
2855 os.chdir(savedir)
2856
2856
2857 @skip_doctest
2857 @skip_doctest
2858 def magic_pwd(self, parameter_s = ''):
2858 def magic_pwd(self, parameter_s = ''):
2859 """Return the current working directory path.
2859 """Return the current working directory path.
2860
2860
2861 Examples
2861 Examples
2862 --------
2862 --------
2863 ::
2863 ::
2864
2864
2865 In [9]: pwd
2865 In [9]: pwd
2866 Out[9]: '/home/tsuser/sprint/ipython'
2866 Out[9]: '/home/tsuser/sprint/ipython'
2867 """
2867 """
2868 return os.getcwdu()
2868 return os.getcwdu()
2869
2869
2870 @skip_doctest
2870 @skip_doctest
2871 def magic_cd(self, parameter_s=''):
2871 def magic_cd(self, parameter_s=''):
2872 """Change the current working directory.
2872 """Change the current working directory.
2873
2873
2874 This command automatically maintains an internal list of directories
2874 This command automatically maintains an internal list of directories
2875 you visit during your IPython session, in the variable _dh. The
2875 you visit during your IPython session, in the variable _dh. The
2876 command %dhist shows this history nicely formatted. You can also
2876 command %dhist shows this history nicely formatted. You can also
2877 do 'cd -<tab>' to see directory history conveniently.
2877 do 'cd -<tab>' to see directory history conveniently.
2878
2878
2879 Usage:
2879 Usage:
2880
2880
2881 cd 'dir': changes to directory 'dir'.
2881 cd 'dir': changes to directory 'dir'.
2882
2882
2883 cd -: changes to the last visited directory.
2883 cd -: changes to the last visited directory.
2884
2884
2885 cd -<n>: changes to the n-th directory in the directory history.
2885 cd -<n>: changes to the n-th directory in the directory history.
2886
2886
2887 cd --foo: change to directory that matches 'foo' in history
2887 cd --foo: change to directory that matches 'foo' in history
2888
2888
2889 cd -b <bookmark_name>: jump to a bookmark set by %bookmark
2889 cd -b <bookmark_name>: jump to a bookmark set by %bookmark
2890 (note: cd <bookmark_name> is enough if there is no
2890 (note: cd <bookmark_name> is enough if there is no
2891 directory <bookmark_name>, but a bookmark with the name exists.)
2891 directory <bookmark_name>, but a bookmark with the name exists.)
2892 'cd -b <tab>' allows you to tab-complete bookmark names.
2892 'cd -b <tab>' allows you to tab-complete bookmark names.
2893
2893
2894 Options:
2894 Options:
2895
2895
2896 -q: quiet. Do not print the working directory after the cd command is
2896 -q: quiet. Do not print the working directory after the cd command is
2897 executed. By default IPython's cd command does print this directory,
2897 executed. By default IPython's cd command does print this directory,
2898 since the default prompts do not display path information.
2898 since the default prompts do not display path information.
2899
2899
2900 Note that !cd doesn't work for this purpose because the shell where
2900 Note that !cd doesn't work for this purpose because the shell where
2901 !command runs is immediately discarded after executing 'command'.
2901 !command runs is immediately discarded after executing 'command'.
2902
2902
2903 Examples
2903 Examples
2904 --------
2904 --------
2905 ::
2905 ::
2906
2906
2907 In [10]: cd parent/child
2907 In [10]: cd parent/child
2908 /home/tsuser/parent/child
2908 /home/tsuser/parent/child
2909 """
2909 """
2910
2910
2911 parameter_s = parameter_s.strip()
2911 parameter_s = parameter_s.strip()
2912 #bkms = self.shell.persist.get("bookmarks",{})
2912 #bkms = self.shell.persist.get("bookmarks",{})
2913
2913
2914 oldcwd = os.getcwdu()
2914 oldcwd = os.getcwdu()
2915 numcd = re.match(r'(-)(\d+)$',parameter_s)
2915 numcd = re.match(r'(-)(\d+)$',parameter_s)
2916 # jump in directory history by number
2916 # jump in directory history by number
2917 if numcd:
2917 if numcd:
2918 nn = int(numcd.group(2))
2918 nn = int(numcd.group(2))
2919 try:
2919 try:
2920 ps = self.shell.user_ns['_dh'][nn]
2920 ps = self.shell.user_ns['_dh'][nn]
2921 except IndexError:
2921 except IndexError:
2922 print 'The requested directory does not exist in history.'
2922 print 'The requested directory does not exist in history.'
2923 return
2923 return
2924 else:
2924 else:
2925 opts = {}
2925 opts = {}
2926 elif parameter_s.startswith('--'):
2926 elif parameter_s.startswith('--'):
2927 ps = None
2927 ps = None
2928 fallback = None
2928 fallback = None
2929 pat = parameter_s[2:]
2929 pat = parameter_s[2:]
2930 dh = self.shell.user_ns['_dh']
2930 dh = self.shell.user_ns['_dh']
2931 # first search only by basename (last component)
2931 # first search only by basename (last component)
2932 for ent in reversed(dh):
2932 for ent in reversed(dh):
2933 if pat in os.path.basename(ent) and os.path.isdir(ent):
2933 if pat in os.path.basename(ent) and os.path.isdir(ent):
2934 ps = ent
2934 ps = ent
2935 break
2935 break
2936
2936
2937 if fallback is None and pat in ent and os.path.isdir(ent):
2937 if fallback is None and pat in ent and os.path.isdir(ent):
2938 fallback = ent
2938 fallback = ent
2939
2939
2940 # if we have no last part match, pick the first full path match
2940 # if we have no last part match, pick the first full path match
2941 if ps is None:
2941 if ps is None:
2942 ps = fallback
2942 ps = fallback
2943
2943
2944 if ps is None:
2944 if ps is None:
2945 print "No matching entry in directory history"
2945 print "No matching entry in directory history"
2946 return
2946 return
2947 else:
2947 else:
2948 opts = {}
2948 opts = {}
2949
2949
2950
2950
2951 else:
2951 else:
2952 #turn all non-space-escaping backslashes to slashes,
2952 #turn all non-space-escaping backslashes to slashes,
2953 # for c:\windows\directory\names\
2953 # for c:\windows\directory\names\
2954 parameter_s = re.sub(r'\\(?! )','/', parameter_s)
2954 parameter_s = re.sub(r'\\(?! )','/', parameter_s)
2955 opts,ps = self.parse_options(parameter_s,'qb',mode='string')
2955 opts,ps = self.parse_options(parameter_s,'qb',mode='string')
2956 # jump to previous
2956 # jump to previous
2957 if ps == '-':
2957 if ps == '-':
2958 try:
2958 try:
2959 ps = self.shell.user_ns['_dh'][-2]
2959 ps = self.shell.user_ns['_dh'][-2]
2960 except IndexError:
2960 except IndexError:
2961 raise UsageError('%cd -: No previous directory to change to.')
2961 raise UsageError('%cd -: No previous directory to change to.')
2962 # jump to bookmark if needed
2962 # jump to bookmark if needed
2963 else:
2963 else:
2964 if not os.path.isdir(ps) or opts.has_key('b'):
2964 if not os.path.isdir(ps) or opts.has_key('b'):
2965 bkms = self.db.get('bookmarks', {})
2965 bkms = self.db.get('bookmarks', {})
2966
2966
2967 if bkms.has_key(ps):
2967 if bkms.has_key(ps):
2968 target = bkms[ps]
2968 target = bkms[ps]
2969 print '(bookmark:%s) -> %s' % (ps,target)
2969 print '(bookmark:%s) -> %s' % (ps,target)
2970 ps = target
2970 ps = target
2971 else:
2971 else:
2972 if opts.has_key('b'):
2972 if opts.has_key('b'):
2973 raise UsageError("Bookmark '%s' not found. "
2973 raise UsageError("Bookmark '%s' not found. "
2974 "Use '%%bookmark -l' to see your bookmarks." % ps)
2974 "Use '%%bookmark -l' to see your bookmarks." % ps)
2975
2975
2976 # strip extra quotes on Windows, because os.chdir doesn't like them
2976 # strip extra quotes on Windows, because os.chdir doesn't like them
2977 ps = unquote_filename(ps)
2977 ps = unquote_filename(ps)
2978 # at this point ps should point to the target dir
2978 # at this point ps should point to the target dir
2979 if ps:
2979 if ps:
2980 try:
2980 try:
2981 os.chdir(os.path.expanduser(ps))
2981 os.chdir(os.path.expanduser(ps))
2982 if hasattr(self.shell, 'term_title') and self.shell.term_title:
2982 if hasattr(self.shell, 'term_title') and self.shell.term_title:
2983 set_term_title('IPython: ' + abbrev_cwd())
2983 set_term_title('IPython: ' + abbrev_cwd())
2984 except OSError:
2984 except OSError:
2985 print sys.exc_info()[1]
2985 print sys.exc_info()[1]
2986 else:
2986 else:
2987 cwd = os.getcwdu()
2987 cwd = os.getcwdu()
2988 dhist = self.shell.user_ns['_dh']
2988 dhist = self.shell.user_ns['_dh']
2989 if oldcwd != cwd:
2989 if oldcwd != cwd:
2990 dhist.append(cwd)
2990 dhist.append(cwd)
2991 self.db['dhist'] = compress_dhist(dhist)[-100:]
2991 self.db['dhist'] = compress_dhist(dhist)[-100:]
2992
2992
2993 else:
2993 else:
2994 os.chdir(self.shell.home_dir)
2994 os.chdir(self.shell.home_dir)
2995 if hasattr(self.shell, 'term_title') and self.shell.term_title:
2995 if hasattr(self.shell, 'term_title') and self.shell.term_title:
2996 set_term_title('IPython: ' + '~')
2996 set_term_title('IPython: ' + '~')
2997 cwd = os.getcwdu()
2997 cwd = os.getcwdu()
2998 dhist = self.shell.user_ns['_dh']
2998 dhist = self.shell.user_ns['_dh']
2999
2999
3000 if oldcwd != cwd:
3000 if oldcwd != cwd:
3001 dhist.append(cwd)
3001 dhist.append(cwd)
3002 self.db['dhist'] = compress_dhist(dhist)[-100:]
3002 self.db['dhist'] = compress_dhist(dhist)[-100:]
3003 if not 'q' in opts and self.shell.user_ns['_dh']:
3003 if not 'q' in opts and self.shell.user_ns['_dh']:
3004 print self.shell.user_ns['_dh'][-1]
3004 print self.shell.user_ns['_dh'][-1]
3005
3005
3006
3006
3007 def magic_env(self, parameter_s=''):
3007 def magic_env(self, parameter_s=''):
3008 """List environment variables."""
3008 """List environment variables."""
3009
3009
3010 return dict(os.environ)
3010 return dict(os.environ)
3011
3011
3012 def magic_pushd(self, parameter_s=''):
3012 def magic_pushd(self, parameter_s=''):
3013 """Place the current dir on stack and change directory.
3013 """Place the current dir on stack and change directory.
3014
3014
3015 Usage:\\
3015 Usage:\\
3016 %pushd ['dirname']
3016 %pushd ['dirname']
3017 """
3017 """
3018
3018
3019 dir_s = self.shell.dir_stack
3019 dir_s = self.shell.dir_stack
3020 tgt = os.path.expanduser(unquote_filename(parameter_s))
3020 tgt = os.path.expanduser(unquote_filename(parameter_s))
3021 cwd = os.getcwdu().replace(self.home_dir,'~')
3021 cwd = os.getcwdu().replace(self.home_dir,'~')
3022 if tgt:
3022 if tgt:
3023 self.magic_cd(parameter_s)
3023 self.magic_cd(parameter_s)
3024 dir_s.insert(0,cwd)
3024 dir_s.insert(0,cwd)
3025 return self.magic_dirs()
3025 return self.magic_dirs()
3026
3026
3027 def magic_popd(self, parameter_s=''):
3027 def magic_popd(self, parameter_s=''):
3028 """Change to directory popped off the top of the stack.
3028 """Change to directory popped off the top of the stack.
3029 """
3029 """
3030 if not self.shell.dir_stack:
3030 if not self.shell.dir_stack:
3031 raise UsageError("%popd on empty stack")
3031 raise UsageError("%popd on empty stack")
3032 top = self.shell.dir_stack.pop(0)
3032 top = self.shell.dir_stack.pop(0)
3033 self.magic_cd(top)
3033 self.magic_cd(top)
3034 print "popd ->",top
3034 print "popd ->",top
3035
3035
3036 def magic_dirs(self, parameter_s=''):
3036 def magic_dirs(self, parameter_s=''):
3037 """Return the current directory stack."""
3037 """Return the current directory stack."""
3038
3038
3039 return self.shell.dir_stack
3039 return self.shell.dir_stack
3040
3040
3041 def magic_dhist(self, parameter_s=''):
3041 def magic_dhist(self, parameter_s=''):
3042 """Print your history of visited directories.
3042 """Print your history of visited directories.
3043
3043
3044 %dhist -> print full history\\
3044 %dhist -> print full history\\
3045 %dhist n -> print last n entries only\\
3045 %dhist n -> print last n entries only\\
3046 %dhist n1 n2 -> print entries between n1 and n2 (n1 not included)\\
3046 %dhist n1 n2 -> print entries between n1 and n2 (n1 not included)\\
3047
3047
3048 This history is automatically maintained by the %cd command, and
3048 This history is automatically maintained by the %cd command, and
3049 always available as the global list variable _dh. You can use %cd -<n>
3049 always available as the global list variable _dh. You can use %cd -<n>
3050 to go to directory number <n>.
3050 to go to directory number <n>.
3051
3051
3052 Note that most of time, you should view directory history by entering
3052 Note that most of time, you should view directory history by entering
3053 cd -<TAB>.
3053 cd -<TAB>.
3054
3054
3055 """
3055 """
3056
3056
3057 dh = self.shell.user_ns['_dh']
3057 dh = self.shell.user_ns['_dh']
3058 if parameter_s:
3058 if parameter_s:
3059 try:
3059 try:
3060 args = map(int,parameter_s.split())
3060 args = map(int,parameter_s.split())
3061 except:
3061 except:
3062 self.arg_err(Magic.magic_dhist)
3062 self.arg_err(Magic.magic_dhist)
3063 return
3063 return
3064 if len(args) == 1:
3064 if len(args) == 1:
3065 ini,fin = max(len(dh)-(args[0]),0),len(dh)
3065 ini,fin = max(len(dh)-(args[0]),0),len(dh)
3066 elif len(args) == 2:
3066 elif len(args) == 2:
3067 ini,fin = args
3067 ini,fin = args
3068 else:
3068 else:
3069 self.arg_err(Magic.magic_dhist)
3069 self.arg_err(Magic.magic_dhist)
3070 return
3070 return
3071 else:
3071 else:
3072 ini,fin = 0,len(dh)
3072 ini,fin = 0,len(dh)
3073 nlprint(dh,
3073 nlprint(dh,
3074 header = 'Directory history (kept in _dh)',
3074 header = 'Directory history (kept in _dh)',
3075 start=ini,stop=fin)
3075 start=ini,stop=fin)
3076
3076
3077 @skip_doctest
3077 @skip_doctest
3078 def magic_sc(self, parameter_s=''):
3078 def magic_sc(self, parameter_s=''):
3079 """Shell capture - execute a shell command and capture its output.
3079 """Shell capture - execute a shell command and capture its output.
3080
3080
3081 DEPRECATED. Suboptimal, retained for backwards compatibility.
3081 DEPRECATED. Suboptimal, retained for backwards compatibility.
3082
3082
3083 You should use the form 'var = !command' instead. Example:
3083 You should use the form 'var = !command' instead. Example:
3084
3084
3085 "%sc -l myfiles = ls ~" should now be written as
3085 "%sc -l myfiles = ls ~" should now be written as
3086
3086
3087 "myfiles = !ls ~"
3087 "myfiles = !ls ~"
3088
3088
3089 myfiles.s, myfiles.l and myfiles.n still apply as documented
3089 myfiles.s, myfiles.l and myfiles.n still apply as documented
3090 below.
3090 below.
3091
3091
3092 --
3092 --
3093 %sc [options] varname=command
3093 %sc [options] varname=command
3094
3094
3095 IPython will run the given command using commands.getoutput(), and
3095 IPython will run the given command using commands.getoutput(), and
3096 will then update the user's interactive namespace with a variable
3096 will then update the user's interactive namespace with a variable
3097 called varname, containing the value of the call. Your command can
3097 called varname, containing the value of the call. Your command can
3098 contain shell wildcards, pipes, etc.
3098 contain shell wildcards, pipes, etc.
3099
3099
3100 The '=' sign in the syntax is mandatory, and the variable name you
3100 The '=' sign in the syntax is mandatory, and the variable name you
3101 supply must follow Python's standard conventions for valid names.
3101 supply must follow Python's standard conventions for valid names.
3102
3102
3103 (A special format without variable name exists for internal use)
3103 (A special format without variable name exists for internal use)
3104
3104
3105 Options:
3105 Options:
3106
3106
3107 -l: list output. Split the output on newlines into a list before
3107 -l: list output. Split the output on newlines into a list before
3108 assigning it to the given variable. By default the output is stored
3108 assigning it to the given variable. By default the output is stored
3109 as a single string.
3109 as a single string.
3110
3110
3111 -v: verbose. Print the contents of the variable.
3111 -v: verbose. Print the contents of the variable.
3112
3112
3113 In most cases you should not need to split as a list, because the
3113 In most cases you should not need to split as a list, because the
3114 returned value is a special type of string which can automatically
3114 returned value is a special type of string which can automatically
3115 provide its contents either as a list (split on newlines) or as a
3115 provide its contents either as a list (split on newlines) or as a
3116 space-separated string. These are convenient, respectively, either
3116 space-separated string. These are convenient, respectively, either
3117 for sequential processing or to be passed to a shell command.
3117 for sequential processing or to be passed to a shell command.
3118
3118
3119 For example::
3119 For example::
3120
3120
3121 # Capture into variable a
3121 # Capture into variable a
3122 In [1]: sc a=ls *py
3122 In [1]: sc a=ls *py
3123
3123
3124 # a is a string with embedded newlines
3124 # a is a string with embedded newlines
3125 In [2]: a
3125 In [2]: a
3126 Out[2]: 'setup.py\\nwin32_manual_post_install.py'
3126 Out[2]: 'setup.py\\nwin32_manual_post_install.py'
3127
3127
3128 # which can be seen as a list:
3128 # which can be seen as a list:
3129 In [3]: a.l
3129 In [3]: a.l
3130 Out[3]: ['setup.py', 'win32_manual_post_install.py']
3130 Out[3]: ['setup.py', 'win32_manual_post_install.py']
3131
3131
3132 # or as a whitespace-separated string:
3132 # or as a whitespace-separated string:
3133 In [4]: a.s
3133 In [4]: a.s
3134 Out[4]: 'setup.py win32_manual_post_install.py'
3134 Out[4]: 'setup.py win32_manual_post_install.py'
3135
3135
3136 # a.s is useful to pass as a single command line:
3136 # a.s is useful to pass as a single command line:
3137 In [5]: !wc -l $a.s
3137 In [5]: !wc -l $a.s
3138 146 setup.py
3138 146 setup.py
3139 130 win32_manual_post_install.py
3139 130 win32_manual_post_install.py
3140 276 total
3140 276 total
3141
3141
3142 # while the list form is useful to loop over:
3142 # while the list form is useful to loop over:
3143 In [6]: for f in a.l:
3143 In [6]: for f in a.l:
3144 ...: !wc -l $f
3144 ...: !wc -l $f
3145 ...:
3145 ...:
3146 146 setup.py
3146 146 setup.py
3147 130 win32_manual_post_install.py
3147 130 win32_manual_post_install.py
3148
3148
3149 Similarly, the lists returned by the -l option are also special, in
3149 Similarly, the lists returned by the -l option are also special, in
3150 the sense that you can equally invoke the .s attribute on them to
3150 the sense that you can equally invoke the .s attribute on them to
3151 automatically get a whitespace-separated string from their contents::
3151 automatically get a whitespace-separated string from their contents::
3152
3152
3153 In [7]: sc -l b=ls *py
3153 In [7]: sc -l b=ls *py
3154
3154
3155 In [8]: b
3155 In [8]: b
3156 Out[8]: ['setup.py', 'win32_manual_post_install.py']
3156 Out[8]: ['setup.py', 'win32_manual_post_install.py']
3157
3157
3158 In [9]: b.s
3158 In [9]: b.s
3159 Out[9]: 'setup.py win32_manual_post_install.py'
3159 Out[9]: 'setup.py win32_manual_post_install.py'
3160
3160
3161 In summary, both the lists and strings used for output capture have
3161 In summary, both the lists and strings used for output capture have
3162 the following special attributes::
3162 the following special attributes::
3163
3163
3164 .l (or .list) : value as list.
3164 .l (or .list) : value as list.
3165 .n (or .nlstr): value as newline-separated string.
3165 .n (or .nlstr): value as newline-separated string.
3166 .s (or .spstr): value as space-separated string.
3166 .s (or .spstr): value as space-separated string.
3167 """
3167 """
3168
3168
3169 opts,args = self.parse_options(parameter_s,'lv')
3169 opts,args = self.parse_options(parameter_s,'lv')
3170 # Try to get a variable name and command to run
3170 # Try to get a variable name and command to run
3171 try:
3171 try:
3172 # the variable name must be obtained from the parse_options
3172 # the variable name must be obtained from the parse_options
3173 # output, which uses shlex.split to strip options out.
3173 # output, which uses shlex.split to strip options out.
3174 var,_ = args.split('=',1)
3174 var,_ = args.split('=',1)
3175 var = var.strip()
3175 var = var.strip()
3176 # But the command has to be extracted from the original input
3176 # But the command has to be extracted from the original input
3177 # parameter_s, not on what parse_options returns, to avoid the
3177 # parameter_s, not on what parse_options returns, to avoid the
3178 # quote stripping which shlex.split performs on it.
3178 # quote stripping which shlex.split performs on it.
3179 _,cmd = parameter_s.split('=',1)
3179 _,cmd = parameter_s.split('=',1)
3180 except ValueError:
3180 except ValueError:
3181 var,cmd = '',''
3181 var,cmd = '',''
3182 # If all looks ok, proceed
3182 # If all looks ok, proceed
3183 split = 'l' in opts
3183 split = 'l' in opts
3184 out = self.shell.getoutput(cmd, split=split)
3184 out = self.shell.getoutput(cmd, split=split)
3185 if opts.has_key('v'):
3185 if opts.has_key('v'):
3186 print '%s ==\n%s' % (var,pformat(out))
3186 print '%s ==\n%s' % (var,pformat(out))
3187 if var:
3187 if var:
3188 self.shell.user_ns.update({var:out})
3188 self.shell.user_ns.update({var:out})
3189 else:
3189 else:
3190 return out
3190 return out
3191
3191
3192 def magic_sx(self, parameter_s=''):
3192 def magic_sx(self, parameter_s=''):
3193 """Shell execute - run a shell command and capture its output.
3193 """Shell execute - run a shell command and capture its output.
3194
3194
3195 %sx command
3195 %sx command
3196
3196
3197 IPython will run the given command using commands.getoutput(), and
3197 IPython will run the given command using commands.getoutput(), and
3198 return the result formatted as a list (split on '\\n'). Since the
3198 return the result formatted as a list (split on '\\n'). Since the
3199 output is _returned_, it will be stored in ipython's regular output
3199 output is _returned_, it will be stored in ipython's regular output
3200 cache Out[N] and in the '_N' automatic variables.
3200 cache Out[N] and in the '_N' automatic variables.
3201
3201
3202 Notes:
3202 Notes:
3203
3203
3204 1) If an input line begins with '!!', then %sx is automatically
3204 1) If an input line begins with '!!', then %sx is automatically
3205 invoked. That is, while::
3205 invoked. That is, while::
3206
3206
3207 !ls
3207 !ls
3208
3208
3209 causes ipython to simply issue system('ls'), typing::
3209 causes ipython to simply issue system('ls'), typing::
3210
3210
3211 !!ls
3211 !!ls
3212
3212
3213 is a shorthand equivalent to::
3213 is a shorthand equivalent to::
3214
3214
3215 %sx ls
3215 %sx ls
3216
3216
3217 2) %sx differs from %sc in that %sx automatically splits into a list,
3217 2) %sx differs from %sc in that %sx automatically splits into a list,
3218 like '%sc -l'. The reason for this is to make it as easy as possible
3218 like '%sc -l'. The reason for this is to make it as easy as possible
3219 to process line-oriented shell output via further python commands.
3219 to process line-oriented shell output via further python commands.
3220 %sc is meant to provide much finer control, but requires more
3220 %sc is meant to provide much finer control, but requires more
3221 typing.
3221 typing.
3222
3222
3223 3) Just like %sc -l, this is a list with special attributes:
3223 3) Just like %sc -l, this is a list with special attributes:
3224 ::
3224 ::
3225
3225
3226 .l (or .list) : value as list.
3226 .l (or .list) : value as list.
3227 .n (or .nlstr): value as newline-separated string.
3227 .n (or .nlstr): value as newline-separated string.
3228 .s (or .spstr): value as whitespace-separated string.
3228 .s (or .spstr): value as whitespace-separated string.
3229
3229
3230 This is very useful when trying to use such lists as arguments to
3230 This is very useful when trying to use such lists as arguments to
3231 system commands."""
3231 system commands."""
3232
3232
3233 if parameter_s:
3233 if parameter_s:
3234 return self.shell.getoutput(parameter_s)
3234 return self.shell.getoutput(parameter_s)
3235
3235
3236
3236
3237 def magic_bookmark(self, parameter_s=''):
3237 def magic_bookmark(self, parameter_s=''):
3238 """Manage IPython's bookmark system.
3238 """Manage IPython's bookmark system.
3239
3239
3240 %bookmark <name> - set bookmark to current dir
3240 %bookmark <name> - set bookmark to current dir
3241 %bookmark <name> <dir> - set bookmark to <dir>
3241 %bookmark <name> <dir> - set bookmark to <dir>
3242 %bookmark -l - list all bookmarks
3242 %bookmark -l - list all bookmarks
3243 %bookmark -d <name> - remove bookmark
3243 %bookmark -d <name> - remove bookmark
3244 %bookmark -r - remove all bookmarks
3244 %bookmark -r - remove all bookmarks
3245
3245
3246 You can later on access a bookmarked folder with::
3246 You can later on access a bookmarked folder with::
3247
3247
3248 %cd -b <name>
3248 %cd -b <name>
3249
3249
3250 or simply '%cd <name>' if there is no directory called <name> AND
3250 or simply '%cd <name>' if there is no directory called <name> AND
3251 there is such a bookmark defined.
3251 there is such a bookmark defined.
3252
3252
3253 Your bookmarks persist through IPython sessions, but they are
3253 Your bookmarks persist through IPython sessions, but they are
3254 associated with each profile."""
3254 associated with each profile."""
3255
3255
3256 opts,args = self.parse_options(parameter_s,'drl',mode='list')
3256 opts,args = self.parse_options(parameter_s,'drl',mode='list')
3257 if len(args) > 2:
3257 if len(args) > 2:
3258 raise UsageError("%bookmark: too many arguments")
3258 raise UsageError("%bookmark: too many arguments")
3259
3259
3260 bkms = self.db.get('bookmarks',{})
3260 bkms = self.db.get('bookmarks',{})
3261
3261
3262 if opts.has_key('d'):
3262 if opts.has_key('d'):
3263 try:
3263 try:
3264 todel = args[0]
3264 todel = args[0]
3265 except IndexError:
3265 except IndexError:
3266 raise UsageError(
3266 raise UsageError(
3267 "%bookmark -d: must provide a bookmark to delete")
3267 "%bookmark -d: must provide a bookmark to delete")
3268 else:
3268 else:
3269 try:
3269 try:
3270 del bkms[todel]
3270 del bkms[todel]
3271 except KeyError:
3271 except KeyError:
3272 raise UsageError(
3272 raise UsageError(
3273 "%%bookmark -d: Can't delete bookmark '%s'" % todel)
3273 "%%bookmark -d: Can't delete bookmark '%s'" % todel)
3274
3274
3275 elif opts.has_key('r'):
3275 elif opts.has_key('r'):
3276 bkms = {}
3276 bkms = {}
3277 elif opts.has_key('l'):
3277 elif opts.has_key('l'):
3278 bks = bkms.keys()
3278 bks = bkms.keys()
3279 bks.sort()
3279 bks.sort()
3280 if bks:
3280 if bks:
3281 size = max(map(len,bks))
3281 size = max(map(len,bks))
3282 else:
3282 else:
3283 size = 0
3283 size = 0
3284 fmt = '%-'+str(size)+'s -> %s'
3284 fmt = '%-'+str(size)+'s -> %s'
3285 print 'Current bookmarks:'
3285 print 'Current bookmarks:'
3286 for bk in bks:
3286 for bk in bks:
3287 print fmt % (bk,bkms[bk])
3287 print fmt % (bk,bkms[bk])
3288 else:
3288 else:
3289 if not args:
3289 if not args:
3290 raise UsageError("%bookmark: You must specify the bookmark name")
3290 raise UsageError("%bookmark: You must specify the bookmark name")
3291 elif len(args)==1:
3291 elif len(args)==1:
3292 bkms[args[0]] = os.getcwdu()
3292 bkms[args[0]] = os.getcwdu()
3293 elif len(args)==2:
3293 elif len(args)==2:
3294 bkms[args[0]] = args[1]
3294 bkms[args[0]] = args[1]
3295 self.db['bookmarks'] = bkms
3295 self.db['bookmarks'] = bkms
3296
3296
3297 def magic_pycat(self, parameter_s=''):
3297 def magic_pycat(self, parameter_s=''):
3298 """Show a syntax-highlighted file through a pager.
3298 """Show a syntax-highlighted file through a pager.
3299
3299
3300 This magic is similar to the cat utility, but it will assume the file
3300 This magic is similar to the cat utility, but it will assume the file
3301 to be Python source and will show it with syntax highlighting. """
3301 to be Python source and will show it with syntax highlighting. """
3302
3302
3303 try:
3303 try:
3304 filename = get_py_filename(parameter_s)
3304 filename = get_py_filename(parameter_s)
3305 cont = file_read(filename)
3305 cont = file_read(filename)
3306 except IOError:
3306 except IOError:
3307 try:
3307 try:
3308 cont = eval(parameter_s,self.user_ns)
3308 cont = eval(parameter_s,self.user_ns)
3309 except NameError:
3309 except NameError:
3310 cont = None
3310 cont = None
3311 if cont is None:
3311 if cont is None:
3312 print "Error: no such file or variable"
3312 print "Error: no such file or variable"
3313 return
3313 return
3314
3314
3315 page.page(self.shell.pycolorize(cont))
3315 page.page(self.shell.pycolorize(cont))
3316
3316
3317 def magic_quickref(self,arg):
3317 def magic_quickref(self,arg):
3318 """ Show a quick reference sheet """
3318 """ Show a quick reference sheet """
3319 import IPython.core.usage
3319 import IPython.core.usage
3320 qr = IPython.core.usage.quick_reference + self.magic_magic('-brief')
3320 qr = IPython.core.usage.quick_reference + self.magic_magic('-brief')
3321
3321
3322 page.page(qr)
3322 page.page(qr)
3323
3323
3324 def magic_doctest_mode(self,parameter_s=''):
3324 def magic_doctest_mode(self,parameter_s=''):
3325 """Toggle doctest mode on and off.
3325 """Toggle doctest mode on and off.
3326
3326
3327 This mode is intended to make IPython behave as much as possible like a
3327 This mode is intended to make IPython behave as much as possible like a
3328 plain Python shell, from the perspective of how its prompts, exceptions
3328 plain Python shell, from the perspective of how its prompts, exceptions
3329 and output look. This makes it easy to copy and paste parts of a
3329 and output look. This makes it easy to copy and paste parts of a
3330 session into doctests. It does so by:
3330 session into doctests. It does so by:
3331
3331
3332 - Changing the prompts to the classic ``>>>`` ones.
3332 - Changing the prompts to the classic ``>>>`` ones.
3333 - Changing the exception reporting mode to 'Plain'.
3333 - Changing the exception reporting mode to 'Plain'.
3334 - Disabling pretty-printing of output.
3334 - Disabling pretty-printing of output.
3335
3335
3336 Note that IPython also supports the pasting of code snippets that have
3336 Note that IPython also supports the pasting of code snippets that have
3337 leading '>>>' and '...' prompts in them. This means that you can paste
3337 leading '>>>' and '...' prompts in them. This means that you can paste
3338 doctests from files or docstrings (even if they have leading
3338 doctests from files or docstrings (even if they have leading
3339 whitespace), and the code will execute correctly. You can then use
3339 whitespace), and the code will execute correctly. You can then use
3340 '%history -t' to see the translated history; this will give you the
3340 '%history -t' to see the translated history; this will give you the
3341 input after removal of all the leading prompts and whitespace, which
3341 input after removal of all the leading prompts and whitespace, which
3342 can be pasted back into an editor.
3342 can be pasted back into an editor.
3343
3343
3344 With these features, you can switch into this mode easily whenever you
3344 With these features, you can switch into this mode easily whenever you
3345 need to do testing and changes to doctests, without having to leave
3345 need to do testing and changes to doctests, without having to leave
3346 your existing IPython session.
3346 your existing IPython session.
3347 """
3347 """
3348
3348
3349 from IPython.utils.ipstruct import Struct
3349 from IPython.utils.ipstruct import Struct
3350
3350
3351 # Shorthands
3351 # Shorthands
3352 shell = self.shell
3352 shell = self.shell
3353 pm = shell.prompt_manager
3353 pm = shell.prompt_manager
3354 meta = shell.meta
3354 meta = shell.meta
3355 disp_formatter = self.shell.display_formatter
3355 disp_formatter = self.shell.display_formatter
3356 ptformatter = disp_formatter.formatters['text/plain']
3356 ptformatter = disp_formatter.formatters['text/plain']
3357 # dstore is a data store kept in the instance metadata bag to track any
3357 # dstore is a data store kept in the instance metadata bag to track any
3358 # changes we make, so we can undo them later.
3358 # changes we make, so we can undo them later.
3359 dstore = meta.setdefault('doctest_mode',Struct())
3359 dstore = meta.setdefault('doctest_mode',Struct())
3360 save_dstore = dstore.setdefault
3360 save_dstore = dstore.setdefault
3361
3361
3362 # save a few values we'll need to recover later
3362 # save a few values we'll need to recover later
3363 mode = save_dstore('mode',False)
3363 mode = save_dstore('mode',False)
3364 save_dstore('rc_pprint',ptformatter.pprint)
3364 save_dstore('rc_pprint',ptformatter.pprint)
3365 save_dstore('xmode',shell.InteractiveTB.mode)
3365 save_dstore('xmode',shell.InteractiveTB.mode)
3366 save_dstore('rc_separate_out',shell.separate_out)
3366 save_dstore('rc_separate_out',shell.separate_out)
3367 save_dstore('rc_separate_out2',shell.separate_out2)
3367 save_dstore('rc_separate_out2',shell.separate_out2)
3368 save_dstore('rc_prompts_pad_left',pm.justify)
3368 save_dstore('rc_prompts_pad_left',pm.justify)
3369 save_dstore('rc_separate_in',shell.separate_in)
3369 save_dstore('rc_separate_in',shell.separate_in)
3370 save_dstore('rc_plain_text_only',disp_formatter.plain_text_only)
3370 save_dstore('rc_plain_text_only',disp_formatter.plain_text_only)
3371 save_dstore('prompt_templates',(pm.in_template, pm.in2_template, pm.out_template))
3371 save_dstore('prompt_templates',(pm.in_template, pm.in2_template, pm.out_template))
3372
3372
3373 if mode == False:
3373 if mode == False:
3374 # turn on
3374 # turn on
3375 pm.in_template = '>>> '
3375 pm.in_template = '>>> '
3376 pm.in2_template = '... '
3376 pm.in2_template = '... '
3377 pm.out_template = ''
3377 pm.out_template = ''
3378
3378
3379 # Prompt separators like plain python
3379 # Prompt separators like plain python
3380 shell.separate_in = ''
3380 shell.separate_in = ''
3381 shell.separate_out = ''
3381 shell.separate_out = ''
3382 shell.separate_out2 = ''
3382 shell.separate_out2 = ''
3383
3383
3384 pm.justify = False
3384 pm.justify = False
3385
3385
3386 ptformatter.pprint = False
3386 ptformatter.pprint = False
3387 disp_formatter.plain_text_only = True
3387 disp_formatter.plain_text_only = True
3388
3388
3389 shell.magic_xmode('Plain')
3389 shell.magic_xmode('Plain')
3390 else:
3390 else:
3391 # turn off
3391 # turn off
3392 pm.in_template, pm.in2_template, pm.out_template = dstore.prompt_templates
3392 pm.in_template, pm.in2_template, pm.out_template = dstore.prompt_templates
3393
3393
3394 shell.separate_in = dstore.rc_separate_in
3394 shell.separate_in = dstore.rc_separate_in
3395
3395
3396 shell.separate_out = dstore.rc_separate_out
3396 shell.separate_out = dstore.rc_separate_out
3397 shell.separate_out2 = dstore.rc_separate_out2
3397 shell.separate_out2 = dstore.rc_separate_out2
3398
3398
3399 pm.justify = dstore.rc_prompts_pad_left
3399 pm.justify = dstore.rc_prompts_pad_left
3400
3400
3401 ptformatter.pprint = dstore.rc_pprint
3401 ptformatter.pprint = dstore.rc_pprint
3402 disp_formatter.plain_text_only = dstore.rc_plain_text_only
3402 disp_formatter.plain_text_only = dstore.rc_plain_text_only
3403
3403
3404 shell.magic_xmode(dstore.xmode)
3404 shell.magic_xmode(dstore.xmode)
3405
3405
3406 # Store new mode and inform
3406 # Store new mode and inform
3407 dstore.mode = bool(1-int(mode))
3407 dstore.mode = bool(1-int(mode))
3408 mode_label = ['OFF','ON'][dstore.mode]
3408 mode_label = ['OFF','ON'][dstore.mode]
3409 print 'Doctest mode is:', mode_label
3409 print 'Doctest mode is:', mode_label
3410
3410
3411 def magic_gui(self, parameter_s=''):
3411 def magic_gui(self, parameter_s=''):
3412 """Enable or disable IPython GUI event loop integration.
3412 """Enable or disable IPython GUI event loop integration.
3413
3413
3414 %gui [GUINAME]
3414 %gui [GUINAME]
3415
3415
3416 This magic replaces IPython's threaded shells that were activated
3416 This magic replaces IPython's threaded shells that were activated
3417 using the (pylab/wthread/etc.) command line flags. GUI toolkits
3417 using the (pylab/wthread/etc.) command line flags. GUI toolkits
3418 can now be enabled at runtime and keyboard
3418 can now be enabled at runtime and keyboard
3419 interrupts should work without any problems. The following toolkits
3419 interrupts should work without any problems. The following toolkits
3420 are supported: wxPython, PyQt4, PyGTK, Tk and Cocoa (OSX)::
3420 are supported: wxPython, PyQt4, PyGTK, Tk and Cocoa (OSX)::
3421
3421
3422 %gui wx # enable wxPython event loop integration
3422 %gui wx # enable wxPython event loop integration
3423 %gui qt4|qt # enable PyQt4 event loop integration
3423 %gui qt4|qt # enable PyQt4 event loop integration
3424 %gui gtk # enable PyGTK event loop integration
3424 %gui gtk # enable PyGTK event loop integration
3425 %gui gtk3 # enable Gtk3 event loop integration
3425 %gui gtk3 # enable Gtk3 event loop integration
3426 %gui tk # enable Tk event loop integration
3426 %gui tk # enable Tk event loop integration
3427 %gui OSX # enable Cocoa event loop integration
3427 %gui OSX # enable Cocoa event loop integration
3428 # (requires %matplotlib 1.1)
3428 # (requires %matplotlib 1.1)
3429 %gui # disable all event loop integration
3429 %gui # disable all event loop integration
3430
3430
3431 WARNING: after any of these has been called you can simply create
3431 WARNING: after any of these has been called you can simply create
3432 an application object, but DO NOT start the event loop yourself, as
3432 an application object, but DO NOT start the event loop yourself, as
3433 we have already handled that.
3433 we have already handled that.
3434 """
3434 """
3435 opts, arg = self.parse_options(parameter_s, '')
3435 opts, arg = self.parse_options(parameter_s, '')
3436 if arg=='': arg = None
3436 if arg=='': arg = None
3437 try:
3437 try:
3438 return self.enable_gui(arg)
3438 return self.enable_gui(arg)
3439 except Exception as e:
3439 except Exception as e:
3440 # print simple error message, rather than traceback if we can't
3440 # print simple error message, rather than traceback if we can't
3441 # hook up the GUI
3441 # hook up the GUI
3442 error(str(e))
3442 error(str(e))
3443
3443
3444 def magic_install_ext(self, parameter_s):
3444 def magic_install_ext(self, parameter_s):
3445 """Download and install an extension from a URL, e.g.::
3445 """Download and install an extension from a URL, e.g.::
3446
3446
3447 %install_ext https://bitbucket.org/birkenfeld/ipython-physics/raw/d1310a2ab15d/physics.py
3447 %install_ext https://bitbucket.org/birkenfeld/ipython-physics/raw/d1310a2ab15d/physics.py
3448
3448
3449 The URL should point to an importable Python module - either a .py file
3449 The URL should point to an importable Python module - either a .py file
3450 or a .zip file.
3450 or a .zip file.
3451
3451
3452 Parameters:
3452 Parameters:
3453
3453
3454 -n filename : Specify a name for the file, rather than taking it from
3454 -n filename : Specify a name for the file, rather than taking it from
3455 the URL.
3455 the URL.
3456 """
3456 """
3457 opts, args = self.parse_options(parameter_s, 'n:')
3457 opts, args = self.parse_options(parameter_s, 'n:')
3458 try:
3458 try:
3459 filename = self.extension_manager.install_extension(args, opts.get('n'))
3459 filename = self.extension_manager.install_extension(args, opts.get('n'))
3460 except ValueError as e:
3460 except ValueError as e:
3461 print e
3461 print e
3462 return
3462 return
3463
3463
3464 filename = os.path.basename(filename)
3464 filename = os.path.basename(filename)
3465 print "Installed %s. To use it, type:" % filename
3465 print "Installed %s. To use it, type:" % filename
3466 print " %%load_ext %s" % os.path.splitext(filename)[0]
3466 print " %%load_ext %s" % os.path.splitext(filename)[0]
3467
3467
3468
3468
3469 def magic_load_ext(self, module_str):
3469 def magic_load_ext(self, module_str):
3470 """Load an IPython extension by its module name."""
3470 """Load an IPython extension by its module name."""
3471 return self.extension_manager.load_extension(module_str)
3471 return self.extension_manager.load_extension(module_str)
3472
3472
3473 def magic_unload_ext(self, module_str):
3473 def magic_unload_ext(self, module_str):
3474 """Unload an IPython extension by its module name."""
3474 """Unload an IPython extension by its module name."""
3475 self.extension_manager.unload_extension(module_str)
3475 self.extension_manager.unload_extension(module_str)
3476
3476
3477 def magic_reload_ext(self, module_str):
3477 def magic_reload_ext(self, module_str):
3478 """Reload an IPython extension by its module name."""
3478 """Reload an IPython extension by its module name."""
3479 self.extension_manager.reload_extension(module_str)
3479 self.extension_manager.reload_extension(module_str)
3480
3480
3481 def magic_install_profiles(self, s):
3481 def magic_install_profiles(self, s):
3482 """%install_profiles has been deprecated."""
3482 """%install_profiles has been deprecated."""
3483 print '\n'.join([
3483 print '\n'.join([
3484 "%install_profiles has been deprecated.",
3484 "%install_profiles has been deprecated.",
3485 "Use `ipython profile list` to view available profiles.",
3485 "Use `ipython profile list` to view available profiles.",
3486 "Requesting a profile with `ipython profile create <name>`",
3486 "Requesting a profile with `ipython profile create <name>`",
3487 "or `ipython --profile=<name>` will start with the bundled",
3487 "or `ipython --profile=<name>` will start with the bundled",
3488 "profile of that name if it exists."
3488 "profile of that name if it exists."
3489 ])
3489 ])
3490
3490
3491 def magic_install_default_config(self, s):
3491 def magic_install_default_config(self, s):
3492 """%install_default_config has been deprecated."""
3492 """%install_default_config has been deprecated."""
3493 print '\n'.join([
3493 print '\n'.join([
3494 "%install_default_config has been deprecated.",
3494 "%install_default_config has been deprecated.",
3495 "Use `ipython profile create <name>` to initialize a profile",
3495 "Use `ipython profile create <name>` to initialize a profile",
3496 "with the default config files.",
3496 "with the default config files.",
3497 "Add `--reset` to overwrite already existing config files with defaults."
3497 "Add `--reset` to overwrite already existing config files with defaults."
3498 ])
3498 ])
3499
3499
3500 # Pylab support: simple wrappers that activate pylab, load gui input
3500 # Pylab support: simple wrappers that activate pylab, load gui input
3501 # handling and modify slightly %run
3501 # handling and modify slightly %run
3502
3502
3503 @skip_doctest
3503 @skip_doctest
3504 def _pylab_magic_run(self, parameter_s=''):
3504 def _pylab_magic_run(self, parameter_s=''):
3505 Magic.magic_run(self, parameter_s,
3505 Magic.magic_run(self, parameter_s,
3506 runner=mpl_runner(self.shell.safe_execfile))
3506 runner=mpl_runner(self.shell.safe_execfile))
3507
3507
3508 _pylab_magic_run.__doc__ = magic_run.__doc__
3508 _pylab_magic_run.__doc__ = magic_run.__doc__
3509
3509
3510 @skip_doctest
3510 @skip_doctest
3511 def magic_pylab(self, s):
3511 def magic_pylab(self, s):
3512 """Load numpy and matplotlib to work interactively.
3512 """Load numpy and matplotlib to work interactively.
3513
3513
3514 %pylab [GUINAME]
3514 %pylab [GUINAME]
3515
3515
3516 This function lets you activate pylab (matplotlib, numpy and
3516 This function lets you activate pylab (matplotlib, numpy and
3517 interactive support) at any point during an IPython session.
3517 interactive support) at any point during an IPython session.
3518
3518
3519 It will import at the top level numpy as np, pyplot as plt, matplotlib,
3519 It will import at the top level numpy as np, pyplot as plt, matplotlib,
3520 pylab and mlab, as well as all names from numpy and pylab.
3520 pylab and mlab, as well as all names from numpy and pylab.
3521
3521
3522 If you are using the inline matplotlib backend for embedded figures,
3522 If you are using the inline matplotlib backend for embedded figures,
3523 you can adjust its behavior via the %config magic::
3523 you can adjust its behavior via the %config magic::
3524
3524
3525 # enable SVG figures, necessary for SVG+XHTML export in the qtconsole
3525 # enable SVG figures, necessary for SVG+XHTML export in the qtconsole
3526 In [1]: %config InlineBackend.figure_format = 'svg'
3526 In [1]: %config InlineBackend.figure_format = 'svg'
3527
3527
3528 # change the behavior of closing all figures at the end of each
3528 # change the behavior of closing all figures at the end of each
3529 # execution (cell), or allowing reuse of active figures across
3529 # execution (cell), or allowing reuse of active figures across
3530 # cells:
3530 # cells:
3531 In [2]: %config InlineBackend.close_figures = False
3531 In [2]: %config InlineBackend.close_figures = False
3532
3532
3533 Parameters
3533 Parameters
3534 ----------
3534 ----------
3535 guiname : optional
3535 guiname : optional
3536 One of the valid arguments to the %gui magic ('qt', 'wx', 'gtk',
3536 One of the valid arguments to the %gui magic ('qt', 'wx', 'gtk',
3537 'osx' or 'tk'). If given, the corresponding Matplotlib backend is
3537 'osx' or 'tk'). If given, the corresponding Matplotlib backend is
3538 used, otherwise matplotlib's default (which you can override in your
3538 used, otherwise matplotlib's default (which you can override in your
3539 matplotlib config file) is used.
3539 matplotlib config file) is used.
3540
3540
3541 Examples
3541 Examples
3542 --------
3542 --------
3543 In this case, where the MPL default is TkAgg::
3543 In this case, where the MPL default is TkAgg::
3544
3544
3545 In [2]: %pylab
3545 In [2]: %pylab
3546
3546
3547 Welcome to pylab, a matplotlib-based Python environment.
3547 Welcome to pylab, a matplotlib-based Python environment.
3548 Backend in use: TkAgg
3548 Backend in use: TkAgg
3549 For more information, type 'help(pylab)'.
3549 For more information, type 'help(pylab)'.
3550
3550
3551 But you can explicitly request a different backend::
3551 But you can explicitly request a different backend::
3552
3552
3553 In [3]: %pylab qt
3553 In [3]: %pylab qt
3554
3554
3555 Welcome to pylab, a matplotlib-based Python environment.
3555 Welcome to pylab, a matplotlib-based Python environment.
3556 Backend in use: Qt4Agg
3556 Backend in use: Qt4Agg
3557 For more information, type 'help(pylab)'.
3557 For more information, type 'help(pylab)'.
3558 """
3558 """
3559
3559
3560 if Application.initialized():
3560 if Application.initialized():
3561 app = Application.instance()
3561 app = Application.instance()
3562 try:
3562 try:
3563 import_all_status = app.pylab_import_all
3563 import_all_status = app.pylab_import_all
3564 except AttributeError:
3564 except AttributeError:
3565 import_all_status = True
3565 import_all_status = True
3566 else:
3566 else:
3567 import_all_status = True
3567 import_all_status = True
3568
3568
3569 self.shell.enable_pylab(s, import_all=import_all_status)
3569 self.shell.enable_pylab(s, import_all=import_all_status)
3570
3570
3571 def magic_tb(self, s):
3571 def magic_tb(self, s):
3572 """Print the last traceback with the currently active exception mode.
3572 """Print the last traceback with the currently active exception mode.
3573
3573
3574 See %xmode for changing exception reporting modes."""
3574 See %xmode for changing exception reporting modes."""
3575 self.shell.showtraceback()
3575 self.shell.showtraceback()
3576
3576
3577 @skip_doctest
3577 @skip_doctest
3578 def magic_precision(self, s=''):
3578 def magic_precision(self, s=''):
3579 """Set floating point precision for pretty printing.
3579 """Set floating point precision for pretty printing.
3580
3580
3581 Can set either integer precision or a format string.
3581 Can set either integer precision or a format string.
3582
3582
3583 If numpy has been imported and precision is an int,
3583 If numpy has been imported and precision is an int,
3584 numpy display precision will also be set, via ``numpy.set_printoptions``.
3584 numpy display precision will also be set, via ``numpy.set_printoptions``.
3585
3585
3586 If no argument is given, defaults will be restored.
3586 If no argument is given, defaults will be restored.
3587
3587
3588 Examples
3588 Examples
3589 --------
3589 --------
3590 ::
3590 ::
3591
3591
3592 In [1]: from math import pi
3592 In [1]: from math import pi
3593
3593
3594 In [2]: %precision 3
3594 In [2]: %precision 3
3595 Out[2]: u'%.3f'
3595 Out[2]: u'%.3f'
3596
3596
3597 In [3]: pi
3597 In [3]: pi
3598 Out[3]: 3.142
3598 Out[3]: 3.142
3599
3599
3600 In [4]: %precision %i
3600 In [4]: %precision %i
3601 Out[4]: u'%i'
3601 Out[4]: u'%i'
3602
3602
3603 In [5]: pi
3603 In [5]: pi
3604 Out[5]: 3
3604 Out[5]: 3
3605
3605
3606 In [6]: %precision %e
3606 In [6]: %precision %e
3607 Out[6]: u'%e'
3607 Out[6]: u'%e'
3608
3608
3609 In [7]: pi**10
3609 In [7]: pi**10
3610 Out[7]: 9.364805e+04
3610 Out[7]: 9.364805e+04
3611
3611
3612 In [8]: %precision
3612 In [8]: %precision
3613 Out[8]: u'%r'
3613 Out[8]: u'%r'
3614
3614
3615 In [9]: pi**10
3615 In [9]: pi**10
3616 Out[9]: 93648.047476082982
3616 Out[9]: 93648.047476082982
3617
3617
3618 """
3618 """
3619
3619
3620 ptformatter = self.shell.display_formatter.formatters['text/plain']
3620 ptformatter = self.shell.display_formatter.formatters['text/plain']
3621 ptformatter.float_precision = s
3621 ptformatter.float_precision = s
3622 return ptformatter.float_format
3622 return ptformatter.float_format
3623
3623
3624
3624
3625 @magic_arguments.magic_arguments()
3625 @magic_arguments.magic_arguments()
3626 @magic_arguments.argument(
3626 @magic_arguments.argument(
3627 '-e', '--export', action='store_true', default=False,
3627 '-e', '--export', action='store_true', default=False,
3628 help='Export IPython history as a notebook. The filename argument '
3628 help='Export IPython history as a notebook. The filename argument '
3629 'is used to specify the notebook name and format. For example '
3629 'is used to specify the notebook name and format. For example '
3630 'a filename of notebook.ipynb will result in a notebook name '
3630 'a filename of notebook.ipynb will result in a notebook name '
3631 'of "notebook" and a format of "xml". Likewise using a ".json" '
3631 'of "notebook" and a format of "xml". Likewise using a ".json" '
3632 'or ".py" file extension will write the notebook in the json '
3632 'or ".py" file extension will write the notebook in the json '
3633 'or py formats.'
3633 'or py formats.'
3634 )
3634 )
3635 @magic_arguments.argument(
3635 @magic_arguments.argument(
3636 '-f', '--format',
3636 '-f', '--format',
3637 help='Convert an existing IPython notebook to a new format. This option '
3637 help='Convert an existing IPython notebook to a new format. This option '
3638 'specifies the new format and can have the values: xml, json, py. '
3638 'specifies the new format and can have the values: xml, json, py. '
3639 'The target filename is chosen automatically based on the new '
3639 'The target filename is chosen automatically based on the new '
3640 'format. The filename argument gives the name of the source file.'
3640 'format. The filename argument gives the name of the source file.'
3641 )
3641 )
3642 @magic_arguments.argument(
3642 @magic_arguments.argument(
3643 'filename', type=unicode,
3643 'filename', type=unicode,
3644 help='Notebook name or filename'
3644 help='Notebook name or filename'
3645 )
3645 )
3646 def magic_notebook(self, s):
3646 def magic_notebook(self, s):
3647 """Export and convert IPython notebooks.
3647 """Export and convert IPython notebooks.
3648
3648
3649 This function can export the current IPython history to a notebook file
3649 This function can export the current IPython history to a notebook file
3650 or can convert an existing notebook file into a different format. For
3650 or can convert an existing notebook file into a different format. For
3651 example, to export the history to "foo.ipynb" do "%notebook -e foo.ipynb".
3651 example, to export the history to "foo.ipynb" do "%notebook -e foo.ipynb".
3652 To export the history to "foo.py" do "%notebook -e foo.py". To convert
3652 To export the history to "foo.py" do "%notebook -e foo.py". To convert
3653 "foo.ipynb" to "foo.json" do "%notebook -f json foo.ipynb". Possible
3653 "foo.ipynb" to "foo.json" do "%notebook -f json foo.ipynb". Possible
3654 formats include (json/ipynb, py).
3654 formats include (json/ipynb, py).
3655 """
3655 """
3656 args = magic_arguments.parse_argstring(self.magic_notebook, s)
3656 args = magic_arguments.parse_argstring(self.magic_notebook, s)
3657
3657
3658 from IPython.nbformat import current
3658 from IPython.nbformat import current
3659 args.filename = unquote_filename(args.filename)
3659 args.filename = unquote_filename(args.filename)
3660 if args.export:
3660 if args.export:
3661 fname, name, format = current.parse_filename(args.filename)
3661 fname, name, format = current.parse_filename(args.filename)
3662 cells = []
3662 cells = []
3663 hist = list(self.history_manager.get_range())
3663 hist = list(self.history_manager.get_range())
3664 for session, prompt_number, input in hist[:-1]:
3664 for session, prompt_number, input in hist[:-1]:
3665 cells.append(current.new_code_cell(prompt_number=prompt_number, input=input))
3665 cells.append(current.new_code_cell(prompt_number=prompt_number, input=input))
3666 worksheet = current.new_worksheet(cells=cells)
3666 worksheet = current.new_worksheet(cells=cells)
3667 nb = current.new_notebook(name=name,worksheets=[worksheet])
3667 nb = current.new_notebook(name=name,worksheets=[worksheet])
3668 with io.open(fname, 'w', encoding='utf-8') as f:
3668 with io.open(fname, 'w', encoding='utf-8') as f:
3669 current.write(nb, f, format);
3669 current.write(nb, f, format);
3670 elif args.format is not None:
3670 elif args.format is not None:
3671 old_fname, old_name, old_format = current.parse_filename(args.filename)
3671 old_fname, old_name, old_format = current.parse_filename(args.filename)
3672 new_format = args.format
3672 new_format = args.format
3673 if new_format == u'xml':
3673 if new_format == u'xml':
3674 raise ValueError('Notebooks cannot be written as xml.')
3674 raise ValueError('Notebooks cannot be written as xml.')
3675 elif new_format == u'ipynb' or new_format == u'json':
3675 elif new_format == u'ipynb' or new_format == u'json':
3676 new_fname = old_name + u'.ipynb'
3676 new_fname = old_name + u'.ipynb'
3677 new_format = u'json'
3677 new_format = u'json'
3678 elif new_format == u'py':
3678 elif new_format == u'py':
3679 new_fname = old_name + u'.py'
3679 new_fname = old_name + u'.py'
3680 else:
3680 else:
3681 raise ValueError('Invalid notebook format: %s' % new_format)
3681 raise ValueError('Invalid notebook format: %s' % new_format)
3682 with io.open(old_fname, 'r', encoding='utf-8') as f:
3682 with io.open(old_fname, 'r', encoding='utf-8') as f:
3683 nb = current.read(f, old_format)
3683 nb = current.read(f, old_format)
3684 with io.open(new_fname, 'w', encoding='utf-8') as f:
3684 with io.open(new_fname, 'w', encoding='utf-8') as f:
3685 current.write(nb, f, new_format)
3685 current.write(nb, f, new_format)
3686
3686
3687 def magic_config(self, s):
3687 def magic_config(self, s):
3688 """configure IPython
3688 """configure IPython
3689
3689
3690 %config Class[.trait=value]
3690 %config Class[.trait=value]
3691
3691
3692 This magic exposes most of the IPython config system. Any
3692 This magic exposes most of the IPython config system. Any
3693 Configurable class should be able to be configured with the simple
3693 Configurable class should be able to be configured with the simple
3694 line::
3694 line::
3695
3695
3696 %config Class.trait=value
3696 %config Class.trait=value
3697
3697
3698 Where `value` will be resolved in the user's namespace, if it is an
3698 Where `value` will be resolved in the user's namespace, if it is an
3699 expression or variable name.
3699 expression or variable name.
3700
3700
3701 Examples
3701 Examples
3702 --------
3702 --------
3703
3703
3704 To see what classes are available for config, pass no arguments::
3704 To see what classes are available for config, pass no arguments::
3705
3705
3706 In [1]: %config
3706 In [1]: %config
3707 Available objects for config:
3707 Available objects for config:
3708 TerminalInteractiveShell
3708 TerminalInteractiveShell
3709 HistoryManager
3709 HistoryManager
3710 PrefilterManager
3710 PrefilterManager
3711 AliasManager
3711 AliasManager
3712 IPCompleter
3712 IPCompleter
3713 PromptManager
3713 PromptManager
3714 DisplayFormatter
3714 DisplayFormatter
3715
3715
3716 To view what is configurable on a given class, just pass the class
3716 To view what is configurable on a given class, just pass the class
3717 name::
3717 name::
3718
3718
3719 In [2]: %config IPCompleter
3719 In [2]: %config IPCompleter
3720 IPCompleter options
3720 IPCompleter options
3721 -----------------
3721 -----------------
3722 IPCompleter.omit__names=<Enum>
3722 IPCompleter.omit__names=<Enum>
3723 Current: 2
3723 Current: 2
3724 Choices: (0, 1, 2)
3724 Choices: (0, 1, 2)
3725 Instruct the completer to omit private method names
3725 Instruct the completer to omit private method names
3726 Specifically, when completing on ``object.<tab>``.
3726 Specifically, when completing on ``object.<tab>``.
3727 When 2 [default]: all names that start with '_' will be excluded.
3727 When 2 [default]: all names that start with '_' will be excluded.
3728 When 1: all 'magic' names (``__foo__``) will be excluded.
3728 When 1: all 'magic' names (``__foo__``) will be excluded.
3729 When 0: nothing will be excluded.
3729 When 0: nothing will be excluded.
3730 IPCompleter.merge_completions=<CBool>
3730 IPCompleter.merge_completions=<CBool>
3731 Current: True
3731 Current: True
3732 Whether to merge completion results into a single list
3732 Whether to merge completion results into a single list
3733 If False, only the completion results from the first non-empty completer
3733 If False, only the completion results from the first non-empty completer
3734 will be returned.
3734 will be returned.
3735 IPCompleter.limit_to__all__=<CBool>
3735 IPCompleter.limit_to__all__=<CBool>
3736 Current: False
3736 Current: False
3737 Instruct the completer to use __all__ for the completion
3737 Instruct the completer to use __all__ for the completion
3738 Specifically, when completing on ``object.<tab>``.
3738 Specifically, when completing on ``object.<tab>``.
3739 When True: only those names in obj.__all__ will be included.
3739 When True: only those names in obj.__all__ will be included.
3740 When False [default]: the __all__ attribute is ignored
3740 When False [default]: the __all__ attribute is ignored
3741 IPCompleter.greedy=<CBool>
3741 IPCompleter.greedy=<CBool>
3742 Current: False
3742 Current: False
3743 Activate greedy completion
3743 Activate greedy completion
3744 This will enable completion on elements of lists, results of function calls,
3744 This will enable completion on elements of lists, results of function calls,
3745 etc., but can be unsafe because the code is actually evaluated on TAB.
3745 etc., but can be unsafe because the code is actually evaluated on TAB.
3746
3746
3747 but the real use is in setting values::
3747 but the real use is in setting values::
3748
3748
3749 In [3]: %config IPCompleter.greedy = True
3749 In [3]: %config IPCompleter.greedy = True
3750
3750
3751 and these values are read from the user_ns if they are variables::
3751 and these values are read from the user_ns if they are variables::
3752
3752
3753 In [4]: feeling_greedy=False
3753 In [4]: feeling_greedy=False
3754
3754
3755 In [5]: %config IPCompleter.greedy = feeling_greedy
3755 In [5]: %config IPCompleter.greedy = feeling_greedy
3756
3756
3757 """
3757 """
3758 from IPython.config.loader import Config
3758 from IPython.config.loader import Config
3759 # some IPython objects are Configurable, but do not yet have
3759 # some IPython objects are Configurable, but do not yet have
3760 # any configurable traits. Exclude them from the effects of
3760 # any configurable traits. Exclude them from the effects of
3761 # this magic, as their presence is just noise:
3761 # this magic, as their presence is just noise:
3762 configurables = [ c for c in self.configurables if c.__class__.class_traits(config=True) ]
3762 configurables = [ c for c in self.configurables if c.__class__.class_traits(config=True) ]
3763 classnames = [ c.__class__.__name__ for c in configurables ]
3763 classnames = [ c.__class__.__name__ for c in configurables ]
3764
3764
3765 line = s.strip()
3765 line = s.strip()
3766 if not line:
3766 if not line:
3767 # print available configurable names
3767 # print available configurable names
3768 print "Available objects for config:"
3768 print "Available objects for config:"
3769 for name in classnames:
3769 for name in classnames:
3770 print " ", name
3770 print " ", name
3771 return
3771 return
3772 elif line in classnames:
3772 elif line in classnames:
3773 # `%config TerminalInteractiveShell` will print trait info for
3773 # `%config TerminalInteractiveShell` will print trait info for
3774 # TerminalInteractiveShell
3774 # TerminalInteractiveShell
3775 c = configurables[classnames.index(line)]
3775 c = configurables[classnames.index(line)]
3776 cls = c.__class__
3776 cls = c.__class__
3777 help = cls.class_get_help(c)
3777 help = cls.class_get_help(c)
3778 # strip leading '--' from cl-args:
3778 # strip leading '--' from cl-args:
3779 help = re.sub(re.compile(r'^--', re.MULTILINE), '', help)
3779 help = re.sub(re.compile(r'^--', re.MULTILINE), '', help)
3780 print help
3780 print help
3781 return
3781 return
3782 elif '=' not in line:
3782 elif '=' not in line:
3783 raise UsageError("Invalid config statement: %r, should be Class.trait = value" % line)
3783 raise UsageError("Invalid config statement: %r, should be Class.trait = value" % line)
3784
3784
3785
3785
3786 # otherwise, assume we are setting configurables.
3786 # otherwise, assume we are setting configurables.
3787 # leave quotes on args when splitting, because we want
3787 # leave quotes on args when splitting, because we want
3788 # unquoted args to eval in user_ns
3788 # unquoted args to eval in user_ns
3789 cfg = Config()
3789 cfg = Config()
3790 exec "cfg."+line in locals(), self.user_ns
3790 exec "cfg."+line in locals(), self.user_ns
3791
3791
3792 for configurable in configurables:
3792 for configurable in configurables:
3793 try:
3793 try:
3794 configurable.update_config(cfg)
3794 configurable.update_config(cfg)
3795 except Exception as e:
3795 except Exception as e:
3796 error(e)
3796 error(e)
3797
3797
3798 # end Magic
3798 # end Magic
@@ -1,75 +1,75 b''
1 # coding: utf-8
1 # coding: utf-8
2 """Tests for the compilerop module.
2 """Tests for the compilerop module.
3 """
3 """
4 #-----------------------------------------------------------------------------
4 #-----------------------------------------------------------------------------
5 # Copyright (C) 2010-2011 The IPython Development Team.
5 # Copyright (C) 2010-2011 The IPython Development Team.
6 #
6 #
7 # Distributed under the terms of the BSD License.
7 # Distributed under the terms of the BSD License.
8 #
8 #
9 # The full license is in the file COPYING.txt, distributed with this software.
9 # The full license is in the file COPYING.txt, distributed with this software.
10 #-----------------------------------------------------------------------------
10 #-----------------------------------------------------------------------------
11
11
12 #-----------------------------------------------------------------------------
12 #-----------------------------------------------------------------------------
13 # Imports
13 # Imports
14 #-----------------------------------------------------------------------------
14 #-----------------------------------------------------------------------------
15 from __future__ import print_function
15 from __future__ import print_function
16
16
17 # Stdlib imports
17 # Stdlib imports
18 import linecache
18 import linecache
19 import sys
19 import sys
20
20
21 # Third-party imports
21 # Third-party imports
22 import nose.tools as nt
22 import nose.tools as nt
23
23
24 # Our own imports
24 # Our own imports
25 from IPython.core import compilerop
25 from IPython.core import compilerop
26 from IPython.utils import py3compat
26 from IPython.utils import py3compat
27
27
28 #-----------------------------------------------------------------------------
28 #-----------------------------------------------------------------------------
29 # Test functions
29 # Test functions
30 #-----------------------------------------------------------------------------
30 #-----------------------------------------------------------------------------
31
31
32 def test_code_name():
32 def test_code_name():
33 code = 'x=1'
33 code = 'x=1'
34 name = compilerop.code_name(code)
34 name = compilerop.code_name(code)
35 nt.assert_true(name.startswith('<ipython-input-0'))
35 nt.assert_true(name.startswith('<ipython-input-0'))
36
36
37
37
38 def test_code_name2():
38 def test_code_name2():
39 code = 'x=1'
39 code = 'x=1'
40 name = compilerop.code_name(code, 9)
40 name = compilerop.code_name(code, 9)
41 nt.assert_true(name.startswith('<ipython-input-9'))
41 nt.assert_true(name.startswith('<ipython-input-9'))
42
42
43
43
44 def test_cache():
44 def test_cache():
45 """Test the compiler correctly compiles and caches inputs
45 """Test the compiler correctly compiles and caches inputs
46 """
46 """
47 cp = compilerop.CachingCompiler()
47 cp = compilerop.CachingCompiler()
48 ncache = len(linecache.cache)
48 ncache = len(linecache.cache)
49 cp.cache('x=1')
49 cp.cache('x=1')
50 nt.assert_true(len(linecache.cache) > ncache)
50 nt.assert_true(len(linecache.cache) > ncache)
51
51
52 def setUp():
52 def setUp():
53 # Check we're in a proper Python 2 environment (some imports, such
53 # Check we're in a proper Python 2 environment (some imports, such
54 # as GTK, can change the default encoding, which can hide bugs.)
54 # as GTK, can change the default encoding, which can hide bugs.)
55 nt.assert_equal(sys.getdefaultencoding(), "utf-8" if py3compat.PY3 else "ascii")
55 nt.assert_equal(py3compat.getdefaultencoding(), "utf-8" if py3compat.PY3 else "ascii")
56
56
57 def test_cache_unicode():
57 def test_cache_unicode():
58 cp = compilerop.CachingCompiler()
58 cp = compilerop.CachingCompiler()
59 ncache = len(linecache.cache)
59 ncache = len(linecache.cache)
60 cp.cache(u"t = 'žćčőđ'")
60 cp.cache(u"t = 'žćčőđ'")
61 nt.assert_true(len(linecache.cache) > ncache)
61 nt.assert_true(len(linecache.cache) > ncache)
62
62
63 def test_compiler_check_cache():
63 def test_compiler_check_cache():
64 """Test the compiler properly manages the cache.
64 """Test the compiler properly manages the cache.
65 """
65 """
66 # Rather simple-minded tests that just exercise the API
66 # Rather simple-minded tests that just exercise the API
67 cp = compilerop.CachingCompiler()
67 cp = compilerop.CachingCompiler()
68 cp.cache('x=1', 99)
68 cp.cache('x=1', 99)
69 # Ensure now that after clearing the cache, our entries survive
69 # Ensure now that after clearing the cache, our entries survive
70 cp.check_cache()
70 cp.check_cache()
71 for k in linecache.cache:
71 for k in linecache.cache:
72 if k.startswith('<ipython-input-99'):
72 if k.startswith('<ipython-input-99'):
73 break
73 break
74 else:
74 else:
75 raise AssertionError('Entry for input-99 missing from linecache')
75 raise AssertionError('Entry for input-99 missing from linecache')
@@ -1,151 +1,151 b''
1 # coding: utf-8
1 # coding: utf-8
2 """Tests for the IPython tab-completion machinery.
2 """Tests for the IPython tab-completion machinery.
3 """
3 """
4 #-----------------------------------------------------------------------------
4 #-----------------------------------------------------------------------------
5 # Module imports
5 # Module imports
6 #-----------------------------------------------------------------------------
6 #-----------------------------------------------------------------------------
7
7
8 # stdlib
8 # stdlib
9 import os
9 import os
10 import shutil
10 import shutil
11 import sys
11 import sys
12 import tempfile
12 import tempfile
13 import unittest
13 import unittest
14 from datetime import datetime
14 from datetime import datetime
15
15
16 # third party
16 # third party
17 import nose.tools as nt
17 import nose.tools as nt
18
18
19 # our own packages
19 # our own packages
20 from IPython.config.loader import Config
20 from IPython.config.loader import Config
21 from IPython.utils.tempdir import TemporaryDirectory
21 from IPython.utils.tempdir import TemporaryDirectory
22 from IPython.core.history import HistoryManager, extract_hist_ranges
22 from IPython.core.history import HistoryManager, extract_hist_ranges
23 from IPython.utils import py3compat
23 from IPython.utils import py3compat
24
24
25 def setUp():
25 def setUp():
26 nt.assert_equal(sys.getdefaultencoding(), "utf-8" if py3compat.PY3 else "ascii")
26 nt.assert_equal(py3compat.getdefaultencoding(), "utf-8" if py3compat.PY3 else "ascii")
27
27
28 def test_history():
28 def test_history():
29 ip = get_ipython()
29 ip = get_ipython()
30 with TemporaryDirectory() as tmpdir:
30 with TemporaryDirectory() as tmpdir:
31 hist_manager_ori = ip.history_manager
31 hist_manager_ori = ip.history_manager
32 hist_file = os.path.join(tmpdir, 'history.sqlite')
32 hist_file = os.path.join(tmpdir, 'history.sqlite')
33 try:
33 try:
34 ip.history_manager = HistoryManager(shell=ip, hist_file=hist_file)
34 ip.history_manager = HistoryManager(shell=ip, hist_file=hist_file)
35 hist = [u'a=1', u'def f():\n test = 1\n return test', u"b='β‚¬Γ†ΒΎΓ·ΓŸ'"]
35 hist = [u'a=1', u'def f():\n test = 1\n return test', u"b='β‚¬Γ†ΒΎΓ·ΓŸ'"]
36 for i, h in enumerate(hist, start=1):
36 for i, h in enumerate(hist, start=1):
37 ip.history_manager.store_inputs(i, h)
37 ip.history_manager.store_inputs(i, h)
38
38
39 ip.history_manager.db_log_output = True
39 ip.history_manager.db_log_output = True
40 # Doesn't match the input, but we'll just check it's stored.
40 # Doesn't match the input, but we'll just check it's stored.
41 ip.history_manager.output_hist_reprs[3] = "spam"
41 ip.history_manager.output_hist_reprs[3] = "spam"
42 ip.history_manager.store_output(3)
42 ip.history_manager.store_output(3)
43
43
44 nt.assert_equal(ip.history_manager.input_hist_raw, [''] + hist)
44 nt.assert_equal(ip.history_manager.input_hist_raw, [''] + hist)
45
45
46 # Detailed tests for _get_range_session
46 # Detailed tests for _get_range_session
47 grs = ip.history_manager._get_range_session
47 grs = ip.history_manager._get_range_session
48 nt.assert_equal(list(grs(start=2,stop=-1)), zip([0], [2], hist[1:-1]))
48 nt.assert_equal(list(grs(start=2,stop=-1)), zip([0], [2], hist[1:-1]))
49 nt.assert_equal(list(grs(start=-2)), zip([0,0], [2,3], hist[-2:]))
49 nt.assert_equal(list(grs(start=-2)), zip([0,0], [2,3], hist[-2:]))
50 nt.assert_equal(list(grs(output=True)), zip([0,0,0], [1,2,3], zip(hist, [None,None,'spam'])))
50 nt.assert_equal(list(grs(output=True)), zip([0,0,0], [1,2,3], zip(hist, [None,None,'spam'])))
51
51
52 # Check whether specifying a range beyond the end of the current
52 # Check whether specifying a range beyond the end of the current
53 # session results in an error (gh-804)
53 # session results in an error (gh-804)
54 ip.magic('%hist 2-500')
54 ip.magic('%hist 2-500')
55
55
56 # Check that we can write non-ascii characters to a file
56 # Check that we can write non-ascii characters to a file
57 ip.magic("%%hist -f %s" % os.path.join(tmpdir, "test1"))
57 ip.magic("%%hist -f %s" % os.path.join(tmpdir, "test1"))
58 ip.magic("%%hist -pf %s" % os.path.join(tmpdir, "test2"))
58 ip.magic("%%hist -pf %s" % os.path.join(tmpdir, "test2"))
59 ip.magic("%%hist -nf %s" % os.path.join(tmpdir, "test3"))
59 ip.magic("%%hist -nf %s" % os.path.join(tmpdir, "test3"))
60 ip.magic("%%save %s 1-10" % os.path.join(tmpdir, "test4"))
60 ip.magic("%%save %s 1-10" % os.path.join(tmpdir, "test4"))
61
61
62 # New session
62 # New session
63 ip.history_manager.reset()
63 ip.history_manager.reset()
64 newcmds = ["z=5","class X(object):\n pass", "k='p'"]
64 newcmds = ["z=5","class X(object):\n pass", "k='p'"]
65 for i, cmd in enumerate(newcmds, start=1):
65 for i, cmd in enumerate(newcmds, start=1):
66 ip.history_manager.store_inputs(i, cmd)
66 ip.history_manager.store_inputs(i, cmd)
67 gothist = ip.history_manager.get_range(start=1, stop=4)
67 gothist = ip.history_manager.get_range(start=1, stop=4)
68 nt.assert_equal(list(gothist), zip([0,0,0],[1,2,3], newcmds))
68 nt.assert_equal(list(gothist), zip([0,0,0],[1,2,3], newcmds))
69 # Previous session:
69 # Previous session:
70 gothist = ip.history_manager.get_range(-1, 1, 4)
70 gothist = ip.history_manager.get_range(-1, 1, 4)
71 nt.assert_equal(list(gothist), zip([1,1,1],[1,2,3], hist))
71 nt.assert_equal(list(gothist), zip([1,1,1],[1,2,3], hist))
72
72
73 # Check get_hist_tail
73 # Check get_hist_tail
74 gothist = ip.history_manager.get_tail(4, output=True,
74 gothist = ip.history_manager.get_tail(4, output=True,
75 include_latest=True)
75 include_latest=True)
76 expected = [(1, 3, (hist[-1], "spam")),
76 expected = [(1, 3, (hist[-1], "spam")),
77 (2, 1, (newcmds[0], None)),
77 (2, 1, (newcmds[0], None)),
78 (2, 2, (newcmds[1], None)),
78 (2, 2, (newcmds[1], None)),
79 (2, 3, (newcmds[2], None)),]
79 (2, 3, (newcmds[2], None)),]
80 nt.assert_equal(list(gothist), expected)
80 nt.assert_equal(list(gothist), expected)
81
81
82 gothist = ip.history_manager.get_tail(2)
82 gothist = ip.history_manager.get_tail(2)
83 expected = [(2, 1, newcmds[0]),
83 expected = [(2, 1, newcmds[0]),
84 (2, 2, newcmds[1])]
84 (2, 2, newcmds[1])]
85 nt.assert_equal(list(gothist), expected)
85 nt.assert_equal(list(gothist), expected)
86
86
87 # Check get_hist_search
87 # Check get_hist_search
88 gothist = ip.history_manager.search("*test*")
88 gothist = ip.history_manager.search("*test*")
89 nt.assert_equal(list(gothist), [(1,2,hist[1])] )
89 nt.assert_equal(list(gothist), [(1,2,hist[1])] )
90 gothist = ip.history_manager.search("b*", output=True)
90 gothist = ip.history_manager.search("b*", output=True)
91 nt.assert_equal(list(gothist), [(1,3,(hist[2],"spam"))] )
91 nt.assert_equal(list(gothist), [(1,3,(hist[2],"spam"))] )
92
92
93 # Cross testing: check that magic %save can get previous session.
93 # Cross testing: check that magic %save can get previous session.
94 testfilename = os.path.realpath(os.path.join(tmpdir, "test.py"))
94 testfilename = os.path.realpath(os.path.join(tmpdir, "test.py"))
95 ip.magic_save(testfilename + " ~1/1-3")
95 ip.magic_save(testfilename + " ~1/1-3")
96 with py3compat.open(testfilename) as testfile:
96 with py3compat.open(testfilename) as testfile:
97 nt.assert_equal(testfile.read(),
97 nt.assert_equal(testfile.read(),
98 u"# coding: utf-8\n" + u"\n".join(hist))
98 u"# coding: utf-8\n" + u"\n".join(hist))
99
99
100 # Duplicate line numbers - check that it doesn't crash, and
100 # Duplicate line numbers - check that it doesn't crash, and
101 # gets a new session
101 # gets a new session
102 ip.history_manager.store_inputs(1, "rogue")
102 ip.history_manager.store_inputs(1, "rogue")
103 ip.history_manager.writeout_cache()
103 ip.history_manager.writeout_cache()
104 nt.assert_equal(ip.history_manager.session_number, 3)
104 nt.assert_equal(ip.history_manager.session_number, 3)
105 finally:
105 finally:
106 # Restore history manager
106 # Restore history manager
107 ip.history_manager = hist_manager_ori
107 ip.history_manager = hist_manager_ori
108
108
109
109
110 def test_extract_hist_ranges():
110 def test_extract_hist_ranges():
111 instr = "1 2/3 ~4/5-6 ~4/7-~4/9 ~9/2-~7/5"
111 instr = "1 2/3 ~4/5-6 ~4/7-~4/9 ~9/2-~7/5"
112 expected = [(0, 1, 2), # 0 == current session
112 expected = [(0, 1, 2), # 0 == current session
113 (2, 3, 4),
113 (2, 3, 4),
114 (-4, 5, 7),
114 (-4, 5, 7),
115 (-4, 7, 10),
115 (-4, 7, 10),
116 (-9, 2, None), # None == to end
116 (-9, 2, None), # None == to end
117 (-8, 1, None),
117 (-8, 1, None),
118 (-7, 1, 6)]
118 (-7, 1, 6)]
119 actual = list(extract_hist_ranges(instr))
119 actual = list(extract_hist_ranges(instr))
120 nt.assert_equal(actual, expected)
120 nt.assert_equal(actual, expected)
121
121
122 def test_magic_rerun():
122 def test_magic_rerun():
123 """Simple test for %rerun (no args -> rerun last line)"""
123 """Simple test for %rerun (no args -> rerun last line)"""
124 ip = get_ipython()
124 ip = get_ipython()
125 ip.run_cell("a = 10", store_history=True)
125 ip.run_cell("a = 10", store_history=True)
126 ip.run_cell("a += 1", store_history=True)
126 ip.run_cell("a += 1", store_history=True)
127 nt.assert_equal(ip.user_ns["a"], 11)
127 nt.assert_equal(ip.user_ns["a"], 11)
128 ip.run_cell("%rerun", store_history=True)
128 ip.run_cell("%rerun", store_history=True)
129 nt.assert_equal(ip.user_ns["a"], 12)
129 nt.assert_equal(ip.user_ns["a"], 12)
130
130
131 def test_timestamp_type():
131 def test_timestamp_type():
132 ip = get_ipython()
132 ip = get_ipython()
133 info = ip.history_manager.get_session_info()
133 info = ip.history_manager.get_session_info()
134 nt.assert_true(isinstance(info[1], datetime))
134 nt.assert_true(isinstance(info[1], datetime))
135
135
136 def test_hist_file_config():
136 def test_hist_file_config():
137 cfg = Config()
137 cfg = Config()
138 tfile = tempfile.NamedTemporaryFile(delete=False)
138 tfile = tempfile.NamedTemporaryFile(delete=False)
139 cfg.HistoryManager.hist_file = tfile.name
139 cfg.HistoryManager.hist_file = tfile.name
140 try:
140 try:
141 hm = HistoryManager(shell=get_ipython(), config=cfg)
141 hm = HistoryManager(shell=get_ipython(), config=cfg)
142 nt.assert_equals(hm.hist_file, cfg.HistoryManager.hist_file)
142 nt.assert_equals(hm.hist_file, cfg.HistoryManager.hist_file)
143 finally:
143 finally:
144 try:
144 try:
145 os.remove(tfile.name)
145 os.remove(tfile.name)
146 except OSError:
146 except OSError:
147 # same catch as in testing.tools.TempFileMixin
147 # same catch as in testing.tools.TempFileMixin
148 # On Windows, even though we close the file, we still can't
148 # On Windows, even though we close the file, we still can't
149 # delete it. I have no clue why
149 # delete it. I have no clue why
150 pass
150 pass
151
151
@@ -1,3708 +1,3708 b''
1 # -*- coding: utf-8 -*-
1 # -*- coding: utf-8 -*-
2 # module pyparsing.py
2 # module pyparsing.py
3 #
3 #
4 # Copyright (c) 2003-2009 Paul T. McGuire
4 # Copyright (c) 2003-2009 Paul T. McGuire
5 #
5 #
6 # Permission is hereby granted, free of charge, to any person obtaining
6 # Permission is hereby granted, free of charge, to any person obtaining
7 # a copy of this software and associated documentation files (the
7 # a copy of this software and associated documentation files (the
8 # "Software"), to deal in the Software without restriction, including
8 # "Software"), to deal in the Software without restriction, including
9 # without limitation the rights to use, copy, modify, merge, publish,
9 # without limitation the rights to use, copy, modify, merge, publish,
10 # distribute, sublicense, and/or sell copies of the Software, and to
10 # distribute, sublicense, and/or sell copies of the Software, and to
11 # permit persons to whom the Software is furnished to do so, subject to
11 # permit persons to whom the Software is furnished to do so, subject to
12 # the following conditions:
12 # the following conditions:
13 #
13 #
14 # The above copyright notice and this permission notice shall be
14 # The above copyright notice and this permission notice shall be
15 # included in all copies or substantial portions of the Software.
15 # included in all copies or substantial portions of the Software.
16 #
16 #
17 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
17 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
18 # EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
18 # EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
19 # MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
19 # MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
20 # IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
20 # IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
21 # CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
21 # CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
22 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
22 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
23 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
23 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
24 #
24 #
25 #from __future__ import generators
25 #from __future__ import generators
26
26
27 __doc__ = \
27 __doc__ = \
28 """
28 """
29 pyparsing module - Classes and methods to define and execute parsing grammars
29 pyparsing module - Classes and methods to define and execute parsing grammars
30
30
31 The pyparsing module is an alternative approach to creating and executing simple grammars,
31 The pyparsing module is an alternative approach to creating and executing simple grammars,
32 vs. the traditional lex/yacc approach, or the use of regular expressions. With pyparsing, you
32 vs. the traditional lex/yacc approach, or the use of regular expressions. With pyparsing, you
33 don't need to learn a new syntax for defining grammars or matching expressions - the parsing module
33 don't need to learn a new syntax for defining grammars or matching expressions - the parsing module
34 provides a library of classes that you use to construct the grammar directly in Python.
34 provides a library of classes that you use to construct the grammar directly in Python.
35
35
36 Here is a program to parse "Hello, World!" (or any greeting of the form "<salutation>, <addressee>!")::
36 Here is a program to parse "Hello, World!" (or any greeting of the form "<salutation>, <addressee>!")::
37
37
38 from pyparsing import Word, alphas
38 from pyparsing import Word, alphas
39
39
40 # define grammar of a greeting
40 # define grammar of a greeting
41 greet = Word( alphas ) + "," + Word( alphas ) + "!"
41 greet = Word( alphas ) + "," + Word( alphas ) + "!"
42
42
43 hello = "Hello, World!"
43 hello = "Hello, World!"
44 print hello, "->", greet.parseString( hello )
44 print hello, "->", greet.parseString( hello )
45
45
46 The program outputs the following::
46 The program outputs the following::
47
47
48 Hello, World! -> ['Hello', ',', 'World', '!']
48 Hello, World! -> ['Hello', ',', 'World', '!']
49
49
50 The Python representation of the grammar is quite readable, owing to the self-explanatory
50 The Python representation of the grammar is quite readable, owing to the self-explanatory
51 class names, and the use of '+', '|' and '^' operators.
51 class names, and the use of '+', '|' and '^' operators.
52
52
53 The parsed results returned from parseString() can be accessed as a nested list, a dictionary, or an
53 The parsed results returned from parseString() can be accessed as a nested list, a dictionary, or an
54 object with named attributes.
54 object with named attributes.
55
55
56 The pyparsing module handles some of the problems that are typically vexing when writing text parsers:
56 The pyparsing module handles some of the problems that are typically vexing when writing text parsers:
57 - extra or missing whitespace (the above program will also handle "Hello,World!", "Hello , World !", etc.)
57 - extra or missing whitespace (the above program will also handle "Hello,World!", "Hello , World !", etc.)
58 - quoted strings
58 - quoted strings
59 - embedded comments
59 - embedded comments
60 """
60 """
61
61
62 __version__ = "1.5.2"
62 __version__ = "1.5.2"
63 __versionTime__ = "17 February 2009 19:45"
63 __versionTime__ = "17 February 2009 19:45"
64 __author__ = "Paul McGuire <ptmcg@users.sourceforge.net>"
64 __author__ = "Paul McGuire <ptmcg@users.sourceforge.net>"
65
65
66 import string
66 import string
67 from weakref import ref as wkref
67 from weakref import ref as wkref
68 import copy
68 import copy
69 import sys
69 import sys
70 import warnings
70 import warnings
71 import re
71 import re
72 import sre_constants
72 import sre_constants
73 #~ sys.stderr.write( "testing pyparsing module, version %s, %s\n" % (__version__,__versionTime__ ) )
73 #~ sys.stderr.write( "testing pyparsing module, version %s, %s\n" % (__version__,__versionTime__ ) )
74
74
75 __all__ = [
75 __all__ = [
76 'And', 'CaselessKeyword', 'CaselessLiteral', 'CharsNotIn', 'Combine', 'Dict', 'Each', 'Empty',
76 'And', 'CaselessKeyword', 'CaselessLiteral', 'CharsNotIn', 'Combine', 'Dict', 'Each', 'Empty',
77 'FollowedBy', 'Forward', 'GoToColumn', 'Group', 'Keyword', 'LineEnd', 'LineStart', 'Literal',
77 'FollowedBy', 'Forward', 'GoToColumn', 'Group', 'Keyword', 'LineEnd', 'LineStart', 'Literal',
78 'MatchFirst', 'NoMatch', 'NotAny', 'OneOrMore', 'OnlyOnce', 'Optional', 'Or',
78 'MatchFirst', 'NoMatch', 'NotAny', 'OneOrMore', 'OnlyOnce', 'Optional', 'Or',
79 'ParseBaseException', 'ParseElementEnhance', 'ParseException', 'ParseExpression', 'ParseFatalException',
79 'ParseBaseException', 'ParseElementEnhance', 'ParseException', 'ParseExpression', 'ParseFatalException',
80 'ParseResults', 'ParseSyntaxException', 'ParserElement', 'QuotedString', 'RecursiveGrammarException',
80 'ParseResults', 'ParseSyntaxException', 'ParserElement', 'QuotedString', 'RecursiveGrammarException',
81 'Regex', 'SkipTo', 'StringEnd', 'StringStart', 'Suppress', 'Token', 'TokenConverter', 'Upcase',
81 'Regex', 'SkipTo', 'StringEnd', 'StringStart', 'Suppress', 'Token', 'TokenConverter', 'Upcase',
82 'White', 'Word', 'WordEnd', 'WordStart', 'ZeroOrMore',
82 'White', 'Word', 'WordEnd', 'WordStart', 'ZeroOrMore',
83 'alphanums', 'alphas', 'alphas8bit', 'anyCloseTag', 'anyOpenTag', 'cStyleComment', 'col',
83 'alphanums', 'alphas', 'alphas8bit', 'anyCloseTag', 'anyOpenTag', 'cStyleComment', 'col',
84 'commaSeparatedList', 'commonHTMLEntity', 'countedArray', 'cppStyleComment', 'dblQuotedString',
84 'commaSeparatedList', 'commonHTMLEntity', 'countedArray', 'cppStyleComment', 'dblQuotedString',
85 'dblSlashComment', 'delimitedList', 'dictOf', 'downcaseTokens', 'empty', 'getTokensEndLoc', 'hexnums',
85 'dblSlashComment', 'delimitedList', 'dictOf', 'downcaseTokens', 'empty', 'getTokensEndLoc', 'hexnums',
86 'htmlComment', 'javaStyleComment', 'keepOriginalText', 'line', 'lineEnd', 'lineStart', 'lineno',
86 'htmlComment', 'javaStyleComment', 'keepOriginalText', 'line', 'lineEnd', 'lineStart', 'lineno',
87 'makeHTMLTags', 'makeXMLTags', 'matchOnlyAtCol', 'matchPreviousExpr', 'matchPreviousLiteral',
87 'makeHTMLTags', 'makeXMLTags', 'matchOnlyAtCol', 'matchPreviousExpr', 'matchPreviousLiteral',
88 'nestedExpr', 'nullDebugAction', 'nums', 'oneOf', 'opAssoc', 'operatorPrecedence', 'printables',
88 'nestedExpr', 'nullDebugAction', 'nums', 'oneOf', 'opAssoc', 'operatorPrecedence', 'printables',
89 'punc8bit', 'pythonStyleComment', 'quotedString', 'removeQuotes', 'replaceHTMLEntity',
89 'punc8bit', 'pythonStyleComment', 'quotedString', 'removeQuotes', 'replaceHTMLEntity',
90 'replaceWith', 'restOfLine', 'sglQuotedString', 'srange', 'stringEnd',
90 'replaceWith', 'restOfLine', 'sglQuotedString', 'srange', 'stringEnd',
91 'stringStart', 'traceParseAction', 'unicodeString', 'upcaseTokens', 'withAttribute',
91 'stringStart', 'traceParseAction', 'unicodeString', 'upcaseTokens', 'withAttribute',
92 'indentedBlock', 'originalTextFor',
92 'indentedBlock', 'originalTextFor',
93 ]
93 ]
94
94
95
95
96 """
96 """
97 Detect if we are running version 3.X and make appropriate changes
97 Detect if we are running version 3.X and make appropriate changes
98 Robert A. Clark
98 Robert A. Clark
99 """
99 """
100 if sys.version_info[0] > 2:
100 if sys.version_info[0] > 2:
101 _PY3K = True
101 _PY3K = True
102 _MAX_INT = sys.maxsize
102 _MAX_INT = sys.maxsize
103 basestring = str
103 basestring = str
104 else:
104 else:
105 _PY3K = False
105 _PY3K = False
106 _MAX_INT = sys.maxint
106 _MAX_INT = sys.maxint
107
107
108 if not _PY3K:
108 if not _PY3K:
109 def _ustr(obj):
109 def _ustr(obj):
110 """Drop-in replacement for str(obj) that tries to be Unicode friendly. It first tries
110 """Drop-in replacement for str(obj) that tries to be Unicode friendly. It first tries
111 str(obj). If that fails with a UnicodeEncodeError, then it tries unicode(obj). It
111 str(obj). If that fails with a UnicodeEncodeError, then it tries unicode(obj). It
112 then < returns the unicode object | encodes it with the default encoding | ... >.
112 then < returns the unicode object | encodes it with the default encoding | ... >.
113 """
113 """
114 if isinstance(obj,unicode):
114 if isinstance(obj,unicode):
115 return obj
115 return obj
116
116
117 try:
117 try:
118 # If this works, then _ustr(obj) has the same behaviour as str(obj), so
118 # If this works, then _ustr(obj) has the same behaviour as str(obj), so
119 # it won't break any existing code.
119 # it won't break any existing code.
120 return str(obj)
120 return str(obj)
121
121
122 except UnicodeEncodeError:
122 except UnicodeEncodeError:
123 # The Python docs (http://docs.python.org/ref/customization.html#l2h-182)
123 # The Python docs (http://docs.python.org/ref/customization.html#l2h-182)
124 # state that "The return value must be a string object". However, does a
124 # state that "The return value must be a string object". However, does a
125 # unicode object (being a subclass of basestring) count as a "string
125 # unicode object (being a subclass of basestring) count as a "string
126 # object"?
126 # object"?
127 # If so, then return a unicode object:
127 # If so, then return a unicode object:
128 return unicode(obj)
128 return unicode(obj)
129 # Else encode it... but how? There are many choices... :)
129 # Else encode it... but how? There are many choices... :)
130 # Replace unprintables with escape codes?
130 # Replace unprintables with escape codes?
131 #return unicode(obj).encode(sys.getdefaultencoding(), 'backslashreplace_errors')
131 #return unicode(obj).encode(py3compat.getdefaultencoding(), 'backslashreplace_errors')
132 # Replace unprintables with question marks?
132 # Replace unprintables with question marks?
133 #return unicode(obj).encode(sys.getdefaultencoding(), 'replace')
133 #return unicode(obj).encode(py3compat.getdefaultencoding(), 'replace')
134 # ...
134 # ...
135 else:
135 else:
136 _ustr = str
136 _ustr = str
137 unichr = chr
137 unichr = chr
138
138
139 if not _PY3K:
139 if not _PY3K:
140 def _str2dict(strg):
140 def _str2dict(strg):
141 return dict( [(c,0) for c in strg] )
141 return dict( [(c,0) for c in strg] )
142 else:
142 else:
143 _str2dict = set
143 _str2dict = set
144
144
145 def _xml_escape(data):
145 def _xml_escape(data):
146 """Escape &, <, >, ", ', etc. in a string of data."""
146 """Escape &, <, >, ", ', etc. in a string of data."""
147
147
148 # ampersand must be replaced first
148 # ampersand must be replaced first
149 from_symbols = '&><"\''
149 from_symbols = '&><"\''
150 to_symbols = ['&'+s+';' for s in "amp gt lt quot apos".split()]
150 to_symbols = ['&'+s+';' for s in "amp gt lt quot apos".split()]
151 for from_,to_ in zip(from_symbols, to_symbols):
151 for from_,to_ in zip(from_symbols, to_symbols):
152 data = data.replace(from_, to_)
152 data = data.replace(from_, to_)
153 return data
153 return data
154
154
155 class _Constants(object):
155 class _Constants(object):
156 pass
156 pass
157
157
158 if not _PY3K:
158 if not _PY3K:
159 alphas = string.lowercase + string.uppercase
159 alphas = string.lowercase + string.uppercase
160 else:
160 else:
161 alphas = string.ascii_lowercase + string.ascii_uppercase
161 alphas = string.ascii_lowercase + string.ascii_uppercase
162 nums = string.digits
162 nums = string.digits
163 hexnums = nums + "ABCDEFabcdef"
163 hexnums = nums + "ABCDEFabcdef"
164 alphanums = alphas + nums
164 alphanums = alphas + nums
165 _bslash = chr(92)
165 _bslash = chr(92)
166 printables = "".join( [ c for c in string.printable if c not in string.whitespace ] )
166 printables = "".join( [ c for c in string.printable if c not in string.whitespace ] )
167
167
168 class ParseBaseException(Exception):
168 class ParseBaseException(Exception):
169 """base exception class for all parsing runtime exceptions"""
169 """base exception class for all parsing runtime exceptions"""
170 # Performance tuning: we construct a *lot* of these, so keep this
170 # Performance tuning: we construct a *lot* of these, so keep this
171 # constructor as small and fast as possible
171 # constructor as small and fast as possible
172 def __init__( self, pstr, loc=0, msg=None, elem=None ):
172 def __init__( self, pstr, loc=0, msg=None, elem=None ):
173 self.loc = loc
173 self.loc = loc
174 if msg is None:
174 if msg is None:
175 self.msg = pstr
175 self.msg = pstr
176 self.pstr = ""
176 self.pstr = ""
177 else:
177 else:
178 self.msg = msg
178 self.msg = msg
179 self.pstr = pstr
179 self.pstr = pstr
180 self.parserElement = elem
180 self.parserElement = elem
181
181
182 def __getattr__( self, aname ):
182 def __getattr__( self, aname ):
183 """supported attributes by name are:
183 """supported attributes by name are:
184 - lineno - returns the line number of the exception text
184 - lineno - returns the line number of the exception text
185 - col - returns the column number of the exception text
185 - col - returns the column number of the exception text
186 - line - returns the line containing the exception text
186 - line - returns the line containing the exception text
187 """
187 """
188 if( aname == "lineno" ):
188 if( aname == "lineno" ):
189 return lineno( self.loc, self.pstr )
189 return lineno( self.loc, self.pstr )
190 elif( aname in ("col", "column") ):
190 elif( aname in ("col", "column") ):
191 return col( self.loc, self.pstr )
191 return col( self.loc, self.pstr )
192 elif( aname == "line" ):
192 elif( aname == "line" ):
193 return line( self.loc, self.pstr )
193 return line( self.loc, self.pstr )
194 else:
194 else:
195 raise AttributeError(aname)
195 raise AttributeError(aname)
196
196
197 def __str__( self ):
197 def __str__( self ):
198 return "%s (at char %d), (line:%d, col:%d)" % \
198 return "%s (at char %d), (line:%d, col:%d)" % \
199 ( self.msg, self.loc, self.lineno, self.column )
199 ( self.msg, self.loc, self.lineno, self.column )
200 def __repr__( self ):
200 def __repr__( self ):
201 return _ustr(self)
201 return _ustr(self)
202 def markInputline( self, markerString = ">!<" ):
202 def markInputline( self, markerString = ">!<" ):
203 """Extracts the exception line from the input string, and marks
203 """Extracts the exception line from the input string, and marks
204 the location of the exception with a special symbol.
204 the location of the exception with a special symbol.
205 """
205 """
206 line_str = self.line
206 line_str = self.line
207 line_column = self.column - 1
207 line_column = self.column - 1
208 if markerString:
208 if markerString:
209 line_str = "".join( [line_str[:line_column],
209 line_str = "".join( [line_str[:line_column],
210 markerString, line_str[line_column:]])
210 markerString, line_str[line_column:]])
211 return line_str.strip()
211 return line_str.strip()
212 def __dir__(self):
212 def __dir__(self):
213 return "loc msg pstr parserElement lineno col line " \
213 return "loc msg pstr parserElement lineno col line " \
214 "markInputLine __str__ __repr__".split()
214 "markInputLine __str__ __repr__".split()
215
215
216 class ParseException(ParseBaseException):
216 class ParseException(ParseBaseException):
217 """exception thrown when parse expressions don't match class;
217 """exception thrown when parse expressions don't match class;
218 supported attributes by name are:
218 supported attributes by name are:
219 - lineno - returns the line number of the exception text
219 - lineno - returns the line number of the exception text
220 - col - returns the column number of the exception text
220 - col - returns the column number of the exception text
221 - line - returns the line containing the exception text
221 - line - returns the line containing the exception text
222 """
222 """
223 pass
223 pass
224
224
225 class ParseFatalException(ParseBaseException):
225 class ParseFatalException(ParseBaseException):
226 """user-throwable exception thrown when inconsistent parse content
226 """user-throwable exception thrown when inconsistent parse content
227 is found; stops all parsing immediately"""
227 is found; stops all parsing immediately"""
228 pass
228 pass
229
229
230 class ParseSyntaxException(ParseFatalException):
230 class ParseSyntaxException(ParseFatalException):
231 """just like ParseFatalException, but thrown internally when an
231 """just like ParseFatalException, but thrown internally when an
232 ErrorStop indicates that parsing is to stop immediately because
232 ErrorStop indicates that parsing is to stop immediately because
233 an unbacktrackable syntax error has been found"""
233 an unbacktrackable syntax error has been found"""
234 def __init__(self, pe):
234 def __init__(self, pe):
235 super(ParseSyntaxException, self).__init__(
235 super(ParseSyntaxException, self).__init__(
236 pe.pstr, pe.loc, pe.msg, pe.parserElement)
236 pe.pstr, pe.loc, pe.msg, pe.parserElement)
237
237
238 #~ class ReparseException(ParseBaseException):
238 #~ class ReparseException(ParseBaseException):
239 #~ """Experimental class - parse actions can raise this exception to cause
239 #~ """Experimental class - parse actions can raise this exception to cause
240 #~ pyparsing to reparse the input string:
240 #~ pyparsing to reparse the input string:
241 #~ - with a modified input string, and/or
241 #~ - with a modified input string, and/or
242 #~ - with a modified start location
242 #~ - with a modified start location
243 #~ Set the values of the ReparseException in the constructor, and raise the
243 #~ Set the values of the ReparseException in the constructor, and raise the
244 #~ exception in a parse action to cause pyparsing to use the new string/location.
244 #~ exception in a parse action to cause pyparsing to use the new string/location.
245 #~ Setting the values as None causes no change to be made.
245 #~ Setting the values as None causes no change to be made.
246 #~ """
246 #~ """
247 #~ def __init_( self, newstring, restartLoc ):
247 #~ def __init_( self, newstring, restartLoc ):
248 #~ self.newParseText = newstring
248 #~ self.newParseText = newstring
249 #~ self.reparseLoc = restartLoc
249 #~ self.reparseLoc = restartLoc
250
250
251 class RecursiveGrammarException(Exception):
251 class RecursiveGrammarException(Exception):
252 """exception thrown by validate() if the grammar could be improperly recursive"""
252 """exception thrown by validate() if the grammar could be improperly recursive"""
253 def __init__( self, parseElementList ):
253 def __init__( self, parseElementList ):
254 self.parseElementTrace = parseElementList
254 self.parseElementTrace = parseElementList
255
255
256 def __str__( self ):
256 def __str__( self ):
257 return "RecursiveGrammarException: %s" % self.parseElementTrace
257 return "RecursiveGrammarException: %s" % self.parseElementTrace
258
258
259 class _ParseResultsWithOffset(object):
259 class _ParseResultsWithOffset(object):
260 def __init__(self,p1,p2):
260 def __init__(self,p1,p2):
261 self.tup = (p1,p2)
261 self.tup = (p1,p2)
262 def __getitem__(self,i):
262 def __getitem__(self,i):
263 return self.tup[i]
263 return self.tup[i]
264 def __repr__(self):
264 def __repr__(self):
265 return repr(self.tup)
265 return repr(self.tup)
266 def setOffset(self,i):
266 def setOffset(self,i):
267 self.tup = (self.tup[0],i)
267 self.tup = (self.tup[0],i)
268
268
269 class ParseResults(object):
269 class ParseResults(object):
270 """Structured parse results, to provide multiple means of access to the parsed data:
270 """Structured parse results, to provide multiple means of access to the parsed data:
271 - as a list (len(results))
271 - as a list (len(results))
272 - by list index (results[0], results[1], etc.)
272 - by list index (results[0], results[1], etc.)
273 - by attribute (results.<resultsName>)
273 - by attribute (results.<resultsName>)
274 """
274 """
275 __slots__ = ( "__toklist", "__tokdict", "__doinit", "__name", "__parent", "__accumNames", "__weakref__" )
275 __slots__ = ( "__toklist", "__tokdict", "__doinit", "__name", "__parent", "__accumNames", "__weakref__" )
276 def __new__(cls, toklist, name=None, asList=True, modal=True ):
276 def __new__(cls, toklist, name=None, asList=True, modal=True ):
277 if isinstance(toklist, cls):
277 if isinstance(toklist, cls):
278 return toklist
278 return toklist
279 retobj = object.__new__(cls)
279 retobj = object.__new__(cls)
280 retobj.__doinit = True
280 retobj.__doinit = True
281 return retobj
281 return retobj
282
282
283 # Performance tuning: we construct a *lot* of these, so keep this
283 # Performance tuning: we construct a *lot* of these, so keep this
284 # constructor as small and fast as possible
284 # constructor as small and fast as possible
285 def __init__( self, toklist, name=None, asList=True, modal=True ):
285 def __init__( self, toklist, name=None, asList=True, modal=True ):
286 if self.__doinit:
286 if self.__doinit:
287 self.__doinit = False
287 self.__doinit = False
288 self.__name = None
288 self.__name = None
289 self.__parent = None
289 self.__parent = None
290 self.__accumNames = {}
290 self.__accumNames = {}
291 if isinstance(toklist, list):
291 if isinstance(toklist, list):
292 self.__toklist = toklist[:]
292 self.__toklist = toklist[:]
293 else:
293 else:
294 self.__toklist = [toklist]
294 self.__toklist = [toklist]
295 self.__tokdict = dict()
295 self.__tokdict = dict()
296
296
297 if name:
297 if name:
298 if not modal:
298 if not modal:
299 self.__accumNames[name] = 0
299 self.__accumNames[name] = 0
300 if isinstance(name,int):
300 if isinstance(name,int):
301 name = _ustr(name) # will always return a str, but use _ustr for consistency
301 name = _ustr(name) # will always return a str, but use _ustr for consistency
302 self.__name = name
302 self.__name = name
303 if not toklist in (None,'',[]):
303 if not toklist in (None,'',[]):
304 if isinstance(toklist,basestring):
304 if isinstance(toklist,basestring):
305 toklist = [ toklist ]
305 toklist = [ toklist ]
306 if asList:
306 if asList:
307 if isinstance(toklist,ParseResults):
307 if isinstance(toklist,ParseResults):
308 self[name] = _ParseResultsWithOffset(toklist.copy(),0)
308 self[name] = _ParseResultsWithOffset(toklist.copy(),0)
309 else:
309 else:
310 self[name] = _ParseResultsWithOffset(ParseResults(toklist[0]),0)
310 self[name] = _ParseResultsWithOffset(ParseResults(toklist[0]),0)
311 self[name].__name = name
311 self[name].__name = name
312 else:
312 else:
313 try:
313 try:
314 self[name] = toklist[0]
314 self[name] = toklist[0]
315 except (KeyError,TypeError,IndexError):
315 except (KeyError,TypeError,IndexError):
316 self[name] = toklist
316 self[name] = toklist
317
317
318 def __getitem__( self, i ):
318 def __getitem__( self, i ):
319 if isinstance( i, (int,slice) ):
319 if isinstance( i, (int,slice) ):
320 return self.__toklist[i]
320 return self.__toklist[i]
321 else:
321 else:
322 if i not in self.__accumNames:
322 if i not in self.__accumNames:
323 return self.__tokdict[i][-1][0]
323 return self.__tokdict[i][-1][0]
324 else:
324 else:
325 return ParseResults([ v[0] for v in self.__tokdict[i] ])
325 return ParseResults([ v[0] for v in self.__tokdict[i] ])
326
326
327 def __setitem__( self, k, v ):
327 def __setitem__( self, k, v ):
328 if isinstance(v,_ParseResultsWithOffset):
328 if isinstance(v,_ParseResultsWithOffset):
329 self.__tokdict[k] = self.__tokdict.get(k,list()) + [v]
329 self.__tokdict[k] = self.__tokdict.get(k,list()) + [v]
330 sub = v[0]
330 sub = v[0]
331 elif isinstance(k,int):
331 elif isinstance(k,int):
332 self.__toklist[k] = v
332 self.__toklist[k] = v
333 sub = v
333 sub = v
334 else:
334 else:
335 self.__tokdict[k] = self.__tokdict.get(k,list()) + [_ParseResultsWithOffset(v,0)]
335 self.__tokdict[k] = self.__tokdict.get(k,list()) + [_ParseResultsWithOffset(v,0)]
336 sub = v
336 sub = v
337 if isinstance(sub,ParseResults):
337 if isinstance(sub,ParseResults):
338 sub.__parent = wkref(self)
338 sub.__parent = wkref(self)
339
339
340 def __delitem__( self, i ):
340 def __delitem__( self, i ):
341 if isinstance(i,(int,slice)):
341 if isinstance(i,(int,slice)):
342 mylen = len( self.__toklist )
342 mylen = len( self.__toklist )
343 del self.__toklist[i]
343 del self.__toklist[i]
344
344
345 # convert int to slice
345 # convert int to slice
346 if isinstance(i, int):
346 if isinstance(i, int):
347 if i < 0:
347 if i < 0:
348 i += mylen
348 i += mylen
349 i = slice(i, i+1)
349 i = slice(i, i+1)
350 # get removed indices
350 # get removed indices
351 removed = list(range(*i.indices(mylen)))
351 removed = list(range(*i.indices(mylen)))
352 removed.reverse()
352 removed.reverse()
353 # fixup indices in token dictionary
353 # fixup indices in token dictionary
354 for name in self.__tokdict:
354 for name in self.__tokdict:
355 occurrences = self.__tokdict[name]
355 occurrences = self.__tokdict[name]
356 for j in removed:
356 for j in removed:
357 for k, (value, position) in enumerate(occurrences):
357 for k, (value, position) in enumerate(occurrences):
358 occurrences[k] = _ParseResultsWithOffset(value, position - (position > j))
358 occurrences[k] = _ParseResultsWithOffset(value, position - (position > j))
359 else:
359 else:
360 del self.__tokdict[i]
360 del self.__tokdict[i]
361
361
362 def __contains__( self, k ):
362 def __contains__( self, k ):
363 return k in self.__tokdict
363 return k in self.__tokdict
364
364
365 def __len__( self ): return len( self.__toklist )
365 def __len__( self ): return len( self.__toklist )
366 def __bool__(self): return len( self.__toklist ) > 0
366 def __bool__(self): return len( self.__toklist ) > 0
367 __nonzero__ = __bool__
367 __nonzero__ = __bool__
368 def __iter__( self ): return iter( self.__toklist )
368 def __iter__( self ): return iter( self.__toklist )
369 def __reversed__( self ): return iter( reversed(self.__toklist) )
369 def __reversed__( self ): return iter( reversed(self.__toklist) )
370 def keys( self ):
370 def keys( self ):
371 """Returns all named result keys."""
371 """Returns all named result keys."""
372 return self.__tokdict.keys()
372 return self.__tokdict.keys()
373
373
374 def pop( self, index=-1 ):
374 def pop( self, index=-1 ):
375 """Removes and returns item at specified index (default=last).
375 """Removes and returns item at specified index (default=last).
376 Will work with either numeric indices or dict-key indicies."""
376 Will work with either numeric indices or dict-key indicies."""
377 ret = self[index]
377 ret = self[index]
378 del self[index]
378 del self[index]
379 return ret
379 return ret
380
380
381 def get(self, key, defaultValue=None):
381 def get(self, key, defaultValue=None):
382 """Returns named result matching the given key, or if there is no
382 """Returns named result matching the given key, or if there is no
383 such name, then returns the given defaultValue or None if no
383 such name, then returns the given defaultValue or None if no
384 defaultValue is specified."""
384 defaultValue is specified."""
385 if key in self:
385 if key in self:
386 return self[key]
386 return self[key]
387 else:
387 else:
388 return defaultValue
388 return defaultValue
389
389
390 def insert( self, index, insStr ):
390 def insert( self, index, insStr ):
391 self.__toklist.insert(index, insStr)
391 self.__toklist.insert(index, insStr)
392 # fixup indices in token dictionary
392 # fixup indices in token dictionary
393 for name in self.__tokdict:
393 for name in self.__tokdict:
394 occurrences = self.__tokdict[name]
394 occurrences = self.__tokdict[name]
395 for k, (value, position) in enumerate(occurrences):
395 for k, (value, position) in enumerate(occurrences):
396 occurrences[k] = _ParseResultsWithOffset(value, position + (position > index))
396 occurrences[k] = _ParseResultsWithOffset(value, position + (position > index))
397
397
398 def items( self ):
398 def items( self ):
399 """Returns all named result keys and values as a list of tuples."""
399 """Returns all named result keys and values as a list of tuples."""
400 return [(k,self[k]) for k in self.__tokdict]
400 return [(k,self[k]) for k in self.__tokdict]
401
401
402 def values( self ):
402 def values( self ):
403 """Returns all named result values."""
403 """Returns all named result values."""
404 return [ v[-1][0] for v in self.__tokdict.itervalues() ]
404 return [ v[-1][0] for v in self.__tokdict.itervalues() ]
405
405
406 def __getattr__( self, name ):
406 def __getattr__( self, name ):
407 if name not in self.__slots__:
407 if name not in self.__slots__:
408 if name in self.__tokdict:
408 if name in self.__tokdict:
409 if name not in self.__accumNames:
409 if name not in self.__accumNames:
410 return self.__tokdict[name][-1][0]
410 return self.__tokdict[name][-1][0]
411 else:
411 else:
412 return ParseResults([ v[0] for v in self.__tokdict[name] ])
412 return ParseResults([ v[0] for v in self.__tokdict[name] ])
413 else:
413 else:
414 return ""
414 return ""
415 return None
415 return None
416
416
417 def __add__( self, other ):
417 def __add__( self, other ):
418 ret = self.copy()
418 ret = self.copy()
419 ret += other
419 ret += other
420 return ret
420 return ret
421
421
422 def __iadd__( self, other ):
422 def __iadd__( self, other ):
423 if other.__tokdict:
423 if other.__tokdict:
424 offset = len(self.__toklist)
424 offset = len(self.__toklist)
425 addoffset = ( lambda a: (a<0 and offset) or (a+offset) )
425 addoffset = ( lambda a: (a<0 and offset) or (a+offset) )
426 otheritems = other.__tokdict.iteritems()
426 otheritems = other.__tokdict.iteritems()
427 otherdictitems = [(k, _ParseResultsWithOffset(v[0],addoffset(v[1])) )
427 otherdictitems = [(k, _ParseResultsWithOffset(v[0],addoffset(v[1])) )
428 for (k,vlist) in otheritems for v in vlist]
428 for (k,vlist) in otheritems for v in vlist]
429 for k,v in otherdictitems:
429 for k,v in otherdictitems:
430 self[k] = v
430 self[k] = v
431 if isinstance(v[0],ParseResults):
431 if isinstance(v[0],ParseResults):
432 v[0].__parent = wkref(self)
432 v[0].__parent = wkref(self)
433
433
434 self.__toklist += other.__toklist
434 self.__toklist += other.__toklist
435 self.__accumNames.update( other.__accumNames )
435 self.__accumNames.update( other.__accumNames )
436 del other
436 del other
437 return self
437 return self
438
438
439 def __repr__( self ):
439 def __repr__( self ):
440 return "(%s, %s)" % ( repr( self.__toklist ), repr( self.__tokdict ) )
440 return "(%s, %s)" % ( repr( self.__toklist ), repr( self.__tokdict ) )
441
441
442 def __str__( self ):
442 def __str__( self ):
443 out = "["
443 out = "["
444 sep = ""
444 sep = ""
445 for i in self.__toklist:
445 for i in self.__toklist:
446 if isinstance(i, ParseResults):
446 if isinstance(i, ParseResults):
447 out += sep + _ustr(i)
447 out += sep + _ustr(i)
448 else:
448 else:
449 out += sep + repr(i)
449 out += sep + repr(i)
450 sep = ", "
450 sep = ", "
451 out += "]"
451 out += "]"
452 return out
452 return out
453
453
454 def _asStringList( self, sep='' ):
454 def _asStringList( self, sep='' ):
455 out = []
455 out = []
456 for item in self.__toklist:
456 for item in self.__toklist:
457 if out and sep:
457 if out and sep:
458 out.append(sep)
458 out.append(sep)
459 if isinstance( item, ParseResults ):
459 if isinstance( item, ParseResults ):
460 out += item._asStringList()
460 out += item._asStringList()
461 else:
461 else:
462 out.append( _ustr(item) )
462 out.append( _ustr(item) )
463 return out
463 return out
464
464
465 def asList( self ):
465 def asList( self ):
466 """Returns the parse results as a nested list of matching tokens, all converted to strings."""
466 """Returns the parse results as a nested list of matching tokens, all converted to strings."""
467 out = []
467 out = []
468 for res in self.__toklist:
468 for res in self.__toklist:
469 if isinstance(res,ParseResults):
469 if isinstance(res,ParseResults):
470 out.append( res.asList() )
470 out.append( res.asList() )
471 else:
471 else:
472 out.append( res )
472 out.append( res )
473 return out
473 return out
474
474
475 def asDict( self ):
475 def asDict( self ):
476 """Returns the named parse results as dictionary."""
476 """Returns the named parse results as dictionary."""
477 return dict( self.items() )
477 return dict( self.items() )
478
478
479 def copy( self ):
479 def copy( self ):
480 """Returns a new copy of a ParseResults object."""
480 """Returns a new copy of a ParseResults object."""
481 ret = ParseResults( self.__toklist )
481 ret = ParseResults( self.__toklist )
482 ret.__tokdict = self.__tokdict.copy()
482 ret.__tokdict = self.__tokdict.copy()
483 ret.__parent = self.__parent
483 ret.__parent = self.__parent
484 ret.__accumNames.update( self.__accumNames )
484 ret.__accumNames.update( self.__accumNames )
485 ret.__name = self.__name
485 ret.__name = self.__name
486 return ret
486 return ret
487
487
488 def asXML( self, doctag=None, namedItemsOnly=False, indent="", formatted=True ):
488 def asXML( self, doctag=None, namedItemsOnly=False, indent="", formatted=True ):
489 """Returns the parse results as XML. Tags are created for tokens and lists that have defined results names."""
489 """Returns the parse results as XML. Tags are created for tokens and lists that have defined results names."""
490 nl = "\n"
490 nl = "\n"
491 out = []
491 out = []
492 namedItems = dict([(v[1],k) for (k,vlist) in self.__tokdict.iteritems()
492 namedItems = dict([(v[1],k) for (k,vlist) in self.__tokdict.iteritems()
493 for v in vlist ] )
493 for v in vlist ] )
494 nextLevelIndent = indent + " "
494 nextLevelIndent = indent + " "
495
495
496 # collapse out indents if formatting is not desired
496 # collapse out indents if formatting is not desired
497 if not formatted:
497 if not formatted:
498 indent = ""
498 indent = ""
499 nextLevelIndent = ""
499 nextLevelIndent = ""
500 nl = ""
500 nl = ""
501
501
502 selfTag = None
502 selfTag = None
503 if doctag is not None:
503 if doctag is not None:
504 selfTag = doctag
504 selfTag = doctag
505 else:
505 else:
506 if self.__name:
506 if self.__name:
507 selfTag = self.__name
507 selfTag = self.__name
508
508
509 if not selfTag:
509 if not selfTag:
510 if namedItemsOnly:
510 if namedItemsOnly:
511 return ""
511 return ""
512 else:
512 else:
513 selfTag = "ITEM"
513 selfTag = "ITEM"
514
514
515 out += [ nl, indent, "<", selfTag, ">" ]
515 out += [ nl, indent, "<", selfTag, ">" ]
516
516
517 worklist = self.__toklist
517 worklist = self.__toklist
518 for i,res in enumerate(worklist):
518 for i,res in enumerate(worklist):
519 if isinstance(res,ParseResults):
519 if isinstance(res,ParseResults):
520 if i in namedItems:
520 if i in namedItems:
521 out += [ res.asXML(namedItems[i],
521 out += [ res.asXML(namedItems[i],
522 namedItemsOnly and doctag is None,
522 namedItemsOnly and doctag is None,
523 nextLevelIndent,
523 nextLevelIndent,
524 formatted)]
524 formatted)]
525 else:
525 else:
526 out += [ res.asXML(None,
526 out += [ res.asXML(None,
527 namedItemsOnly and doctag is None,
527 namedItemsOnly and doctag is None,
528 nextLevelIndent,
528 nextLevelIndent,
529 formatted)]
529 formatted)]
530 else:
530 else:
531 # individual token, see if there is a name for it
531 # individual token, see if there is a name for it
532 resTag = None
532 resTag = None
533 if i in namedItems:
533 if i in namedItems:
534 resTag = namedItems[i]
534 resTag = namedItems[i]
535 if not resTag:
535 if not resTag:
536 if namedItemsOnly:
536 if namedItemsOnly:
537 continue
537 continue
538 else:
538 else:
539 resTag = "ITEM"
539 resTag = "ITEM"
540 xmlBodyText = _xml_escape(_ustr(res))
540 xmlBodyText = _xml_escape(_ustr(res))
541 out += [ nl, nextLevelIndent, "<", resTag, ">",
541 out += [ nl, nextLevelIndent, "<", resTag, ">",
542 xmlBodyText,
542 xmlBodyText,
543 "</", resTag, ">" ]
543 "</", resTag, ">" ]
544
544
545 out += [ nl, indent, "</", selfTag, ">" ]
545 out += [ nl, indent, "</", selfTag, ">" ]
546 return "".join(out)
546 return "".join(out)
547
547
548 def __lookup(self,sub):
548 def __lookup(self,sub):
549 for k,vlist in self.__tokdict.iteritems():
549 for k,vlist in self.__tokdict.iteritems():
550 for v,loc in vlist:
550 for v,loc in vlist:
551 if sub is v:
551 if sub is v:
552 return k
552 return k
553 return None
553 return None
554
554
555 def getName(self):
555 def getName(self):
556 """Returns the results name for this token expression."""
556 """Returns the results name for this token expression."""
557 if self.__name:
557 if self.__name:
558 return self.__name
558 return self.__name
559 elif self.__parent:
559 elif self.__parent:
560 par = self.__parent()
560 par = self.__parent()
561 if par:
561 if par:
562 return par.__lookup(self)
562 return par.__lookup(self)
563 else:
563 else:
564 return None
564 return None
565 elif (len(self) == 1 and
565 elif (len(self) == 1 and
566 len(self.__tokdict) == 1 and
566 len(self.__tokdict) == 1 and
567 self.__tokdict.values()[0][0][1] in (0,-1)):
567 self.__tokdict.values()[0][0][1] in (0,-1)):
568 return self.__tokdict.keys()[0]
568 return self.__tokdict.keys()[0]
569 else:
569 else:
570 return None
570 return None
571
571
572 def dump(self,indent='',depth=0):
572 def dump(self,indent='',depth=0):
573 """Diagnostic method for listing out the contents of a ParseResults.
573 """Diagnostic method for listing out the contents of a ParseResults.
574 Accepts an optional indent argument so that this string can be embedded
574 Accepts an optional indent argument so that this string can be embedded
575 in a nested display of other data."""
575 in a nested display of other data."""
576 out = []
576 out = []
577 out.append( indent+_ustr(self.asList()) )
577 out.append( indent+_ustr(self.asList()) )
578 keys = self.items()
578 keys = self.items()
579 keys.sort()
579 keys.sort()
580 for k,v in keys:
580 for k,v in keys:
581 if out:
581 if out:
582 out.append('\n')
582 out.append('\n')
583 out.append( "%s%s- %s: " % (indent,(' '*depth), k) )
583 out.append( "%s%s- %s: " % (indent,(' '*depth), k) )
584 if isinstance(v,ParseResults):
584 if isinstance(v,ParseResults):
585 if v.keys():
585 if v.keys():
586 #~ out.append('\n')
586 #~ out.append('\n')
587 out.append( v.dump(indent,depth+1) )
587 out.append( v.dump(indent,depth+1) )
588 #~ out.append('\n')
588 #~ out.append('\n')
589 else:
589 else:
590 out.append(_ustr(v))
590 out.append(_ustr(v))
591 else:
591 else:
592 out.append(_ustr(v))
592 out.append(_ustr(v))
593 #~ out.append('\n')
593 #~ out.append('\n')
594 return "".join(out)
594 return "".join(out)
595
595
596 # add support for pickle protocol
596 # add support for pickle protocol
597 def __getstate__(self):
597 def __getstate__(self):
598 return ( self.__toklist,
598 return ( self.__toklist,
599 ( self.__tokdict.copy(),
599 ( self.__tokdict.copy(),
600 self.__parent is not None and self.__parent() or None,
600 self.__parent is not None and self.__parent() or None,
601 self.__accumNames,
601 self.__accumNames,
602 self.__name ) )
602 self.__name ) )
603
603
604 def __setstate__(self,state):
604 def __setstate__(self,state):
605 self.__toklist = state[0]
605 self.__toklist = state[0]
606 self.__tokdict, \
606 self.__tokdict, \
607 par, \
607 par, \
608 inAccumNames, \
608 inAccumNames, \
609 self.__name = state[1]
609 self.__name = state[1]
610 self.__accumNames = {}
610 self.__accumNames = {}
611 self.__accumNames.update(inAccumNames)
611 self.__accumNames.update(inAccumNames)
612 if par is not None:
612 if par is not None:
613 self.__parent = wkref(par)
613 self.__parent = wkref(par)
614 else:
614 else:
615 self.__parent = None
615 self.__parent = None
616
616
617 def __dir__(self):
617 def __dir__(self):
618 return dir(super(ParseResults,self)) + self.keys()
618 return dir(super(ParseResults,self)) + self.keys()
619
619
620 def col (loc,strg):
620 def col (loc,strg):
621 """Returns current column within a string, counting newlines as line separators.
621 """Returns current column within a string, counting newlines as line separators.
622 The first column is number 1.
622 The first column is number 1.
623
623
624 Note: the default parsing behavior is to expand tabs in the input string
624 Note: the default parsing behavior is to expand tabs in the input string
625 before starting the parsing process. See L{I{ParserElement.parseString}<ParserElement.parseString>} for more information
625 before starting the parsing process. See L{I{ParserElement.parseString}<ParserElement.parseString>} for more information
626 on parsing strings containing <TAB>s, and suggested methods to maintain a
626 on parsing strings containing <TAB>s, and suggested methods to maintain a
627 consistent view of the parsed string, the parse location, and line and column
627 consistent view of the parsed string, the parse location, and line and column
628 positions within the parsed string.
628 positions within the parsed string.
629 """
629 """
630 return (loc<len(strg) and strg[loc] == '\n') and 1 or loc - strg.rfind("\n", 0, loc)
630 return (loc<len(strg) and strg[loc] == '\n') and 1 or loc - strg.rfind("\n", 0, loc)
631
631
632 def lineno(loc,strg):
632 def lineno(loc,strg):
633 """Returns current line number within a string, counting newlines as line separators.
633 """Returns current line number within a string, counting newlines as line separators.
634 The first line is number 1.
634 The first line is number 1.
635
635
636 Note: the default parsing behavior is to expand tabs in the input string
636 Note: the default parsing behavior is to expand tabs in the input string
637 before starting the parsing process. See L{I{ParserElement.parseString}<ParserElement.parseString>} for more information
637 before starting the parsing process. See L{I{ParserElement.parseString}<ParserElement.parseString>} for more information
638 on parsing strings containing <TAB>s, and suggested methods to maintain a
638 on parsing strings containing <TAB>s, and suggested methods to maintain a
639 consistent view of the parsed string, the parse location, and line and column
639 consistent view of the parsed string, the parse location, and line and column
640 positions within the parsed string.
640 positions within the parsed string.
641 """
641 """
642 return strg.count("\n",0,loc) + 1
642 return strg.count("\n",0,loc) + 1
643
643
644 def line( loc, strg ):
644 def line( loc, strg ):
645 """Returns the line of text containing loc within a string, counting newlines as line separators.
645 """Returns the line of text containing loc within a string, counting newlines as line separators.
646 """
646 """
647 lastCR = strg.rfind("\n", 0, loc)
647 lastCR = strg.rfind("\n", 0, loc)
648 nextCR = strg.find("\n", loc)
648 nextCR = strg.find("\n", loc)
649 if nextCR > 0:
649 if nextCR > 0:
650 return strg[lastCR+1:nextCR]
650 return strg[lastCR+1:nextCR]
651 else:
651 else:
652 return strg[lastCR+1:]
652 return strg[lastCR+1:]
653
653
654 def _defaultStartDebugAction( instring, loc, expr ):
654 def _defaultStartDebugAction( instring, loc, expr ):
655 print ("Match " + _ustr(expr) + " at loc " + _ustr(loc) + "(%d,%d)" % ( lineno(loc,instring), col(loc,instring) ))
655 print ("Match " + _ustr(expr) + " at loc " + _ustr(loc) + "(%d,%d)" % ( lineno(loc,instring), col(loc,instring) ))
656
656
657 def _defaultSuccessDebugAction( instring, startloc, endloc, expr, toks ):
657 def _defaultSuccessDebugAction( instring, startloc, endloc, expr, toks ):
658 print ("Matched " + _ustr(expr) + " -> " + str(toks.asList()))
658 print ("Matched " + _ustr(expr) + " -> " + str(toks.asList()))
659
659
660 def _defaultExceptionDebugAction( instring, loc, expr, exc ):
660 def _defaultExceptionDebugAction( instring, loc, expr, exc ):
661 print ("Exception raised:" + _ustr(exc))
661 print ("Exception raised:" + _ustr(exc))
662
662
663 def nullDebugAction(*args):
663 def nullDebugAction(*args):
664 """'Do-nothing' debug action, to suppress debugging output during parsing."""
664 """'Do-nothing' debug action, to suppress debugging output during parsing."""
665 pass
665 pass
666
666
667 class ParserElement(object):
667 class ParserElement(object):
668 """Abstract base level parser element class."""
668 """Abstract base level parser element class."""
669 DEFAULT_WHITE_CHARS = " \n\t\r"
669 DEFAULT_WHITE_CHARS = " \n\t\r"
670
670
671 def setDefaultWhitespaceChars( chars ):
671 def setDefaultWhitespaceChars( chars ):
672 """Overrides the default whitespace chars
672 """Overrides the default whitespace chars
673 """
673 """
674 ParserElement.DEFAULT_WHITE_CHARS = chars
674 ParserElement.DEFAULT_WHITE_CHARS = chars
675 setDefaultWhitespaceChars = staticmethod(setDefaultWhitespaceChars)
675 setDefaultWhitespaceChars = staticmethod(setDefaultWhitespaceChars)
676
676
677 def __init__( self, savelist=False ):
677 def __init__( self, savelist=False ):
678 self.parseAction = list()
678 self.parseAction = list()
679 self.failAction = None
679 self.failAction = None
680 #~ self.name = "<unknown>" # don't define self.name, let subclasses try/except upcall
680 #~ self.name = "<unknown>" # don't define self.name, let subclasses try/except upcall
681 self.strRepr = None
681 self.strRepr = None
682 self.resultsName = None
682 self.resultsName = None
683 self.saveAsList = savelist
683 self.saveAsList = savelist
684 self.skipWhitespace = True
684 self.skipWhitespace = True
685 self.whiteChars = ParserElement.DEFAULT_WHITE_CHARS
685 self.whiteChars = ParserElement.DEFAULT_WHITE_CHARS
686 self.copyDefaultWhiteChars = True
686 self.copyDefaultWhiteChars = True
687 self.mayReturnEmpty = False # used when checking for left-recursion
687 self.mayReturnEmpty = False # used when checking for left-recursion
688 self.keepTabs = False
688 self.keepTabs = False
689 self.ignoreExprs = list()
689 self.ignoreExprs = list()
690 self.debug = False
690 self.debug = False
691 self.streamlined = False
691 self.streamlined = False
692 self.mayIndexError = True # used to optimize exception handling for subclasses that don't advance parse index
692 self.mayIndexError = True # used to optimize exception handling for subclasses that don't advance parse index
693 self.errmsg = ""
693 self.errmsg = ""
694 self.modalResults = True # used to mark results names as modal (report only last) or cumulative (list all)
694 self.modalResults = True # used to mark results names as modal (report only last) or cumulative (list all)
695 self.debugActions = ( None, None, None ) #custom debug actions
695 self.debugActions = ( None, None, None ) #custom debug actions
696 self.re = None
696 self.re = None
697 self.callPreparse = True # used to avoid redundant calls to preParse
697 self.callPreparse = True # used to avoid redundant calls to preParse
698 self.callDuringTry = False
698 self.callDuringTry = False
699
699
700 def copy( self ):
700 def copy( self ):
701 """Make a copy of this ParserElement. Useful for defining different parse actions
701 """Make a copy of this ParserElement. Useful for defining different parse actions
702 for the same parsing pattern, using copies of the original parse element."""
702 for the same parsing pattern, using copies of the original parse element."""
703 cpy = copy.copy( self )
703 cpy = copy.copy( self )
704 cpy.parseAction = self.parseAction[:]
704 cpy.parseAction = self.parseAction[:]
705 cpy.ignoreExprs = self.ignoreExprs[:]
705 cpy.ignoreExprs = self.ignoreExprs[:]
706 if self.copyDefaultWhiteChars:
706 if self.copyDefaultWhiteChars:
707 cpy.whiteChars = ParserElement.DEFAULT_WHITE_CHARS
707 cpy.whiteChars = ParserElement.DEFAULT_WHITE_CHARS
708 return cpy
708 return cpy
709
709
710 def setName( self, name ):
710 def setName( self, name ):
711 """Define name for this expression, for use in debugging."""
711 """Define name for this expression, for use in debugging."""
712 self.name = name
712 self.name = name
713 self.errmsg = "Expected " + self.name
713 self.errmsg = "Expected " + self.name
714 if hasattr(self,"exception"):
714 if hasattr(self,"exception"):
715 self.exception.msg = self.errmsg
715 self.exception.msg = self.errmsg
716 return self
716 return self
717
717
718 def setResultsName( self, name, listAllMatches=False ):
718 def setResultsName( self, name, listAllMatches=False ):
719 """Define name for referencing matching tokens as a nested attribute
719 """Define name for referencing matching tokens as a nested attribute
720 of the returned parse results.
720 of the returned parse results.
721 NOTE: this returns a *copy* of the original ParserElement object;
721 NOTE: this returns a *copy* of the original ParserElement object;
722 this is so that the client can define a basic element, such as an
722 this is so that the client can define a basic element, such as an
723 integer, and reference it in multiple places with different names.
723 integer, and reference it in multiple places with different names.
724 """
724 """
725 newself = self.copy()
725 newself = self.copy()
726 newself.resultsName = name
726 newself.resultsName = name
727 newself.modalResults = not listAllMatches
727 newself.modalResults = not listAllMatches
728 return newself
728 return newself
729
729
730 def setBreak(self,breakFlag = True):
730 def setBreak(self,breakFlag = True):
731 """Method to invoke the Python pdb debugger when this element is
731 """Method to invoke the Python pdb debugger when this element is
732 about to be parsed. Set breakFlag to True to enable, False to
732 about to be parsed. Set breakFlag to True to enable, False to
733 disable.
733 disable.
734 """
734 """
735 if breakFlag:
735 if breakFlag:
736 _parseMethod = self._parse
736 _parseMethod = self._parse
737 def breaker(instring, loc, doActions=True, callPreParse=True):
737 def breaker(instring, loc, doActions=True, callPreParse=True):
738 import pdb
738 import pdb
739 pdb.set_trace()
739 pdb.set_trace()
740 return _parseMethod( instring, loc, doActions, callPreParse )
740 return _parseMethod( instring, loc, doActions, callPreParse )
741 breaker._originalParseMethod = _parseMethod
741 breaker._originalParseMethod = _parseMethod
742 self._parse = breaker
742 self._parse = breaker
743 else:
743 else:
744 if hasattr(self._parse,"_originalParseMethod"):
744 if hasattr(self._parse,"_originalParseMethod"):
745 self._parse = self._parse._originalParseMethod
745 self._parse = self._parse._originalParseMethod
746 return self
746 return self
747
747
748 def _normalizeParseActionArgs( f ):
748 def _normalizeParseActionArgs( f ):
749 """Internal method used to decorate parse actions that take fewer than 3 arguments,
749 """Internal method used to decorate parse actions that take fewer than 3 arguments,
750 so that all parse actions can be called as f(s,l,t)."""
750 so that all parse actions can be called as f(s,l,t)."""
751 STAR_ARGS = 4
751 STAR_ARGS = 4
752
752
753 try:
753 try:
754 restore = None
754 restore = None
755 if isinstance(f,type):
755 if isinstance(f,type):
756 restore = f
756 restore = f
757 f = f.__init__
757 f = f.__init__
758 if not _PY3K:
758 if not _PY3K:
759 codeObj = f.func_code
759 codeObj = f.func_code
760 else:
760 else:
761 codeObj = f.code
761 codeObj = f.code
762 if codeObj.co_flags & STAR_ARGS:
762 if codeObj.co_flags & STAR_ARGS:
763 return f
763 return f
764 numargs = codeObj.co_argcount
764 numargs = codeObj.co_argcount
765 if not _PY3K:
765 if not _PY3K:
766 if hasattr(f,"im_self"):
766 if hasattr(f,"im_self"):
767 numargs -= 1
767 numargs -= 1
768 else:
768 else:
769 if hasattr(f,"__self__"):
769 if hasattr(f,"__self__"):
770 numargs -= 1
770 numargs -= 1
771 if restore:
771 if restore:
772 f = restore
772 f = restore
773 except AttributeError:
773 except AttributeError:
774 try:
774 try:
775 if not _PY3K:
775 if not _PY3K:
776 call_im_func_code = f.__call__.im_func.func_code
776 call_im_func_code = f.__call__.im_func.func_code
777 else:
777 else:
778 call_im_func_code = f.__code__
778 call_im_func_code = f.__code__
779
779
780 # not a function, must be a callable object, get info from the
780 # not a function, must be a callable object, get info from the
781 # im_func binding of its bound __call__ method
781 # im_func binding of its bound __call__ method
782 if call_im_func_code.co_flags & STAR_ARGS:
782 if call_im_func_code.co_flags & STAR_ARGS:
783 return f
783 return f
784 numargs = call_im_func_code.co_argcount
784 numargs = call_im_func_code.co_argcount
785 if not _PY3K:
785 if not _PY3K:
786 if hasattr(f.__call__,"im_self"):
786 if hasattr(f.__call__,"im_self"):
787 numargs -= 1
787 numargs -= 1
788 else:
788 else:
789 if hasattr(f.__call__,"__self__"):
789 if hasattr(f.__call__,"__self__"):
790 numargs -= 0
790 numargs -= 0
791 except AttributeError:
791 except AttributeError:
792 if not _PY3K:
792 if not _PY3K:
793 call_func_code = f.__call__.func_code
793 call_func_code = f.__call__.func_code
794 else:
794 else:
795 call_func_code = f.__call__.__code__
795 call_func_code = f.__call__.__code__
796 # not a bound method, get info directly from __call__ method
796 # not a bound method, get info directly from __call__ method
797 if call_func_code.co_flags & STAR_ARGS:
797 if call_func_code.co_flags & STAR_ARGS:
798 return f
798 return f
799 numargs = call_func_code.co_argcount
799 numargs = call_func_code.co_argcount
800 if not _PY3K:
800 if not _PY3K:
801 if hasattr(f.__call__,"im_self"):
801 if hasattr(f.__call__,"im_self"):
802 numargs -= 1
802 numargs -= 1
803 else:
803 else:
804 if hasattr(f.__call__,"__self__"):
804 if hasattr(f.__call__,"__self__"):
805 numargs -= 1
805 numargs -= 1
806
806
807
807
808 #~ print ("adding function %s with %d args" % (f.func_name,numargs))
808 #~ print ("adding function %s with %d args" % (f.func_name,numargs))
809 if numargs == 3:
809 if numargs == 3:
810 return f
810 return f
811 else:
811 else:
812 if numargs > 3:
812 if numargs > 3:
813 def tmp(s,l,t):
813 def tmp(s,l,t):
814 return f(f.__call__.__self__, s,l,t)
814 return f(f.__call__.__self__, s,l,t)
815 if numargs == 2:
815 if numargs == 2:
816 def tmp(s,l,t):
816 def tmp(s,l,t):
817 return f(l,t)
817 return f(l,t)
818 elif numargs == 1:
818 elif numargs == 1:
819 def tmp(s,l,t):
819 def tmp(s,l,t):
820 return f(t)
820 return f(t)
821 else: #~ numargs == 0:
821 else: #~ numargs == 0:
822 def tmp(s,l,t):
822 def tmp(s,l,t):
823 return f()
823 return f()
824 try:
824 try:
825 tmp.__name__ = f.__name__
825 tmp.__name__ = f.__name__
826 except (AttributeError,TypeError):
826 except (AttributeError,TypeError):
827 # no need for special handling if attribute doesnt exist
827 # no need for special handling if attribute doesnt exist
828 pass
828 pass
829 try:
829 try:
830 tmp.__doc__ = f.__doc__
830 tmp.__doc__ = f.__doc__
831 except (AttributeError,TypeError):
831 except (AttributeError,TypeError):
832 # no need for special handling if attribute doesnt exist
832 # no need for special handling if attribute doesnt exist
833 pass
833 pass
834 try:
834 try:
835 tmp.__dict__.update(f.__dict__)
835 tmp.__dict__.update(f.__dict__)
836 except (AttributeError,TypeError):
836 except (AttributeError,TypeError):
837 # no need for special handling if attribute doesnt exist
837 # no need for special handling if attribute doesnt exist
838 pass
838 pass
839 return tmp
839 return tmp
840 _normalizeParseActionArgs = staticmethod(_normalizeParseActionArgs)
840 _normalizeParseActionArgs = staticmethod(_normalizeParseActionArgs)
841
841
842 def setParseAction( self, *fns, **kwargs ):
842 def setParseAction( self, *fns, **kwargs ):
843 """Define action to perform when successfully matching parse element definition.
843 """Define action to perform when successfully matching parse element definition.
844 Parse action fn is a callable method with 0-3 arguments, called as fn(s,loc,toks),
844 Parse action fn is a callable method with 0-3 arguments, called as fn(s,loc,toks),
845 fn(loc,toks), fn(toks), or just fn(), where:
845 fn(loc,toks), fn(toks), or just fn(), where:
846 - s = the original string being parsed (see note below)
846 - s = the original string being parsed (see note below)
847 - loc = the location of the matching substring
847 - loc = the location of the matching substring
848 - toks = a list of the matched tokens, packaged as a ParseResults object
848 - toks = a list of the matched tokens, packaged as a ParseResults object
849 If the functions in fns modify the tokens, they can return them as the return
849 If the functions in fns modify the tokens, they can return them as the return
850 value from fn, and the modified list of tokens will replace the original.
850 value from fn, and the modified list of tokens will replace the original.
851 Otherwise, fn does not need to return any value.
851 Otherwise, fn does not need to return any value.
852
852
853 Note: the default parsing behavior is to expand tabs in the input string
853 Note: the default parsing behavior is to expand tabs in the input string
854 before starting the parsing process. See L{I{parseString}<parseString>} for more information
854 before starting the parsing process. See L{I{parseString}<parseString>} for more information
855 on parsing strings containing <TAB>s, and suggested methods to maintain a
855 on parsing strings containing <TAB>s, and suggested methods to maintain a
856 consistent view of the parsed string, the parse location, and line and column
856 consistent view of the parsed string, the parse location, and line and column
857 positions within the parsed string.
857 positions within the parsed string.
858 """
858 """
859 self.parseAction = list(map(self._normalizeParseActionArgs, list(fns)))
859 self.parseAction = list(map(self._normalizeParseActionArgs, list(fns)))
860 self.callDuringTry = ("callDuringTry" in kwargs and kwargs["callDuringTry"])
860 self.callDuringTry = ("callDuringTry" in kwargs and kwargs["callDuringTry"])
861 return self
861 return self
862
862
863 def addParseAction( self, *fns, **kwargs ):
863 def addParseAction( self, *fns, **kwargs ):
864 """Add parse action to expression's list of parse actions. See L{I{setParseAction}<setParseAction>}."""
864 """Add parse action to expression's list of parse actions. See L{I{setParseAction}<setParseAction>}."""
865 self.parseAction += list(map(self._normalizeParseActionArgs, list(fns)))
865 self.parseAction += list(map(self._normalizeParseActionArgs, list(fns)))
866 self.callDuringTry = self.callDuringTry or ("callDuringTry" in kwargs and kwargs["callDuringTry"])
866 self.callDuringTry = self.callDuringTry or ("callDuringTry" in kwargs and kwargs["callDuringTry"])
867 return self
867 return self
868
868
869 def setFailAction( self, fn ):
869 def setFailAction( self, fn ):
870 """Define action to perform if parsing fails at this expression.
870 """Define action to perform if parsing fails at this expression.
871 Fail acton fn is a callable function that takes the arguments
871 Fail acton fn is a callable function that takes the arguments
872 fn(s,loc,expr,err) where:
872 fn(s,loc,expr,err) where:
873 - s = string being parsed
873 - s = string being parsed
874 - loc = location where expression match was attempted and failed
874 - loc = location where expression match was attempted and failed
875 - expr = the parse expression that failed
875 - expr = the parse expression that failed
876 - err = the exception thrown
876 - err = the exception thrown
877 The function returns no value. It may throw ParseFatalException
877 The function returns no value. It may throw ParseFatalException
878 if it is desired to stop parsing immediately."""
878 if it is desired to stop parsing immediately."""
879 self.failAction = fn
879 self.failAction = fn
880 return self
880 return self
881
881
882 def _skipIgnorables( self, instring, loc ):
882 def _skipIgnorables( self, instring, loc ):
883 exprsFound = True
883 exprsFound = True
884 while exprsFound:
884 while exprsFound:
885 exprsFound = False
885 exprsFound = False
886 for e in self.ignoreExprs:
886 for e in self.ignoreExprs:
887 try:
887 try:
888 while 1:
888 while 1:
889 loc,dummy = e._parse( instring, loc )
889 loc,dummy = e._parse( instring, loc )
890 exprsFound = True
890 exprsFound = True
891 except ParseException:
891 except ParseException:
892 pass
892 pass
893 return loc
893 return loc
894
894
895 def preParse( self, instring, loc ):
895 def preParse( self, instring, loc ):
896 if self.ignoreExprs:
896 if self.ignoreExprs:
897 loc = self._skipIgnorables( instring, loc )
897 loc = self._skipIgnorables( instring, loc )
898
898
899 if self.skipWhitespace:
899 if self.skipWhitespace:
900 wt = self.whiteChars
900 wt = self.whiteChars
901 instrlen = len(instring)
901 instrlen = len(instring)
902 while loc < instrlen and instring[loc] in wt:
902 while loc < instrlen and instring[loc] in wt:
903 loc += 1
903 loc += 1
904
904
905 return loc
905 return loc
906
906
907 def parseImpl( self, instring, loc, doActions=True ):
907 def parseImpl( self, instring, loc, doActions=True ):
908 return loc, []
908 return loc, []
909
909
910 def postParse( self, instring, loc, tokenlist ):
910 def postParse( self, instring, loc, tokenlist ):
911 return tokenlist
911 return tokenlist
912
912
913 #~ @profile
913 #~ @profile
914 def _parseNoCache( self, instring, loc, doActions=True, callPreParse=True ):
914 def _parseNoCache( self, instring, loc, doActions=True, callPreParse=True ):
915 debugging = ( self.debug ) #and doActions )
915 debugging = ( self.debug ) #and doActions )
916
916
917 if debugging or self.failAction:
917 if debugging or self.failAction:
918 #~ print ("Match",self,"at loc",loc,"(%d,%d)" % ( lineno(loc,instring), col(loc,instring) ))
918 #~ print ("Match",self,"at loc",loc,"(%d,%d)" % ( lineno(loc,instring), col(loc,instring) ))
919 if (self.debugActions[0] ):
919 if (self.debugActions[0] ):
920 self.debugActions[0]( instring, loc, self )
920 self.debugActions[0]( instring, loc, self )
921 if callPreParse and self.callPreparse:
921 if callPreParse and self.callPreparse:
922 preloc = self.preParse( instring, loc )
922 preloc = self.preParse( instring, loc )
923 else:
923 else:
924 preloc = loc
924 preloc = loc
925 tokensStart = loc
925 tokensStart = loc
926 try:
926 try:
927 try:
927 try:
928 loc,tokens = self.parseImpl( instring, preloc, doActions )
928 loc,tokens = self.parseImpl( instring, preloc, doActions )
929 except IndexError:
929 except IndexError:
930 raise ParseException( instring, len(instring), self.errmsg, self )
930 raise ParseException( instring, len(instring), self.errmsg, self )
931 except ParseBaseException, err:
931 except ParseBaseException, err:
932 #~ print ("Exception raised:", err)
932 #~ print ("Exception raised:", err)
933 if self.debugActions[2]:
933 if self.debugActions[2]:
934 self.debugActions[2]( instring, tokensStart, self, err )
934 self.debugActions[2]( instring, tokensStart, self, err )
935 if self.failAction:
935 if self.failAction:
936 self.failAction( instring, tokensStart, self, err )
936 self.failAction( instring, tokensStart, self, err )
937 raise
937 raise
938 else:
938 else:
939 if callPreParse and self.callPreparse:
939 if callPreParse and self.callPreparse:
940 preloc = self.preParse( instring, loc )
940 preloc = self.preParse( instring, loc )
941 else:
941 else:
942 preloc = loc
942 preloc = loc
943 tokensStart = loc
943 tokensStart = loc
944 if self.mayIndexError or loc >= len(instring):
944 if self.mayIndexError or loc >= len(instring):
945 try:
945 try:
946 loc,tokens = self.parseImpl( instring, preloc, doActions )
946 loc,tokens = self.parseImpl( instring, preloc, doActions )
947 except IndexError:
947 except IndexError:
948 raise ParseException( instring, len(instring), self.errmsg, self )
948 raise ParseException( instring, len(instring), self.errmsg, self )
949 else:
949 else:
950 loc,tokens = self.parseImpl( instring, preloc, doActions )
950 loc,tokens = self.parseImpl( instring, preloc, doActions )
951
951
952 tokens = self.postParse( instring, loc, tokens )
952 tokens = self.postParse( instring, loc, tokens )
953
953
954 retTokens = ParseResults( tokens, self.resultsName, asList=self.saveAsList, modal=self.modalResults )
954 retTokens = ParseResults( tokens, self.resultsName, asList=self.saveAsList, modal=self.modalResults )
955 if self.parseAction and (doActions or self.callDuringTry):
955 if self.parseAction and (doActions or self.callDuringTry):
956 if debugging:
956 if debugging:
957 try:
957 try:
958 for fn in self.parseAction:
958 for fn in self.parseAction:
959 tokens = fn( instring, tokensStart, retTokens )
959 tokens = fn( instring, tokensStart, retTokens )
960 if tokens is not None:
960 if tokens is not None:
961 retTokens = ParseResults( tokens,
961 retTokens = ParseResults( tokens,
962 self.resultsName,
962 self.resultsName,
963 asList=self.saveAsList and isinstance(tokens,(ParseResults,list)),
963 asList=self.saveAsList and isinstance(tokens,(ParseResults,list)),
964 modal=self.modalResults )
964 modal=self.modalResults )
965 except ParseBaseException, err:
965 except ParseBaseException, err:
966 #~ print "Exception raised in user parse action:", err
966 #~ print "Exception raised in user parse action:", err
967 if (self.debugActions[2] ):
967 if (self.debugActions[2] ):
968 self.debugActions[2]( instring, tokensStart, self, err )
968 self.debugActions[2]( instring, tokensStart, self, err )
969 raise
969 raise
970 else:
970 else:
971 for fn in self.parseAction:
971 for fn in self.parseAction:
972 tokens = fn( instring, tokensStart, retTokens )
972 tokens = fn( instring, tokensStart, retTokens )
973 if tokens is not None:
973 if tokens is not None:
974 retTokens = ParseResults( tokens,
974 retTokens = ParseResults( tokens,
975 self.resultsName,
975 self.resultsName,
976 asList=self.saveAsList and isinstance(tokens,(ParseResults,list)),
976 asList=self.saveAsList and isinstance(tokens,(ParseResults,list)),
977 modal=self.modalResults )
977 modal=self.modalResults )
978
978
979 if debugging:
979 if debugging:
980 #~ print ("Matched",self,"->",retTokens.asList())
980 #~ print ("Matched",self,"->",retTokens.asList())
981 if (self.debugActions[1] ):
981 if (self.debugActions[1] ):
982 self.debugActions[1]( instring, tokensStart, loc, self, retTokens )
982 self.debugActions[1]( instring, tokensStart, loc, self, retTokens )
983
983
984 return loc, retTokens
984 return loc, retTokens
985
985
986 def tryParse( self, instring, loc ):
986 def tryParse( self, instring, loc ):
987 try:
987 try:
988 return self._parse( instring, loc, doActions=False )[0]
988 return self._parse( instring, loc, doActions=False )[0]
989 except ParseFatalException:
989 except ParseFatalException:
990 raise ParseException( instring, loc, self.errmsg, self)
990 raise ParseException( instring, loc, self.errmsg, self)
991
991
992 # this method gets repeatedly called during backtracking with the same arguments -
992 # this method gets repeatedly called during backtracking with the same arguments -
993 # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression
993 # we can cache these arguments and save ourselves the trouble of re-parsing the contained expression
994 def _parseCache( self, instring, loc, doActions=True, callPreParse=True ):
994 def _parseCache( self, instring, loc, doActions=True, callPreParse=True ):
995 lookup = (self,instring,loc,callPreParse,doActions)
995 lookup = (self,instring,loc,callPreParse,doActions)
996 if lookup in ParserElement._exprArgCache:
996 if lookup in ParserElement._exprArgCache:
997 value = ParserElement._exprArgCache[ lookup ]
997 value = ParserElement._exprArgCache[ lookup ]
998 if isinstance(value,Exception):
998 if isinstance(value,Exception):
999 raise value
999 raise value
1000 return value
1000 return value
1001 else:
1001 else:
1002 try:
1002 try:
1003 value = self._parseNoCache( instring, loc, doActions, callPreParse )
1003 value = self._parseNoCache( instring, loc, doActions, callPreParse )
1004 ParserElement._exprArgCache[ lookup ] = (value[0],value[1].copy())
1004 ParserElement._exprArgCache[ lookup ] = (value[0],value[1].copy())
1005 return value
1005 return value
1006 except ParseBaseException, pe:
1006 except ParseBaseException, pe:
1007 ParserElement._exprArgCache[ lookup ] = pe
1007 ParserElement._exprArgCache[ lookup ] = pe
1008 raise
1008 raise
1009
1009
1010 _parse = _parseNoCache
1010 _parse = _parseNoCache
1011
1011
1012 # argument cache for optimizing repeated calls when backtracking through recursive expressions
1012 # argument cache for optimizing repeated calls when backtracking through recursive expressions
1013 _exprArgCache = {}
1013 _exprArgCache = {}
1014 def resetCache():
1014 def resetCache():
1015 ParserElement._exprArgCache.clear()
1015 ParserElement._exprArgCache.clear()
1016 resetCache = staticmethod(resetCache)
1016 resetCache = staticmethod(resetCache)
1017
1017
1018 _packratEnabled = False
1018 _packratEnabled = False
1019 def enablePackrat():
1019 def enablePackrat():
1020 """Enables "packrat" parsing, which adds memoizing to the parsing logic.
1020 """Enables "packrat" parsing, which adds memoizing to the parsing logic.
1021 Repeated parse attempts at the same string location (which happens
1021 Repeated parse attempts at the same string location (which happens
1022 often in many complex grammars) can immediately return a cached value,
1022 often in many complex grammars) can immediately return a cached value,
1023 instead of re-executing parsing/validating code. Memoizing is done of
1023 instead of re-executing parsing/validating code. Memoizing is done of
1024 both valid results and parsing exceptions.
1024 both valid results and parsing exceptions.
1025
1025
1026 This speedup may break existing programs that use parse actions that
1026 This speedup may break existing programs that use parse actions that
1027 have side-effects. For this reason, packrat parsing is disabled when
1027 have side-effects. For this reason, packrat parsing is disabled when
1028 you first import pyparsing. To activate the packrat feature, your
1028 you first import pyparsing. To activate the packrat feature, your
1029 program must call the class method ParserElement.enablePackrat(). If
1029 program must call the class method ParserElement.enablePackrat(). If
1030 your program uses psyco to "compile as you go", you must call
1030 your program uses psyco to "compile as you go", you must call
1031 enablePackrat before calling psyco.full(). If you do not do this,
1031 enablePackrat before calling psyco.full(). If you do not do this,
1032 Python will crash. For best results, call enablePackrat() immediately
1032 Python will crash. For best results, call enablePackrat() immediately
1033 after importing pyparsing.
1033 after importing pyparsing.
1034 """
1034 """
1035 if not ParserElement._packratEnabled:
1035 if not ParserElement._packratEnabled:
1036 ParserElement._packratEnabled = True
1036 ParserElement._packratEnabled = True
1037 ParserElement._parse = ParserElement._parseCache
1037 ParserElement._parse = ParserElement._parseCache
1038 enablePackrat = staticmethod(enablePackrat)
1038 enablePackrat = staticmethod(enablePackrat)
1039
1039
1040 def parseString( self, instring, parseAll=False ):
1040 def parseString( self, instring, parseAll=False ):
1041 """Execute the parse expression with the given string.
1041 """Execute the parse expression with the given string.
1042 This is the main interface to the client code, once the complete
1042 This is the main interface to the client code, once the complete
1043 expression has been built.
1043 expression has been built.
1044
1044
1045 If you want the grammar to require that the entire input string be
1045 If you want the grammar to require that the entire input string be
1046 successfully parsed, then set parseAll to True (equivalent to ending
1046 successfully parsed, then set parseAll to True (equivalent to ending
1047 the grammar with StringEnd()).
1047 the grammar with StringEnd()).
1048
1048
1049 Note: parseString implicitly calls expandtabs() on the input string,
1049 Note: parseString implicitly calls expandtabs() on the input string,
1050 in order to report proper column numbers in parse actions.
1050 in order to report proper column numbers in parse actions.
1051 If the input string contains tabs and
1051 If the input string contains tabs and
1052 the grammar uses parse actions that use the loc argument to index into the
1052 the grammar uses parse actions that use the loc argument to index into the
1053 string being parsed, you can ensure you have a consistent view of the input
1053 string being parsed, you can ensure you have a consistent view of the input
1054 string by:
1054 string by:
1055 - calling parseWithTabs on your grammar before calling parseString
1055 - calling parseWithTabs on your grammar before calling parseString
1056 (see L{I{parseWithTabs}<parseWithTabs>})
1056 (see L{I{parseWithTabs}<parseWithTabs>})
1057 - define your parse action using the full (s,loc,toks) signature, and
1057 - define your parse action using the full (s,loc,toks) signature, and
1058 reference the input string using the parse action's s argument
1058 reference the input string using the parse action's s argument
1059 - explictly expand the tabs in your input string before calling
1059 - explictly expand the tabs in your input string before calling
1060 parseString
1060 parseString
1061 """
1061 """
1062 ParserElement.resetCache()
1062 ParserElement.resetCache()
1063 if not self.streamlined:
1063 if not self.streamlined:
1064 self.streamline()
1064 self.streamline()
1065 #~ self.saveAsList = True
1065 #~ self.saveAsList = True
1066 for e in self.ignoreExprs:
1066 for e in self.ignoreExprs:
1067 e.streamline()
1067 e.streamline()
1068 if not self.keepTabs:
1068 if not self.keepTabs:
1069 instring = instring.expandtabs()
1069 instring = instring.expandtabs()
1070 try:
1070 try:
1071 loc, tokens = self._parse( instring, 0 )
1071 loc, tokens = self._parse( instring, 0 )
1072 if parseAll:
1072 if parseAll:
1073 loc = self.preParse( instring, loc )
1073 loc = self.preParse( instring, loc )
1074 StringEnd()._parse( instring, loc )
1074 StringEnd()._parse( instring, loc )
1075 except ParseBaseException, exc:
1075 except ParseBaseException, exc:
1076 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1076 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1077 raise exc
1077 raise exc
1078 else:
1078 else:
1079 return tokens
1079 return tokens
1080
1080
1081 def scanString( self, instring, maxMatches=_MAX_INT ):
1081 def scanString( self, instring, maxMatches=_MAX_INT ):
1082 """Scan the input string for expression matches. Each match will return the
1082 """Scan the input string for expression matches. Each match will return the
1083 matching tokens, start location, and end location. May be called with optional
1083 matching tokens, start location, and end location. May be called with optional
1084 maxMatches argument, to clip scanning after 'n' matches are found.
1084 maxMatches argument, to clip scanning after 'n' matches are found.
1085
1085
1086 Note that the start and end locations are reported relative to the string
1086 Note that the start and end locations are reported relative to the string
1087 being parsed. See L{I{parseString}<parseString>} for more information on parsing
1087 being parsed. See L{I{parseString}<parseString>} for more information on parsing
1088 strings with embedded tabs."""
1088 strings with embedded tabs."""
1089 if not self.streamlined:
1089 if not self.streamlined:
1090 self.streamline()
1090 self.streamline()
1091 for e in self.ignoreExprs:
1091 for e in self.ignoreExprs:
1092 e.streamline()
1092 e.streamline()
1093
1093
1094 if not self.keepTabs:
1094 if not self.keepTabs:
1095 instring = _ustr(instring).expandtabs()
1095 instring = _ustr(instring).expandtabs()
1096 instrlen = len(instring)
1096 instrlen = len(instring)
1097 loc = 0
1097 loc = 0
1098 preparseFn = self.preParse
1098 preparseFn = self.preParse
1099 parseFn = self._parse
1099 parseFn = self._parse
1100 ParserElement.resetCache()
1100 ParserElement.resetCache()
1101 matches = 0
1101 matches = 0
1102 try:
1102 try:
1103 while loc <= instrlen and matches < maxMatches:
1103 while loc <= instrlen and matches < maxMatches:
1104 try:
1104 try:
1105 preloc = preparseFn( instring, loc )
1105 preloc = preparseFn( instring, loc )
1106 nextLoc,tokens = parseFn( instring, preloc, callPreParse=False )
1106 nextLoc,tokens = parseFn( instring, preloc, callPreParse=False )
1107 except ParseException:
1107 except ParseException:
1108 loc = preloc+1
1108 loc = preloc+1
1109 else:
1109 else:
1110 matches += 1
1110 matches += 1
1111 yield tokens, preloc, nextLoc
1111 yield tokens, preloc, nextLoc
1112 loc = nextLoc
1112 loc = nextLoc
1113 except ParseBaseException, pe:
1113 except ParseBaseException, pe:
1114 raise pe
1114 raise pe
1115
1115
1116 def transformString( self, instring ):
1116 def transformString( self, instring ):
1117 """Extension to scanString, to modify matching text with modified tokens that may
1117 """Extension to scanString, to modify matching text with modified tokens that may
1118 be returned from a parse action. To use transformString, define a grammar and
1118 be returned from a parse action. To use transformString, define a grammar and
1119 attach a parse action to it that modifies the returned token list.
1119 attach a parse action to it that modifies the returned token list.
1120 Invoking transformString() on a target string will then scan for matches,
1120 Invoking transformString() on a target string will then scan for matches,
1121 and replace the matched text patterns according to the logic in the parse
1121 and replace the matched text patterns according to the logic in the parse
1122 action. transformString() returns the resulting transformed string."""
1122 action. transformString() returns the resulting transformed string."""
1123 out = []
1123 out = []
1124 lastE = 0
1124 lastE = 0
1125 # force preservation of <TAB>s, to minimize unwanted transformation of string, and to
1125 # force preservation of <TAB>s, to minimize unwanted transformation of string, and to
1126 # keep string locs straight between transformString and scanString
1126 # keep string locs straight between transformString and scanString
1127 self.keepTabs = True
1127 self.keepTabs = True
1128 try:
1128 try:
1129 for t,s,e in self.scanString( instring ):
1129 for t,s,e in self.scanString( instring ):
1130 out.append( instring[lastE:s] )
1130 out.append( instring[lastE:s] )
1131 if t:
1131 if t:
1132 if isinstance(t,ParseResults):
1132 if isinstance(t,ParseResults):
1133 out += t.asList()
1133 out += t.asList()
1134 elif isinstance(t,list):
1134 elif isinstance(t,list):
1135 out += t
1135 out += t
1136 else:
1136 else:
1137 out.append(t)
1137 out.append(t)
1138 lastE = e
1138 lastE = e
1139 out.append(instring[lastE:])
1139 out.append(instring[lastE:])
1140 return "".join(map(_ustr,out))
1140 return "".join(map(_ustr,out))
1141 except ParseBaseException, pe:
1141 except ParseBaseException, pe:
1142 raise pe
1142 raise pe
1143
1143
1144 def searchString( self, instring, maxMatches=_MAX_INT ):
1144 def searchString( self, instring, maxMatches=_MAX_INT ):
1145 """Another extension to scanString, simplifying the access to the tokens found
1145 """Another extension to scanString, simplifying the access to the tokens found
1146 to match the given parse expression. May be called with optional
1146 to match the given parse expression. May be called with optional
1147 maxMatches argument, to clip searching after 'n' matches are found.
1147 maxMatches argument, to clip searching after 'n' matches are found.
1148 """
1148 """
1149 try:
1149 try:
1150 return ParseResults([ t for t,s,e in self.scanString( instring, maxMatches ) ])
1150 return ParseResults([ t for t,s,e in self.scanString( instring, maxMatches ) ])
1151 except ParseBaseException, pe:
1151 except ParseBaseException, pe:
1152 raise pe
1152 raise pe
1153
1153
1154 def __add__(self, other ):
1154 def __add__(self, other ):
1155 """Implementation of + operator - returns And"""
1155 """Implementation of + operator - returns And"""
1156 if isinstance( other, basestring ):
1156 if isinstance( other, basestring ):
1157 other = Literal( other )
1157 other = Literal( other )
1158 if not isinstance( other, ParserElement ):
1158 if not isinstance( other, ParserElement ):
1159 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1159 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1160 SyntaxWarning, stacklevel=2)
1160 SyntaxWarning, stacklevel=2)
1161 return None
1161 return None
1162 return And( [ self, other ] )
1162 return And( [ self, other ] )
1163
1163
1164 def __radd__(self, other ):
1164 def __radd__(self, other ):
1165 """Implementation of + operator when left operand is not a ParserElement"""
1165 """Implementation of + operator when left operand is not a ParserElement"""
1166 if isinstance( other, basestring ):
1166 if isinstance( other, basestring ):
1167 other = Literal( other )
1167 other = Literal( other )
1168 if not isinstance( other, ParserElement ):
1168 if not isinstance( other, ParserElement ):
1169 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1169 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1170 SyntaxWarning, stacklevel=2)
1170 SyntaxWarning, stacklevel=2)
1171 return None
1171 return None
1172 return other + self
1172 return other + self
1173
1173
1174 def __sub__(self, other):
1174 def __sub__(self, other):
1175 """Implementation of - operator, returns And with error stop"""
1175 """Implementation of - operator, returns And with error stop"""
1176 if isinstance( other, basestring ):
1176 if isinstance( other, basestring ):
1177 other = Literal( other )
1177 other = Literal( other )
1178 if not isinstance( other, ParserElement ):
1178 if not isinstance( other, ParserElement ):
1179 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1179 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1180 SyntaxWarning, stacklevel=2)
1180 SyntaxWarning, stacklevel=2)
1181 return None
1181 return None
1182 return And( [ self, And._ErrorStop(), other ] )
1182 return And( [ self, And._ErrorStop(), other ] )
1183
1183
1184 def __rsub__(self, other ):
1184 def __rsub__(self, other ):
1185 """Implementation of - operator when left operand is not a ParserElement"""
1185 """Implementation of - operator when left operand is not a ParserElement"""
1186 if isinstance( other, basestring ):
1186 if isinstance( other, basestring ):
1187 other = Literal( other )
1187 other = Literal( other )
1188 if not isinstance( other, ParserElement ):
1188 if not isinstance( other, ParserElement ):
1189 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1189 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1190 SyntaxWarning, stacklevel=2)
1190 SyntaxWarning, stacklevel=2)
1191 return None
1191 return None
1192 return other - self
1192 return other - self
1193
1193
1194 def __mul__(self,other):
1194 def __mul__(self,other):
1195 if isinstance(other,int):
1195 if isinstance(other,int):
1196 minElements, optElements = other,0
1196 minElements, optElements = other,0
1197 elif isinstance(other,tuple):
1197 elif isinstance(other,tuple):
1198 other = (other + (None, None))[:2]
1198 other = (other + (None, None))[:2]
1199 if other[0] is None:
1199 if other[0] is None:
1200 other = (0, other[1])
1200 other = (0, other[1])
1201 if isinstance(other[0],int) and other[1] is None:
1201 if isinstance(other[0],int) and other[1] is None:
1202 if other[0] == 0:
1202 if other[0] == 0:
1203 return ZeroOrMore(self)
1203 return ZeroOrMore(self)
1204 if other[0] == 1:
1204 if other[0] == 1:
1205 return OneOrMore(self)
1205 return OneOrMore(self)
1206 else:
1206 else:
1207 return self*other[0] + ZeroOrMore(self)
1207 return self*other[0] + ZeroOrMore(self)
1208 elif isinstance(other[0],int) and isinstance(other[1],int):
1208 elif isinstance(other[0],int) and isinstance(other[1],int):
1209 minElements, optElements = other
1209 minElements, optElements = other
1210 optElements -= minElements
1210 optElements -= minElements
1211 else:
1211 else:
1212 raise TypeError("cannot multiply 'ParserElement' and ('%s','%s') objects", type(other[0]),type(other[1]))
1212 raise TypeError("cannot multiply 'ParserElement' and ('%s','%s') objects", type(other[0]),type(other[1]))
1213 else:
1213 else:
1214 raise TypeError("cannot multiply 'ParserElement' and '%s' objects", type(other))
1214 raise TypeError("cannot multiply 'ParserElement' and '%s' objects", type(other))
1215
1215
1216 if minElements < 0:
1216 if minElements < 0:
1217 raise ValueError("cannot multiply ParserElement by negative value")
1217 raise ValueError("cannot multiply ParserElement by negative value")
1218 if optElements < 0:
1218 if optElements < 0:
1219 raise ValueError("second tuple value must be greater or equal to first tuple value")
1219 raise ValueError("second tuple value must be greater or equal to first tuple value")
1220 if minElements == optElements == 0:
1220 if minElements == optElements == 0:
1221 raise ValueError("cannot multiply ParserElement by 0 or (0,0)")
1221 raise ValueError("cannot multiply ParserElement by 0 or (0,0)")
1222
1222
1223 if (optElements):
1223 if (optElements):
1224 def makeOptionalList(n):
1224 def makeOptionalList(n):
1225 if n>1:
1225 if n>1:
1226 return Optional(self + makeOptionalList(n-1))
1226 return Optional(self + makeOptionalList(n-1))
1227 else:
1227 else:
1228 return Optional(self)
1228 return Optional(self)
1229 if minElements:
1229 if minElements:
1230 if minElements == 1:
1230 if minElements == 1:
1231 ret = self + makeOptionalList(optElements)
1231 ret = self + makeOptionalList(optElements)
1232 else:
1232 else:
1233 ret = And([self]*minElements) + makeOptionalList(optElements)
1233 ret = And([self]*minElements) + makeOptionalList(optElements)
1234 else:
1234 else:
1235 ret = makeOptionalList(optElements)
1235 ret = makeOptionalList(optElements)
1236 else:
1236 else:
1237 if minElements == 1:
1237 if minElements == 1:
1238 ret = self
1238 ret = self
1239 else:
1239 else:
1240 ret = And([self]*minElements)
1240 ret = And([self]*minElements)
1241 return ret
1241 return ret
1242
1242
1243 def __rmul__(self, other):
1243 def __rmul__(self, other):
1244 return self.__mul__(other)
1244 return self.__mul__(other)
1245
1245
1246 def __or__(self, other ):
1246 def __or__(self, other ):
1247 """Implementation of | operator - returns MatchFirst"""
1247 """Implementation of | operator - returns MatchFirst"""
1248 if isinstance( other, basestring ):
1248 if isinstance( other, basestring ):
1249 other = Literal( other )
1249 other = Literal( other )
1250 if not isinstance( other, ParserElement ):
1250 if not isinstance( other, ParserElement ):
1251 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1251 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1252 SyntaxWarning, stacklevel=2)
1252 SyntaxWarning, stacklevel=2)
1253 return None
1253 return None
1254 return MatchFirst( [ self, other ] )
1254 return MatchFirst( [ self, other ] )
1255
1255
1256 def __ror__(self, other ):
1256 def __ror__(self, other ):
1257 """Implementation of | operator when left operand is not a ParserElement"""
1257 """Implementation of | operator when left operand is not a ParserElement"""
1258 if isinstance( other, basestring ):
1258 if isinstance( other, basestring ):
1259 other = Literal( other )
1259 other = Literal( other )
1260 if not isinstance( other, ParserElement ):
1260 if not isinstance( other, ParserElement ):
1261 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1261 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1262 SyntaxWarning, stacklevel=2)
1262 SyntaxWarning, stacklevel=2)
1263 return None
1263 return None
1264 return other | self
1264 return other | self
1265
1265
1266 def __xor__(self, other ):
1266 def __xor__(self, other ):
1267 """Implementation of ^ operator - returns Or"""
1267 """Implementation of ^ operator - returns Or"""
1268 if isinstance( other, basestring ):
1268 if isinstance( other, basestring ):
1269 other = Literal( other )
1269 other = Literal( other )
1270 if not isinstance( other, ParserElement ):
1270 if not isinstance( other, ParserElement ):
1271 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1271 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1272 SyntaxWarning, stacklevel=2)
1272 SyntaxWarning, stacklevel=2)
1273 return None
1273 return None
1274 return Or( [ self, other ] )
1274 return Or( [ self, other ] )
1275
1275
1276 def __rxor__(self, other ):
1276 def __rxor__(self, other ):
1277 """Implementation of ^ operator when left operand is not a ParserElement"""
1277 """Implementation of ^ operator when left operand is not a ParserElement"""
1278 if isinstance( other, basestring ):
1278 if isinstance( other, basestring ):
1279 other = Literal( other )
1279 other = Literal( other )
1280 if not isinstance( other, ParserElement ):
1280 if not isinstance( other, ParserElement ):
1281 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1281 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1282 SyntaxWarning, stacklevel=2)
1282 SyntaxWarning, stacklevel=2)
1283 return None
1283 return None
1284 return other ^ self
1284 return other ^ self
1285
1285
1286 def __and__(self, other ):
1286 def __and__(self, other ):
1287 """Implementation of & operator - returns Each"""
1287 """Implementation of & operator - returns Each"""
1288 if isinstance( other, basestring ):
1288 if isinstance( other, basestring ):
1289 other = Literal( other )
1289 other = Literal( other )
1290 if not isinstance( other, ParserElement ):
1290 if not isinstance( other, ParserElement ):
1291 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1291 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1292 SyntaxWarning, stacklevel=2)
1292 SyntaxWarning, stacklevel=2)
1293 return None
1293 return None
1294 return Each( [ self, other ] )
1294 return Each( [ self, other ] )
1295
1295
1296 def __rand__(self, other ):
1296 def __rand__(self, other ):
1297 """Implementation of & operator when left operand is not a ParserElement"""
1297 """Implementation of & operator when left operand is not a ParserElement"""
1298 if isinstance( other, basestring ):
1298 if isinstance( other, basestring ):
1299 other = Literal( other )
1299 other = Literal( other )
1300 if not isinstance( other, ParserElement ):
1300 if not isinstance( other, ParserElement ):
1301 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1301 warnings.warn("Cannot combine element of type %s with ParserElement" % type(other),
1302 SyntaxWarning, stacklevel=2)
1302 SyntaxWarning, stacklevel=2)
1303 return None
1303 return None
1304 return other & self
1304 return other & self
1305
1305
1306 def __invert__( self ):
1306 def __invert__( self ):
1307 """Implementation of ~ operator - returns NotAny"""
1307 """Implementation of ~ operator - returns NotAny"""
1308 return NotAny( self )
1308 return NotAny( self )
1309
1309
1310 def __call__(self, name):
1310 def __call__(self, name):
1311 """Shortcut for setResultsName, with listAllMatches=default::
1311 """Shortcut for setResultsName, with listAllMatches=default::
1312 userdata = Word(alphas).setResultsName("name") + Word(nums+"-").setResultsName("socsecno")
1312 userdata = Word(alphas).setResultsName("name") + Word(nums+"-").setResultsName("socsecno")
1313 could be written as::
1313 could be written as::
1314 userdata = Word(alphas)("name") + Word(nums+"-")("socsecno")
1314 userdata = Word(alphas)("name") + Word(nums+"-")("socsecno")
1315 """
1315 """
1316 return self.setResultsName(name)
1316 return self.setResultsName(name)
1317
1317
1318 def suppress( self ):
1318 def suppress( self ):
1319 """Suppresses the output of this ParserElement; useful to keep punctuation from
1319 """Suppresses the output of this ParserElement; useful to keep punctuation from
1320 cluttering up returned output.
1320 cluttering up returned output.
1321 """
1321 """
1322 return Suppress( self )
1322 return Suppress( self )
1323
1323
1324 def leaveWhitespace( self ):
1324 def leaveWhitespace( self ):
1325 """Disables the skipping of whitespace before matching the characters in the
1325 """Disables the skipping of whitespace before matching the characters in the
1326 ParserElement's defined pattern. This is normally only used internally by
1326 ParserElement's defined pattern. This is normally only used internally by
1327 the pyparsing module, but may be needed in some whitespace-sensitive grammars.
1327 the pyparsing module, but may be needed in some whitespace-sensitive grammars.
1328 """
1328 """
1329 self.skipWhitespace = False
1329 self.skipWhitespace = False
1330 return self
1330 return self
1331
1331
1332 def setWhitespaceChars( self, chars ):
1332 def setWhitespaceChars( self, chars ):
1333 """Overrides the default whitespace chars
1333 """Overrides the default whitespace chars
1334 """
1334 """
1335 self.skipWhitespace = True
1335 self.skipWhitespace = True
1336 self.whiteChars = chars
1336 self.whiteChars = chars
1337 self.copyDefaultWhiteChars = False
1337 self.copyDefaultWhiteChars = False
1338 return self
1338 return self
1339
1339
1340 def parseWithTabs( self ):
1340 def parseWithTabs( self ):
1341 """Overrides default behavior to expand <TAB>s to spaces before parsing the input string.
1341 """Overrides default behavior to expand <TAB>s to spaces before parsing the input string.
1342 Must be called before parseString when the input grammar contains elements that
1342 Must be called before parseString when the input grammar contains elements that
1343 match <TAB> characters."""
1343 match <TAB> characters."""
1344 self.keepTabs = True
1344 self.keepTabs = True
1345 return self
1345 return self
1346
1346
1347 def ignore( self, other ):
1347 def ignore( self, other ):
1348 """Define expression to be ignored (e.g., comments) while doing pattern
1348 """Define expression to be ignored (e.g., comments) while doing pattern
1349 matching; may be called repeatedly, to define multiple comment or other
1349 matching; may be called repeatedly, to define multiple comment or other
1350 ignorable patterns.
1350 ignorable patterns.
1351 """
1351 """
1352 if isinstance( other, Suppress ):
1352 if isinstance( other, Suppress ):
1353 if other not in self.ignoreExprs:
1353 if other not in self.ignoreExprs:
1354 self.ignoreExprs.append( other )
1354 self.ignoreExprs.append( other )
1355 else:
1355 else:
1356 self.ignoreExprs.append( Suppress( other ) )
1356 self.ignoreExprs.append( Suppress( other ) )
1357 return self
1357 return self
1358
1358
1359 def setDebugActions( self, startAction, successAction, exceptionAction ):
1359 def setDebugActions( self, startAction, successAction, exceptionAction ):
1360 """Enable display of debugging messages while doing pattern matching."""
1360 """Enable display of debugging messages while doing pattern matching."""
1361 self.debugActions = (startAction or _defaultStartDebugAction,
1361 self.debugActions = (startAction or _defaultStartDebugAction,
1362 successAction or _defaultSuccessDebugAction,
1362 successAction or _defaultSuccessDebugAction,
1363 exceptionAction or _defaultExceptionDebugAction)
1363 exceptionAction or _defaultExceptionDebugAction)
1364 self.debug = True
1364 self.debug = True
1365 return self
1365 return self
1366
1366
1367 def setDebug( self, flag=True ):
1367 def setDebug( self, flag=True ):
1368 """Enable display of debugging messages while doing pattern matching.
1368 """Enable display of debugging messages while doing pattern matching.
1369 Set flag to True to enable, False to disable."""
1369 Set flag to True to enable, False to disable."""
1370 if flag:
1370 if flag:
1371 self.setDebugActions( _defaultStartDebugAction, _defaultSuccessDebugAction, _defaultExceptionDebugAction )
1371 self.setDebugActions( _defaultStartDebugAction, _defaultSuccessDebugAction, _defaultExceptionDebugAction )
1372 else:
1372 else:
1373 self.debug = False
1373 self.debug = False
1374 return self
1374 return self
1375
1375
1376 def __str__( self ):
1376 def __str__( self ):
1377 return self.name
1377 return self.name
1378
1378
1379 def __repr__( self ):
1379 def __repr__( self ):
1380 return _ustr(self)
1380 return _ustr(self)
1381
1381
1382 def streamline( self ):
1382 def streamline( self ):
1383 self.streamlined = True
1383 self.streamlined = True
1384 self.strRepr = None
1384 self.strRepr = None
1385 return self
1385 return self
1386
1386
1387 def checkRecursion( self, parseElementList ):
1387 def checkRecursion( self, parseElementList ):
1388 pass
1388 pass
1389
1389
1390 def validate( self, validateTrace=[] ):
1390 def validate( self, validateTrace=[] ):
1391 """Check defined expressions for valid structure, check for infinite recursive definitions."""
1391 """Check defined expressions for valid structure, check for infinite recursive definitions."""
1392 self.checkRecursion( [] )
1392 self.checkRecursion( [] )
1393
1393
1394 def parseFile( self, file_or_filename, parseAll=False ):
1394 def parseFile( self, file_or_filename, parseAll=False ):
1395 """Execute the parse expression on the given file or filename.
1395 """Execute the parse expression on the given file or filename.
1396 If a filename is specified (instead of a file object),
1396 If a filename is specified (instead of a file object),
1397 the entire file is opened, read, and closed before parsing.
1397 the entire file is opened, read, and closed before parsing.
1398 """
1398 """
1399 try:
1399 try:
1400 file_contents = file_or_filename.read()
1400 file_contents = file_or_filename.read()
1401 except AttributeError:
1401 except AttributeError:
1402 f = open(file_or_filename, "rb")
1402 f = open(file_or_filename, "rb")
1403 file_contents = f.read()
1403 file_contents = f.read()
1404 f.close()
1404 f.close()
1405 try:
1405 try:
1406 return self.parseString(file_contents, parseAll)
1406 return self.parseString(file_contents, parseAll)
1407 except ParseBaseException, exc:
1407 except ParseBaseException, exc:
1408 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1408 # catch and re-raise exception from here, clears out pyparsing internal stack trace
1409 raise exc
1409 raise exc
1410
1410
1411 def getException(self):
1411 def getException(self):
1412 return ParseException("",0,self.errmsg,self)
1412 return ParseException("",0,self.errmsg,self)
1413
1413
1414 def __getattr__(self,aname):
1414 def __getattr__(self,aname):
1415 if aname == "myException":
1415 if aname == "myException":
1416 self.myException = ret = self.getException();
1416 self.myException = ret = self.getException();
1417 return ret;
1417 return ret;
1418 else:
1418 else:
1419 raise AttributeError("no such attribute " + aname)
1419 raise AttributeError("no such attribute " + aname)
1420
1420
1421 def __eq__(self,other):
1421 def __eq__(self,other):
1422 if isinstance(other, ParserElement):
1422 if isinstance(other, ParserElement):
1423 return self is other or self.__dict__ == other.__dict__
1423 return self is other or self.__dict__ == other.__dict__
1424 elif isinstance(other, basestring):
1424 elif isinstance(other, basestring):
1425 try:
1425 try:
1426 self.parseString(_ustr(other), parseAll=True)
1426 self.parseString(_ustr(other), parseAll=True)
1427 return True
1427 return True
1428 except ParseBaseException:
1428 except ParseBaseException:
1429 return False
1429 return False
1430 else:
1430 else:
1431 return super(ParserElement,self)==other
1431 return super(ParserElement,self)==other
1432
1432
1433 def __ne__(self,other):
1433 def __ne__(self,other):
1434 return not (self == other)
1434 return not (self == other)
1435
1435
1436 def __hash__(self):
1436 def __hash__(self):
1437 return hash(id(self))
1437 return hash(id(self))
1438
1438
1439 def __req__(self,other):
1439 def __req__(self,other):
1440 return self == other
1440 return self == other
1441
1441
1442 def __rne__(self,other):
1442 def __rne__(self,other):
1443 return not (self == other)
1443 return not (self == other)
1444
1444
1445
1445
1446 class Token(ParserElement):
1446 class Token(ParserElement):
1447 """Abstract ParserElement subclass, for defining atomic matching patterns."""
1447 """Abstract ParserElement subclass, for defining atomic matching patterns."""
1448 def __init__( self ):
1448 def __init__( self ):
1449 super(Token,self).__init__( savelist=False )
1449 super(Token,self).__init__( savelist=False )
1450 #self.myException = ParseException("",0,"",self)
1450 #self.myException = ParseException("",0,"",self)
1451
1451
1452 def setName(self, name):
1452 def setName(self, name):
1453 s = super(Token,self).setName(name)
1453 s = super(Token,self).setName(name)
1454 self.errmsg = "Expected " + self.name
1454 self.errmsg = "Expected " + self.name
1455 #s.myException.msg = self.errmsg
1455 #s.myException.msg = self.errmsg
1456 return s
1456 return s
1457
1457
1458
1458
1459 class Empty(Token):
1459 class Empty(Token):
1460 """An empty token, will always match."""
1460 """An empty token, will always match."""
1461 def __init__( self ):
1461 def __init__( self ):
1462 super(Empty,self).__init__()
1462 super(Empty,self).__init__()
1463 self.name = "Empty"
1463 self.name = "Empty"
1464 self.mayReturnEmpty = True
1464 self.mayReturnEmpty = True
1465 self.mayIndexError = False
1465 self.mayIndexError = False
1466
1466
1467
1467
1468 class NoMatch(Token):
1468 class NoMatch(Token):
1469 """A token that will never match."""
1469 """A token that will never match."""
1470 def __init__( self ):
1470 def __init__( self ):
1471 super(NoMatch,self).__init__()
1471 super(NoMatch,self).__init__()
1472 self.name = "NoMatch"
1472 self.name = "NoMatch"
1473 self.mayReturnEmpty = True
1473 self.mayReturnEmpty = True
1474 self.mayIndexError = False
1474 self.mayIndexError = False
1475 self.errmsg = "Unmatchable token"
1475 self.errmsg = "Unmatchable token"
1476 #self.myException.msg = self.errmsg
1476 #self.myException.msg = self.errmsg
1477
1477
1478 def parseImpl( self, instring, loc, doActions=True ):
1478 def parseImpl( self, instring, loc, doActions=True ):
1479 exc = self.myException
1479 exc = self.myException
1480 exc.loc = loc
1480 exc.loc = loc
1481 exc.pstr = instring
1481 exc.pstr = instring
1482 raise exc
1482 raise exc
1483
1483
1484
1484
1485 class Literal(Token):
1485 class Literal(Token):
1486 """Token to exactly match a specified string."""
1486 """Token to exactly match a specified string."""
1487 def __init__( self, matchString ):
1487 def __init__( self, matchString ):
1488 super(Literal,self).__init__()
1488 super(Literal,self).__init__()
1489 self.match = matchString
1489 self.match = matchString
1490 self.matchLen = len(matchString)
1490 self.matchLen = len(matchString)
1491 try:
1491 try:
1492 self.firstMatchChar = matchString[0]
1492 self.firstMatchChar = matchString[0]
1493 except IndexError:
1493 except IndexError:
1494 warnings.warn("null string passed to Literal; use Empty() instead",
1494 warnings.warn("null string passed to Literal; use Empty() instead",
1495 SyntaxWarning, stacklevel=2)
1495 SyntaxWarning, stacklevel=2)
1496 self.__class__ = Empty
1496 self.__class__ = Empty
1497 self.name = '"%s"' % _ustr(self.match)
1497 self.name = '"%s"' % _ustr(self.match)
1498 self.errmsg = "Expected " + self.name
1498 self.errmsg = "Expected " + self.name
1499 self.mayReturnEmpty = False
1499 self.mayReturnEmpty = False
1500 #self.myException.msg = self.errmsg
1500 #self.myException.msg = self.errmsg
1501 self.mayIndexError = False
1501 self.mayIndexError = False
1502
1502
1503 # Performance tuning: this routine gets called a *lot*
1503 # Performance tuning: this routine gets called a *lot*
1504 # if this is a single character match string and the first character matches,
1504 # if this is a single character match string and the first character matches,
1505 # short-circuit as quickly as possible, and avoid calling startswith
1505 # short-circuit as quickly as possible, and avoid calling startswith
1506 #~ @profile
1506 #~ @profile
1507 def parseImpl( self, instring, loc, doActions=True ):
1507 def parseImpl( self, instring, loc, doActions=True ):
1508 if (instring[loc] == self.firstMatchChar and
1508 if (instring[loc] == self.firstMatchChar and
1509 (self.matchLen==1 or instring.startswith(self.match,loc)) ):
1509 (self.matchLen==1 or instring.startswith(self.match,loc)) ):
1510 return loc+self.matchLen, self.match
1510 return loc+self.matchLen, self.match
1511 #~ raise ParseException( instring, loc, self.errmsg )
1511 #~ raise ParseException( instring, loc, self.errmsg )
1512 exc = self.myException
1512 exc = self.myException
1513 exc.loc = loc
1513 exc.loc = loc
1514 exc.pstr = instring
1514 exc.pstr = instring
1515 raise exc
1515 raise exc
1516 _L = Literal
1516 _L = Literal
1517
1517
1518 class Keyword(Token):
1518 class Keyword(Token):
1519 """Token to exactly match a specified string as a keyword, that is, it must be
1519 """Token to exactly match a specified string as a keyword, that is, it must be
1520 immediately followed by a non-keyword character. Compare with Literal::
1520 immediately followed by a non-keyword character. Compare with Literal::
1521 Literal("if") will match the leading 'if' in 'ifAndOnlyIf'.
1521 Literal("if") will match the leading 'if' in 'ifAndOnlyIf'.
1522 Keyword("if") will not; it will only match the leading 'if in 'if x=1', or 'if(y==2)'
1522 Keyword("if") will not; it will only match the leading 'if in 'if x=1', or 'if(y==2)'
1523 Accepts two optional constructor arguments in addition to the keyword string:
1523 Accepts two optional constructor arguments in addition to the keyword string:
1524 identChars is a string of characters that would be valid identifier characters,
1524 identChars is a string of characters that would be valid identifier characters,
1525 defaulting to all alphanumerics + "_" and "$"; caseless allows case-insensitive
1525 defaulting to all alphanumerics + "_" and "$"; caseless allows case-insensitive
1526 matching, default is False.
1526 matching, default is False.
1527 """
1527 """
1528 DEFAULT_KEYWORD_CHARS = alphanums+"_$"
1528 DEFAULT_KEYWORD_CHARS = alphanums+"_$"
1529
1529
1530 def __init__( self, matchString, identChars=DEFAULT_KEYWORD_CHARS, caseless=False ):
1530 def __init__( self, matchString, identChars=DEFAULT_KEYWORD_CHARS, caseless=False ):
1531 super(Keyword,self).__init__()
1531 super(Keyword,self).__init__()
1532 self.match = matchString
1532 self.match = matchString
1533 self.matchLen = len(matchString)
1533 self.matchLen = len(matchString)
1534 try:
1534 try:
1535 self.firstMatchChar = matchString[0]
1535 self.firstMatchChar = matchString[0]
1536 except IndexError:
1536 except IndexError:
1537 warnings.warn("null string passed to Keyword; use Empty() instead",
1537 warnings.warn("null string passed to Keyword; use Empty() instead",
1538 SyntaxWarning, stacklevel=2)
1538 SyntaxWarning, stacklevel=2)
1539 self.name = '"%s"' % self.match
1539 self.name = '"%s"' % self.match
1540 self.errmsg = "Expected " + self.name
1540 self.errmsg = "Expected " + self.name
1541 self.mayReturnEmpty = False
1541 self.mayReturnEmpty = False
1542 #self.myException.msg = self.errmsg
1542 #self.myException.msg = self.errmsg
1543 self.mayIndexError = False
1543 self.mayIndexError = False
1544 self.caseless = caseless
1544 self.caseless = caseless
1545 if caseless:
1545 if caseless:
1546 self.caselessmatch = matchString.upper()
1546 self.caselessmatch = matchString.upper()
1547 identChars = identChars.upper()
1547 identChars = identChars.upper()
1548 self.identChars = _str2dict(identChars)
1548 self.identChars = _str2dict(identChars)
1549
1549
1550 def parseImpl( self, instring, loc, doActions=True ):
1550 def parseImpl( self, instring, loc, doActions=True ):
1551 if self.caseless:
1551 if self.caseless:
1552 if ( (instring[ loc:loc+self.matchLen ].upper() == self.caselessmatch) and
1552 if ( (instring[ loc:loc+self.matchLen ].upper() == self.caselessmatch) and
1553 (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen].upper() not in self.identChars) and
1553 (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen].upper() not in self.identChars) and
1554 (loc == 0 or instring[loc-1].upper() not in self.identChars) ):
1554 (loc == 0 or instring[loc-1].upper() not in self.identChars) ):
1555 return loc+self.matchLen, self.match
1555 return loc+self.matchLen, self.match
1556 else:
1556 else:
1557 if (instring[loc] == self.firstMatchChar and
1557 if (instring[loc] == self.firstMatchChar and
1558 (self.matchLen==1 or instring.startswith(self.match,loc)) and
1558 (self.matchLen==1 or instring.startswith(self.match,loc)) and
1559 (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen] not in self.identChars) and
1559 (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen] not in self.identChars) and
1560 (loc == 0 or instring[loc-1] not in self.identChars) ):
1560 (loc == 0 or instring[loc-1] not in self.identChars) ):
1561 return loc+self.matchLen, self.match
1561 return loc+self.matchLen, self.match
1562 #~ raise ParseException( instring, loc, self.errmsg )
1562 #~ raise ParseException( instring, loc, self.errmsg )
1563 exc = self.myException
1563 exc = self.myException
1564 exc.loc = loc
1564 exc.loc = loc
1565 exc.pstr = instring
1565 exc.pstr = instring
1566 raise exc
1566 raise exc
1567
1567
1568 def copy(self):
1568 def copy(self):
1569 c = super(Keyword,self).copy()
1569 c = super(Keyword,self).copy()
1570 c.identChars = Keyword.DEFAULT_KEYWORD_CHARS
1570 c.identChars = Keyword.DEFAULT_KEYWORD_CHARS
1571 return c
1571 return c
1572
1572
1573 def setDefaultKeywordChars( chars ):
1573 def setDefaultKeywordChars( chars ):
1574 """Overrides the default Keyword chars
1574 """Overrides the default Keyword chars
1575 """
1575 """
1576 Keyword.DEFAULT_KEYWORD_CHARS = chars
1576 Keyword.DEFAULT_KEYWORD_CHARS = chars
1577 setDefaultKeywordChars = staticmethod(setDefaultKeywordChars)
1577 setDefaultKeywordChars = staticmethod(setDefaultKeywordChars)
1578
1578
1579 class CaselessLiteral(Literal):
1579 class CaselessLiteral(Literal):
1580 """Token to match a specified string, ignoring case of letters.
1580 """Token to match a specified string, ignoring case of letters.
1581 Note: the matched results will always be in the case of the given
1581 Note: the matched results will always be in the case of the given
1582 match string, NOT the case of the input text.
1582 match string, NOT the case of the input text.
1583 """
1583 """
1584 def __init__( self, matchString ):
1584 def __init__( self, matchString ):
1585 super(CaselessLiteral,self).__init__( matchString.upper() )
1585 super(CaselessLiteral,self).__init__( matchString.upper() )
1586 # Preserve the defining literal.
1586 # Preserve the defining literal.
1587 self.returnString = matchString
1587 self.returnString = matchString
1588 self.name = "'%s'" % self.returnString
1588 self.name = "'%s'" % self.returnString
1589 self.errmsg = "Expected " + self.name
1589 self.errmsg = "Expected " + self.name
1590 #self.myException.msg = self.errmsg
1590 #self.myException.msg = self.errmsg
1591
1591
1592 def parseImpl( self, instring, loc, doActions=True ):
1592 def parseImpl( self, instring, loc, doActions=True ):
1593 if instring[ loc:loc+self.matchLen ].upper() == self.match:
1593 if instring[ loc:loc+self.matchLen ].upper() == self.match:
1594 return loc+self.matchLen, self.returnString
1594 return loc+self.matchLen, self.returnString
1595 #~ raise ParseException( instring, loc, self.errmsg )
1595 #~ raise ParseException( instring, loc, self.errmsg )
1596 exc = self.myException
1596 exc = self.myException
1597 exc.loc = loc
1597 exc.loc = loc
1598 exc.pstr = instring
1598 exc.pstr = instring
1599 raise exc
1599 raise exc
1600
1600
1601 class CaselessKeyword(Keyword):
1601 class CaselessKeyword(Keyword):
1602 def __init__( self, matchString, identChars=Keyword.DEFAULT_KEYWORD_CHARS ):
1602 def __init__( self, matchString, identChars=Keyword.DEFAULT_KEYWORD_CHARS ):
1603 super(CaselessKeyword,self).__init__( matchString, identChars, caseless=True )
1603 super(CaselessKeyword,self).__init__( matchString, identChars, caseless=True )
1604
1604
1605 def parseImpl( self, instring, loc, doActions=True ):
1605 def parseImpl( self, instring, loc, doActions=True ):
1606 if ( (instring[ loc:loc+self.matchLen ].upper() == self.caselessmatch) and
1606 if ( (instring[ loc:loc+self.matchLen ].upper() == self.caselessmatch) and
1607 (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen].upper() not in self.identChars) ):
1607 (loc >= len(instring)-self.matchLen or instring[loc+self.matchLen].upper() not in self.identChars) ):
1608 return loc+self.matchLen, self.match
1608 return loc+self.matchLen, self.match
1609 #~ raise ParseException( instring, loc, self.errmsg )
1609 #~ raise ParseException( instring, loc, self.errmsg )
1610 exc = self.myException
1610 exc = self.myException
1611 exc.loc = loc
1611 exc.loc = loc
1612 exc.pstr = instring
1612 exc.pstr = instring
1613 raise exc
1613 raise exc
1614
1614
1615 class Word(Token):
1615 class Word(Token):
1616 """Token for matching words composed of allowed character sets.
1616 """Token for matching words composed of allowed character sets.
1617 Defined with string containing all allowed initial characters,
1617 Defined with string containing all allowed initial characters,
1618 an optional string containing allowed body characters (if omitted,
1618 an optional string containing allowed body characters (if omitted,
1619 defaults to the initial character set), and an optional minimum,
1619 defaults to the initial character set), and an optional minimum,
1620 maximum, and/or exact length. The default value for min is 1 (a
1620 maximum, and/or exact length. The default value for min is 1 (a
1621 minimum value < 1 is not valid); the default values for max and exact
1621 minimum value < 1 is not valid); the default values for max and exact
1622 are 0, meaning no maximum or exact length restriction.
1622 are 0, meaning no maximum or exact length restriction.
1623 """
1623 """
1624 def __init__( self, initChars, bodyChars=None, min=1, max=0, exact=0, asKeyword=False ):
1624 def __init__( self, initChars, bodyChars=None, min=1, max=0, exact=0, asKeyword=False ):
1625 super(Word,self).__init__()
1625 super(Word,self).__init__()
1626 self.initCharsOrig = initChars
1626 self.initCharsOrig = initChars
1627 self.initChars = _str2dict(initChars)
1627 self.initChars = _str2dict(initChars)
1628 if bodyChars :
1628 if bodyChars :
1629 self.bodyCharsOrig = bodyChars
1629 self.bodyCharsOrig = bodyChars
1630 self.bodyChars = _str2dict(bodyChars)
1630 self.bodyChars = _str2dict(bodyChars)
1631 else:
1631 else:
1632 self.bodyCharsOrig = initChars
1632 self.bodyCharsOrig = initChars
1633 self.bodyChars = _str2dict(initChars)
1633 self.bodyChars = _str2dict(initChars)
1634
1634
1635 self.maxSpecified = max > 0
1635 self.maxSpecified = max > 0
1636
1636
1637 if min < 1:
1637 if min < 1:
1638 raise ValueError("cannot specify a minimum length < 1; use Optional(Word()) if zero-length word is permitted")
1638 raise ValueError("cannot specify a minimum length < 1; use Optional(Word()) if zero-length word is permitted")
1639
1639
1640 self.minLen = min
1640 self.minLen = min
1641
1641
1642 if max > 0:
1642 if max > 0:
1643 self.maxLen = max
1643 self.maxLen = max
1644 else:
1644 else:
1645 self.maxLen = _MAX_INT
1645 self.maxLen = _MAX_INT
1646
1646
1647 if exact > 0:
1647 if exact > 0:
1648 self.maxLen = exact
1648 self.maxLen = exact
1649 self.minLen = exact
1649 self.minLen = exact
1650
1650
1651 self.name = _ustr(self)
1651 self.name = _ustr(self)
1652 self.errmsg = "Expected " + self.name
1652 self.errmsg = "Expected " + self.name
1653 #self.myException.msg = self.errmsg
1653 #self.myException.msg = self.errmsg
1654 self.mayIndexError = False
1654 self.mayIndexError = False
1655 self.asKeyword = asKeyword
1655 self.asKeyword = asKeyword
1656
1656
1657 if ' ' not in self.initCharsOrig+self.bodyCharsOrig and (min==1 and max==0 and exact==0):
1657 if ' ' not in self.initCharsOrig+self.bodyCharsOrig and (min==1 and max==0 and exact==0):
1658 if self.bodyCharsOrig == self.initCharsOrig:
1658 if self.bodyCharsOrig == self.initCharsOrig:
1659 self.reString = "[%s]+" % _escapeRegexRangeChars(self.initCharsOrig)
1659 self.reString = "[%s]+" % _escapeRegexRangeChars(self.initCharsOrig)
1660 elif len(self.bodyCharsOrig) == 1:
1660 elif len(self.bodyCharsOrig) == 1:
1661 self.reString = "%s[%s]*" % \
1661 self.reString = "%s[%s]*" % \
1662 (re.escape(self.initCharsOrig),
1662 (re.escape(self.initCharsOrig),
1663 _escapeRegexRangeChars(self.bodyCharsOrig),)
1663 _escapeRegexRangeChars(self.bodyCharsOrig),)
1664 else:
1664 else:
1665 self.reString = "[%s][%s]*" % \
1665 self.reString = "[%s][%s]*" % \
1666 (_escapeRegexRangeChars(self.initCharsOrig),
1666 (_escapeRegexRangeChars(self.initCharsOrig),
1667 _escapeRegexRangeChars(self.bodyCharsOrig),)
1667 _escapeRegexRangeChars(self.bodyCharsOrig),)
1668 if self.asKeyword:
1668 if self.asKeyword:
1669 self.reString = r"\b"+self.reString+r"\b"
1669 self.reString = r"\b"+self.reString+r"\b"
1670 try:
1670 try:
1671 self.re = re.compile( self.reString )
1671 self.re = re.compile( self.reString )
1672 except:
1672 except:
1673 self.re = None
1673 self.re = None
1674
1674
1675 def parseImpl( self, instring, loc, doActions=True ):
1675 def parseImpl( self, instring, loc, doActions=True ):
1676 if self.re:
1676 if self.re:
1677 result = self.re.match(instring,loc)
1677 result = self.re.match(instring,loc)
1678 if not result:
1678 if not result:
1679 exc = self.myException
1679 exc = self.myException
1680 exc.loc = loc
1680 exc.loc = loc
1681 exc.pstr = instring
1681 exc.pstr = instring
1682 raise exc
1682 raise exc
1683
1683
1684 loc = result.end()
1684 loc = result.end()
1685 return loc,result.group()
1685 return loc,result.group()
1686
1686
1687 if not(instring[ loc ] in self.initChars):
1687 if not(instring[ loc ] in self.initChars):
1688 #~ raise ParseException( instring, loc, self.errmsg )
1688 #~ raise ParseException( instring, loc, self.errmsg )
1689 exc = self.myException
1689 exc = self.myException
1690 exc.loc = loc
1690 exc.loc = loc
1691 exc.pstr = instring
1691 exc.pstr = instring
1692 raise exc
1692 raise exc
1693 start = loc
1693 start = loc
1694 loc += 1
1694 loc += 1
1695 instrlen = len(instring)
1695 instrlen = len(instring)
1696 bodychars = self.bodyChars
1696 bodychars = self.bodyChars
1697 maxloc = start + self.maxLen
1697 maxloc = start + self.maxLen
1698 maxloc = min( maxloc, instrlen )
1698 maxloc = min( maxloc, instrlen )
1699 while loc < maxloc and instring[loc] in bodychars:
1699 while loc < maxloc and instring[loc] in bodychars:
1700 loc += 1
1700 loc += 1
1701
1701
1702 throwException = False
1702 throwException = False
1703 if loc - start < self.minLen:
1703 if loc - start < self.minLen:
1704 throwException = True
1704 throwException = True
1705 if self.maxSpecified and loc < instrlen and instring[loc] in bodychars:
1705 if self.maxSpecified and loc < instrlen and instring[loc] in bodychars:
1706 throwException = True
1706 throwException = True
1707 if self.asKeyword:
1707 if self.asKeyword:
1708 if (start>0 and instring[start-1] in bodychars) or (loc<instrlen and instring[loc] in bodychars):
1708 if (start>0 and instring[start-1] in bodychars) or (loc<instrlen and instring[loc] in bodychars):
1709 throwException = True
1709 throwException = True
1710
1710
1711 if throwException:
1711 if throwException:
1712 #~ raise ParseException( instring, loc, self.errmsg )
1712 #~ raise ParseException( instring, loc, self.errmsg )
1713 exc = self.myException
1713 exc = self.myException
1714 exc.loc = loc
1714 exc.loc = loc
1715 exc.pstr = instring
1715 exc.pstr = instring
1716 raise exc
1716 raise exc
1717
1717
1718 return loc, instring[start:loc]
1718 return loc, instring[start:loc]
1719
1719
1720 def __str__( self ):
1720 def __str__( self ):
1721 try:
1721 try:
1722 return super(Word,self).__str__()
1722 return super(Word,self).__str__()
1723 except:
1723 except:
1724 pass
1724 pass
1725
1725
1726
1726
1727 if self.strRepr is None:
1727 if self.strRepr is None:
1728
1728
1729 def charsAsStr(s):
1729 def charsAsStr(s):
1730 if len(s)>4:
1730 if len(s)>4:
1731 return s[:4]+"..."
1731 return s[:4]+"..."
1732 else:
1732 else:
1733 return s
1733 return s
1734
1734
1735 if ( self.initCharsOrig != self.bodyCharsOrig ):
1735 if ( self.initCharsOrig != self.bodyCharsOrig ):
1736 self.strRepr = "W:(%s,%s)" % ( charsAsStr(self.initCharsOrig), charsAsStr(self.bodyCharsOrig) )
1736 self.strRepr = "W:(%s,%s)" % ( charsAsStr(self.initCharsOrig), charsAsStr(self.bodyCharsOrig) )
1737 else:
1737 else:
1738 self.strRepr = "W:(%s)" % charsAsStr(self.initCharsOrig)
1738 self.strRepr = "W:(%s)" % charsAsStr(self.initCharsOrig)
1739
1739
1740 return self.strRepr
1740 return self.strRepr
1741
1741
1742
1742
1743 class Regex(Token):
1743 class Regex(Token):
1744 """Token for matching strings that match a given regular expression.
1744 """Token for matching strings that match a given regular expression.
1745 Defined with string specifying the regular expression in a form recognized by the inbuilt Python re module.
1745 Defined with string specifying the regular expression in a form recognized by the inbuilt Python re module.
1746 """
1746 """
1747 def __init__( self, pattern, flags=0):
1747 def __init__( self, pattern, flags=0):
1748 """The parameters pattern and flags are passed to the re.compile() function as-is. See the Python re module for an explanation of the acceptable patterns and flags."""
1748 """The parameters pattern and flags are passed to the re.compile() function as-is. See the Python re module for an explanation of the acceptable patterns and flags."""
1749 super(Regex,self).__init__()
1749 super(Regex,self).__init__()
1750
1750
1751 if len(pattern) == 0:
1751 if len(pattern) == 0:
1752 warnings.warn("null string passed to Regex; use Empty() instead",
1752 warnings.warn("null string passed to Regex; use Empty() instead",
1753 SyntaxWarning, stacklevel=2)
1753 SyntaxWarning, stacklevel=2)
1754
1754
1755 self.pattern = pattern
1755 self.pattern = pattern
1756 self.flags = flags
1756 self.flags = flags
1757
1757
1758 try:
1758 try:
1759 self.re = re.compile(self.pattern, self.flags)
1759 self.re = re.compile(self.pattern, self.flags)
1760 self.reString = self.pattern
1760 self.reString = self.pattern
1761 except sre_constants.error:
1761 except sre_constants.error:
1762 warnings.warn("invalid pattern (%s) passed to Regex" % pattern,
1762 warnings.warn("invalid pattern (%s) passed to Regex" % pattern,
1763 SyntaxWarning, stacklevel=2)
1763 SyntaxWarning, stacklevel=2)
1764 raise
1764 raise
1765
1765
1766 self.name = _ustr(self)
1766 self.name = _ustr(self)
1767 self.errmsg = "Expected " + self.name
1767 self.errmsg = "Expected " + self.name
1768 #self.myException.msg = self.errmsg
1768 #self.myException.msg = self.errmsg
1769 self.mayIndexError = False
1769 self.mayIndexError = False
1770 self.mayReturnEmpty = True
1770 self.mayReturnEmpty = True
1771
1771
1772 def parseImpl( self, instring, loc, doActions=True ):
1772 def parseImpl( self, instring, loc, doActions=True ):
1773 result = self.re.match(instring,loc)
1773 result = self.re.match(instring,loc)
1774 if not result:
1774 if not result:
1775 exc = self.myException
1775 exc = self.myException
1776 exc.loc = loc
1776 exc.loc = loc
1777 exc.pstr = instring
1777 exc.pstr = instring
1778 raise exc
1778 raise exc
1779
1779
1780 loc = result.end()
1780 loc = result.end()
1781 d = result.groupdict()
1781 d = result.groupdict()
1782 ret = ParseResults(result.group())
1782 ret = ParseResults(result.group())
1783 if d:
1783 if d:
1784 for k in d:
1784 for k in d:
1785 ret[k] = d[k]
1785 ret[k] = d[k]
1786 return loc,ret
1786 return loc,ret
1787
1787
1788 def __str__( self ):
1788 def __str__( self ):
1789 try:
1789 try:
1790 return super(Regex,self).__str__()
1790 return super(Regex,self).__str__()
1791 except:
1791 except:
1792 pass
1792 pass
1793
1793
1794 if self.strRepr is None:
1794 if self.strRepr is None:
1795 self.strRepr = "Re:(%s)" % repr(self.pattern)
1795 self.strRepr = "Re:(%s)" % repr(self.pattern)
1796
1796
1797 return self.strRepr
1797 return self.strRepr
1798
1798
1799
1799
1800 class QuotedString(Token):
1800 class QuotedString(Token):
1801 """Token for matching strings that are delimited by quoting characters.
1801 """Token for matching strings that are delimited by quoting characters.
1802 """
1802 """
1803 def __init__( self, quoteChar, escChar=None, escQuote=None, multiline=False, unquoteResults=True, endQuoteChar=None):
1803 def __init__( self, quoteChar, escChar=None, escQuote=None, multiline=False, unquoteResults=True, endQuoteChar=None):
1804 """
1804 """
1805 Defined with the following parameters:
1805 Defined with the following parameters:
1806 - quoteChar - string of one or more characters defining the quote delimiting string
1806 - quoteChar - string of one or more characters defining the quote delimiting string
1807 - escChar - character to escape quotes, typically backslash (default=None)
1807 - escChar - character to escape quotes, typically backslash (default=None)
1808 - escQuote - special quote sequence to escape an embedded quote string (such as SQL's "" to escape an embedded ") (default=None)
1808 - escQuote - special quote sequence to escape an embedded quote string (such as SQL's "" to escape an embedded ") (default=None)
1809 - multiline - boolean indicating whether quotes can span multiple lines (default=False)
1809 - multiline - boolean indicating whether quotes can span multiple lines (default=False)
1810 - unquoteResults - boolean indicating whether the matched text should be unquoted (default=True)
1810 - unquoteResults - boolean indicating whether the matched text should be unquoted (default=True)
1811 - endQuoteChar - string of one or more characters defining the end of the quote delimited string (default=None => same as quoteChar)
1811 - endQuoteChar - string of one or more characters defining the end of the quote delimited string (default=None => same as quoteChar)
1812 """
1812 """
1813 super(QuotedString,self).__init__()
1813 super(QuotedString,self).__init__()
1814
1814
1815 # remove white space from quote chars - wont work anyway
1815 # remove white space from quote chars - wont work anyway
1816 quoteChar = quoteChar.strip()
1816 quoteChar = quoteChar.strip()
1817 if len(quoteChar) == 0:
1817 if len(quoteChar) == 0:
1818 warnings.warn("quoteChar cannot be the empty string",SyntaxWarning,stacklevel=2)
1818 warnings.warn("quoteChar cannot be the empty string",SyntaxWarning,stacklevel=2)
1819 raise SyntaxError()
1819 raise SyntaxError()
1820
1820
1821 if endQuoteChar is None:
1821 if endQuoteChar is None:
1822 endQuoteChar = quoteChar
1822 endQuoteChar = quoteChar
1823 else:
1823 else:
1824 endQuoteChar = endQuoteChar.strip()
1824 endQuoteChar = endQuoteChar.strip()
1825 if len(endQuoteChar) == 0:
1825 if len(endQuoteChar) == 0:
1826 warnings.warn("endQuoteChar cannot be the empty string",SyntaxWarning,stacklevel=2)
1826 warnings.warn("endQuoteChar cannot be the empty string",SyntaxWarning,stacklevel=2)
1827 raise SyntaxError()
1827 raise SyntaxError()
1828
1828
1829 self.quoteChar = quoteChar
1829 self.quoteChar = quoteChar
1830 self.quoteCharLen = len(quoteChar)
1830 self.quoteCharLen = len(quoteChar)
1831 self.firstQuoteChar = quoteChar[0]
1831 self.firstQuoteChar = quoteChar[0]
1832 self.endQuoteChar = endQuoteChar
1832 self.endQuoteChar = endQuoteChar
1833 self.endQuoteCharLen = len(endQuoteChar)
1833 self.endQuoteCharLen = len(endQuoteChar)
1834 self.escChar = escChar
1834 self.escChar = escChar
1835 self.escQuote = escQuote
1835 self.escQuote = escQuote
1836 self.unquoteResults = unquoteResults
1836 self.unquoteResults = unquoteResults
1837
1837
1838 if multiline:
1838 if multiline:
1839 self.flags = re.MULTILINE | re.DOTALL
1839 self.flags = re.MULTILINE | re.DOTALL
1840 self.pattern = r'%s(?:[^%s%s]' % \
1840 self.pattern = r'%s(?:[^%s%s]' % \
1841 ( re.escape(self.quoteChar),
1841 ( re.escape(self.quoteChar),
1842 _escapeRegexRangeChars(self.endQuoteChar[0]),
1842 _escapeRegexRangeChars(self.endQuoteChar[0]),
1843 (escChar is not None and _escapeRegexRangeChars(escChar) or '') )
1843 (escChar is not None and _escapeRegexRangeChars(escChar) or '') )
1844 else:
1844 else:
1845 self.flags = 0
1845 self.flags = 0
1846 self.pattern = r'%s(?:[^%s\n\r%s]' % \
1846 self.pattern = r'%s(?:[^%s\n\r%s]' % \
1847 ( re.escape(self.quoteChar),
1847 ( re.escape(self.quoteChar),
1848 _escapeRegexRangeChars(self.endQuoteChar[0]),
1848 _escapeRegexRangeChars(self.endQuoteChar[0]),
1849 (escChar is not None and _escapeRegexRangeChars(escChar) or '') )
1849 (escChar is not None and _escapeRegexRangeChars(escChar) or '') )
1850 if len(self.endQuoteChar) > 1:
1850 if len(self.endQuoteChar) > 1:
1851 self.pattern += (
1851 self.pattern += (
1852 '|(?:' + ')|(?:'.join(["%s[^%s]" % (re.escape(self.endQuoteChar[:i]),
1852 '|(?:' + ')|(?:'.join(["%s[^%s]" % (re.escape(self.endQuoteChar[:i]),
1853 _escapeRegexRangeChars(self.endQuoteChar[i]))
1853 _escapeRegexRangeChars(self.endQuoteChar[i]))
1854 for i in range(len(self.endQuoteChar)-1,0,-1)]) + ')'
1854 for i in range(len(self.endQuoteChar)-1,0,-1)]) + ')'
1855 )
1855 )
1856 if escQuote:
1856 if escQuote:
1857 self.pattern += (r'|(?:%s)' % re.escape(escQuote))
1857 self.pattern += (r'|(?:%s)' % re.escape(escQuote))
1858 if escChar:
1858 if escChar:
1859 self.pattern += (r'|(?:%s.)' % re.escape(escChar))
1859 self.pattern += (r'|(?:%s.)' % re.escape(escChar))
1860 self.escCharReplacePattern = re.escape(self.escChar)+"(.)"
1860 self.escCharReplacePattern = re.escape(self.escChar)+"(.)"
1861 self.pattern += (r')*%s' % re.escape(self.endQuoteChar))
1861 self.pattern += (r')*%s' % re.escape(self.endQuoteChar))
1862
1862
1863 try:
1863 try:
1864 self.re = re.compile(self.pattern, self.flags)
1864 self.re = re.compile(self.pattern, self.flags)
1865 self.reString = self.pattern
1865 self.reString = self.pattern
1866 except sre_constants.error:
1866 except sre_constants.error:
1867 warnings.warn("invalid pattern (%s) passed to Regex" % self.pattern,
1867 warnings.warn("invalid pattern (%s) passed to Regex" % self.pattern,
1868 SyntaxWarning, stacklevel=2)
1868 SyntaxWarning, stacklevel=2)
1869 raise
1869 raise
1870
1870
1871 self.name = _ustr(self)
1871 self.name = _ustr(self)
1872 self.errmsg = "Expected " + self.name
1872 self.errmsg = "Expected " + self.name
1873 #self.myException.msg = self.errmsg
1873 #self.myException.msg = self.errmsg
1874 self.mayIndexError = False
1874 self.mayIndexError = False
1875 self.mayReturnEmpty = True
1875 self.mayReturnEmpty = True
1876
1876
1877 def parseImpl( self, instring, loc, doActions=True ):
1877 def parseImpl( self, instring, loc, doActions=True ):
1878 result = instring[loc] == self.firstQuoteChar and self.re.match(instring,loc) or None
1878 result = instring[loc] == self.firstQuoteChar and self.re.match(instring,loc) or None
1879 if not result:
1879 if not result:
1880 exc = self.myException
1880 exc = self.myException
1881 exc.loc = loc
1881 exc.loc = loc
1882 exc.pstr = instring
1882 exc.pstr = instring
1883 raise exc
1883 raise exc
1884
1884
1885 loc = result.end()
1885 loc = result.end()
1886 ret = result.group()
1886 ret = result.group()
1887
1887
1888 if self.unquoteResults:
1888 if self.unquoteResults:
1889
1889
1890 # strip off quotes
1890 # strip off quotes
1891 ret = ret[self.quoteCharLen:-self.endQuoteCharLen]
1891 ret = ret[self.quoteCharLen:-self.endQuoteCharLen]
1892
1892
1893 if isinstance(ret,basestring):
1893 if isinstance(ret,basestring):
1894 # replace escaped characters
1894 # replace escaped characters
1895 if self.escChar:
1895 if self.escChar:
1896 ret = re.sub(self.escCharReplacePattern,"\g<1>",ret)
1896 ret = re.sub(self.escCharReplacePattern,"\g<1>",ret)
1897
1897
1898 # replace escaped quotes
1898 # replace escaped quotes
1899 if self.escQuote:
1899 if self.escQuote:
1900 ret = ret.replace(self.escQuote, self.endQuoteChar)
1900 ret = ret.replace(self.escQuote, self.endQuoteChar)
1901
1901
1902 return loc, ret
1902 return loc, ret
1903
1903
1904 def __str__( self ):
1904 def __str__( self ):
1905 try:
1905 try:
1906 return super(QuotedString,self).__str__()
1906 return super(QuotedString,self).__str__()
1907 except:
1907 except:
1908 pass
1908 pass
1909
1909
1910 if self.strRepr is None:
1910 if self.strRepr is None:
1911 self.strRepr = "quoted string, starting with %s ending with %s" % (self.quoteChar, self.endQuoteChar)
1911 self.strRepr = "quoted string, starting with %s ending with %s" % (self.quoteChar, self.endQuoteChar)
1912
1912
1913 return self.strRepr
1913 return self.strRepr
1914
1914
1915
1915
1916 class CharsNotIn(Token):
1916 class CharsNotIn(Token):
1917 """Token for matching words composed of characters *not* in a given set.
1917 """Token for matching words composed of characters *not* in a given set.
1918 Defined with string containing all disallowed characters, and an optional
1918 Defined with string containing all disallowed characters, and an optional
1919 minimum, maximum, and/or exact length. The default value for min is 1 (a
1919 minimum, maximum, and/or exact length. The default value for min is 1 (a
1920 minimum value < 1 is not valid); the default values for max and exact
1920 minimum value < 1 is not valid); the default values for max and exact
1921 are 0, meaning no maximum or exact length restriction.
1921 are 0, meaning no maximum or exact length restriction.
1922 """
1922 """
1923 def __init__( self, notChars, min=1, max=0, exact=0 ):
1923 def __init__( self, notChars, min=1, max=0, exact=0 ):
1924 super(CharsNotIn,self).__init__()
1924 super(CharsNotIn,self).__init__()
1925 self.skipWhitespace = False
1925 self.skipWhitespace = False
1926 self.notChars = notChars
1926 self.notChars = notChars
1927
1927
1928 if min < 1:
1928 if min < 1:
1929 raise ValueError("cannot specify a minimum length < 1; use Optional(CharsNotIn()) if zero-length char group is permitted")
1929 raise ValueError("cannot specify a minimum length < 1; use Optional(CharsNotIn()) if zero-length char group is permitted")
1930
1930
1931 self.minLen = min
1931 self.minLen = min
1932
1932
1933 if max > 0:
1933 if max > 0:
1934 self.maxLen = max
1934 self.maxLen = max
1935 else:
1935 else:
1936 self.maxLen = _MAX_INT
1936 self.maxLen = _MAX_INT
1937
1937
1938 if exact > 0:
1938 if exact > 0:
1939 self.maxLen = exact
1939 self.maxLen = exact
1940 self.minLen = exact
1940 self.minLen = exact
1941
1941
1942 self.name = _ustr(self)
1942 self.name = _ustr(self)
1943 self.errmsg = "Expected " + self.name
1943 self.errmsg = "Expected " + self.name
1944 self.mayReturnEmpty = ( self.minLen == 0 )
1944 self.mayReturnEmpty = ( self.minLen == 0 )
1945 #self.myException.msg = self.errmsg
1945 #self.myException.msg = self.errmsg
1946 self.mayIndexError = False
1946 self.mayIndexError = False
1947
1947
1948 def parseImpl( self, instring, loc, doActions=True ):
1948 def parseImpl( self, instring, loc, doActions=True ):
1949 if instring[loc] in self.notChars:
1949 if instring[loc] in self.notChars:
1950 #~ raise ParseException( instring, loc, self.errmsg )
1950 #~ raise ParseException( instring, loc, self.errmsg )
1951 exc = self.myException
1951 exc = self.myException
1952 exc.loc = loc
1952 exc.loc = loc
1953 exc.pstr = instring
1953 exc.pstr = instring
1954 raise exc
1954 raise exc
1955
1955
1956 start = loc
1956 start = loc
1957 loc += 1
1957 loc += 1
1958 notchars = self.notChars
1958 notchars = self.notChars
1959 maxlen = min( start+self.maxLen, len(instring) )
1959 maxlen = min( start+self.maxLen, len(instring) )
1960 while loc < maxlen and \
1960 while loc < maxlen and \
1961 (instring[loc] not in notchars):
1961 (instring[loc] not in notchars):
1962 loc += 1
1962 loc += 1
1963
1963
1964 if loc - start < self.minLen:
1964 if loc - start < self.minLen:
1965 #~ raise ParseException( instring, loc, self.errmsg )
1965 #~ raise ParseException( instring, loc, self.errmsg )
1966 exc = self.myException
1966 exc = self.myException
1967 exc.loc = loc
1967 exc.loc = loc
1968 exc.pstr = instring
1968 exc.pstr = instring
1969 raise exc
1969 raise exc
1970
1970
1971 return loc, instring[start:loc]
1971 return loc, instring[start:loc]
1972
1972
1973 def __str__( self ):
1973 def __str__( self ):
1974 try:
1974 try:
1975 return super(CharsNotIn, self).__str__()
1975 return super(CharsNotIn, self).__str__()
1976 except:
1976 except:
1977 pass
1977 pass
1978
1978
1979 if self.strRepr is None:
1979 if self.strRepr is None:
1980 if len(self.notChars) > 4:
1980 if len(self.notChars) > 4:
1981 self.strRepr = "!W:(%s...)" % self.notChars[:4]
1981 self.strRepr = "!W:(%s...)" % self.notChars[:4]
1982 else:
1982 else:
1983 self.strRepr = "!W:(%s)" % self.notChars
1983 self.strRepr = "!W:(%s)" % self.notChars
1984
1984
1985 return self.strRepr
1985 return self.strRepr
1986
1986
1987 class White(Token):
1987 class White(Token):
1988 """Special matching class for matching whitespace. Normally, whitespace is ignored
1988 """Special matching class for matching whitespace. Normally, whitespace is ignored
1989 by pyparsing grammars. This class is included when some whitespace structures
1989 by pyparsing grammars. This class is included when some whitespace structures
1990 are significant. Define with a string containing the whitespace characters to be
1990 are significant. Define with a string containing the whitespace characters to be
1991 matched; default is " \\t\\r\\n". Also takes optional min, max, and exact arguments,
1991 matched; default is " \\t\\r\\n". Also takes optional min, max, and exact arguments,
1992 as defined for the Word class."""
1992 as defined for the Word class."""
1993 whiteStrs = {
1993 whiteStrs = {
1994 " " : "<SPC>",
1994 " " : "<SPC>",
1995 "\t": "<TAB>",
1995 "\t": "<TAB>",
1996 "\n": "<LF>",
1996 "\n": "<LF>",
1997 "\r": "<CR>",
1997 "\r": "<CR>",
1998 "\f": "<FF>",
1998 "\f": "<FF>",
1999 }
1999 }
2000 def __init__(self, ws=" \t\r\n", min=1, max=0, exact=0):
2000 def __init__(self, ws=" \t\r\n", min=1, max=0, exact=0):
2001 super(White,self).__init__()
2001 super(White,self).__init__()
2002 self.matchWhite = ws
2002 self.matchWhite = ws
2003 self.setWhitespaceChars( "".join([c for c in self.whiteChars if c not in self.matchWhite]) )
2003 self.setWhitespaceChars( "".join([c for c in self.whiteChars if c not in self.matchWhite]) )
2004 #~ self.leaveWhitespace()
2004 #~ self.leaveWhitespace()
2005 self.name = ("".join([White.whiteStrs[c] for c in self.matchWhite]))
2005 self.name = ("".join([White.whiteStrs[c] for c in self.matchWhite]))
2006 self.mayReturnEmpty = True
2006 self.mayReturnEmpty = True
2007 self.errmsg = "Expected " + self.name
2007 self.errmsg = "Expected " + self.name
2008 #self.myException.msg = self.errmsg
2008 #self.myException.msg = self.errmsg
2009
2009
2010 self.minLen = min
2010 self.minLen = min
2011
2011
2012 if max > 0:
2012 if max > 0:
2013 self.maxLen = max
2013 self.maxLen = max
2014 else:
2014 else:
2015 self.maxLen = _MAX_INT
2015 self.maxLen = _MAX_INT
2016
2016
2017 if exact > 0:
2017 if exact > 0:
2018 self.maxLen = exact
2018 self.maxLen = exact
2019 self.minLen = exact
2019 self.minLen = exact
2020
2020
2021 def parseImpl( self, instring, loc, doActions=True ):
2021 def parseImpl( self, instring, loc, doActions=True ):
2022 if not(instring[ loc ] in self.matchWhite):
2022 if not(instring[ loc ] in self.matchWhite):
2023 #~ raise ParseException( instring, loc, self.errmsg )
2023 #~ raise ParseException( instring, loc, self.errmsg )
2024 exc = self.myException
2024 exc = self.myException
2025 exc.loc = loc
2025 exc.loc = loc
2026 exc.pstr = instring
2026 exc.pstr = instring
2027 raise exc
2027 raise exc
2028 start = loc
2028 start = loc
2029 loc += 1
2029 loc += 1
2030 maxloc = start + self.maxLen
2030 maxloc = start + self.maxLen
2031 maxloc = min( maxloc, len(instring) )
2031 maxloc = min( maxloc, len(instring) )
2032 while loc < maxloc and instring[loc] in self.matchWhite:
2032 while loc < maxloc and instring[loc] in self.matchWhite:
2033 loc += 1
2033 loc += 1
2034
2034
2035 if loc - start < self.minLen:
2035 if loc - start < self.minLen:
2036 #~ raise ParseException( instring, loc, self.errmsg )
2036 #~ raise ParseException( instring, loc, self.errmsg )
2037 exc = self.myException
2037 exc = self.myException
2038 exc.loc = loc
2038 exc.loc = loc
2039 exc.pstr = instring
2039 exc.pstr = instring
2040 raise exc
2040 raise exc
2041
2041
2042 return loc, instring[start:loc]
2042 return loc, instring[start:loc]
2043
2043
2044
2044
2045 class _PositionToken(Token):
2045 class _PositionToken(Token):
2046 def __init__( self ):
2046 def __init__( self ):
2047 super(_PositionToken,self).__init__()
2047 super(_PositionToken,self).__init__()
2048 self.name=self.__class__.__name__
2048 self.name=self.__class__.__name__
2049 self.mayReturnEmpty = True
2049 self.mayReturnEmpty = True
2050 self.mayIndexError = False
2050 self.mayIndexError = False
2051
2051
2052 class GoToColumn(_PositionToken):
2052 class GoToColumn(_PositionToken):
2053 """Token to advance to a specific column of input text; useful for tabular report scraping."""
2053 """Token to advance to a specific column of input text; useful for tabular report scraping."""
2054 def __init__( self, colno ):
2054 def __init__( self, colno ):
2055 super(GoToColumn,self).__init__()
2055 super(GoToColumn,self).__init__()
2056 self.col = colno
2056 self.col = colno
2057
2057
2058 def preParse( self, instring, loc ):
2058 def preParse( self, instring, loc ):
2059 if col(loc,instring) != self.col:
2059 if col(loc,instring) != self.col:
2060 instrlen = len(instring)
2060 instrlen = len(instring)
2061 if self.ignoreExprs:
2061 if self.ignoreExprs:
2062 loc = self._skipIgnorables( instring, loc )
2062 loc = self._skipIgnorables( instring, loc )
2063 while loc < instrlen and instring[loc].isspace() and col( loc, instring ) != self.col :
2063 while loc < instrlen and instring[loc].isspace() and col( loc, instring ) != self.col :
2064 loc += 1
2064 loc += 1
2065 return loc
2065 return loc
2066
2066
2067 def parseImpl( self, instring, loc, doActions=True ):
2067 def parseImpl( self, instring, loc, doActions=True ):
2068 thiscol = col( loc, instring )
2068 thiscol = col( loc, instring )
2069 if thiscol > self.col:
2069 if thiscol > self.col:
2070 raise ParseException( instring, loc, "Text not in expected column", self )
2070 raise ParseException( instring, loc, "Text not in expected column", self )
2071 newloc = loc + self.col - thiscol
2071 newloc = loc + self.col - thiscol
2072 ret = instring[ loc: newloc ]
2072 ret = instring[ loc: newloc ]
2073 return newloc, ret
2073 return newloc, ret
2074
2074
2075 class LineStart(_PositionToken):
2075 class LineStart(_PositionToken):
2076 """Matches if current position is at the beginning of a line within the parse string"""
2076 """Matches if current position is at the beginning of a line within the parse string"""
2077 def __init__( self ):
2077 def __init__( self ):
2078 super(LineStart,self).__init__()
2078 super(LineStart,self).__init__()
2079 self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") )
2079 self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") )
2080 self.errmsg = "Expected start of line"
2080 self.errmsg = "Expected start of line"
2081 #self.myException.msg = self.errmsg
2081 #self.myException.msg = self.errmsg
2082
2082
2083 def preParse( self, instring, loc ):
2083 def preParse( self, instring, loc ):
2084 preloc = super(LineStart,self).preParse(instring,loc)
2084 preloc = super(LineStart,self).preParse(instring,loc)
2085 if instring[preloc] == "\n":
2085 if instring[preloc] == "\n":
2086 loc += 1
2086 loc += 1
2087 return loc
2087 return loc
2088
2088
2089 def parseImpl( self, instring, loc, doActions=True ):
2089 def parseImpl( self, instring, loc, doActions=True ):
2090 if not( loc==0 or
2090 if not( loc==0 or
2091 (loc == self.preParse( instring, 0 )) or
2091 (loc == self.preParse( instring, 0 )) or
2092 (instring[loc-1] == "\n") ): #col(loc, instring) != 1:
2092 (instring[loc-1] == "\n") ): #col(loc, instring) != 1:
2093 #~ raise ParseException( instring, loc, "Expected start of line" )
2093 #~ raise ParseException( instring, loc, "Expected start of line" )
2094 exc = self.myException
2094 exc = self.myException
2095 exc.loc = loc
2095 exc.loc = loc
2096 exc.pstr = instring
2096 exc.pstr = instring
2097 raise exc
2097 raise exc
2098 return loc, []
2098 return loc, []
2099
2099
2100 class LineEnd(_PositionToken):
2100 class LineEnd(_PositionToken):
2101 """Matches if current position is at the end of a line within the parse string"""
2101 """Matches if current position is at the end of a line within the parse string"""
2102 def __init__( self ):
2102 def __init__( self ):
2103 super(LineEnd,self).__init__()
2103 super(LineEnd,self).__init__()
2104 self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") )
2104 self.setWhitespaceChars( ParserElement.DEFAULT_WHITE_CHARS.replace("\n","") )
2105 self.errmsg = "Expected end of line"
2105 self.errmsg = "Expected end of line"
2106 #self.myException.msg = self.errmsg
2106 #self.myException.msg = self.errmsg
2107
2107
2108 def parseImpl( self, instring, loc, doActions=True ):
2108 def parseImpl( self, instring, loc, doActions=True ):
2109 if loc<len(instring):
2109 if loc<len(instring):
2110 if instring[loc] == "\n":
2110 if instring[loc] == "\n":
2111 return loc+1, "\n"
2111 return loc+1, "\n"
2112 else:
2112 else:
2113 #~ raise ParseException( instring, loc, "Expected end of line" )
2113 #~ raise ParseException( instring, loc, "Expected end of line" )
2114 exc = self.myException
2114 exc = self.myException
2115 exc.loc = loc
2115 exc.loc = loc
2116 exc.pstr = instring
2116 exc.pstr = instring
2117 raise exc
2117 raise exc
2118 elif loc == len(instring):
2118 elif loc == len(instring):
2119 return loc+1, []
2119 return loc+1, []
2120 else:
2120 else:
2121 exc = self.myException
2121 exc = self.myException
2122 exc.loc = loc
2122 exc.loc = loc
2123 exc.pstr = instring
2123 exc.pstr = instring
2124 raise exc
2124 raise exc
2125
2125
2126 class StringStart(_PositionToken):
2126 class StringStart(_PositionToken):
2127 """Matches if current position is at the beginning of the parse string"""
2127 """Matches if current position is at the beginning of the parse string"""
2128 def __init__( self ):
2128 def __init__( self ):
2129 super(StringStart,self).__init__()
2129 super(StringStart,self).__init__()
2130 self.errmsg = "Expected start of text"
2130 self.errmsg = "Expected start of text"
2131 #self.myException.msg = self.errmsg
2131 #self.myException.msg = self.errmsg
2132
2132
2133 def parseImpl( self, instring, loc, doActions=True ):
2133 def parseImpl( self, instring, loc, doActions=True ):
2134 if loc != 0:
2134 if loc != 0:
2135 # see if entire string up to here is just whitespace and ignoreables
2135 # see if entire string up to here is just whitespace and ignoreables
2136 if loc != self.preParse( instring, 0 ):
2136 if loc != self.preParse( instring, 0 ):
2137 #~ raise ParseException( instring, loc, "Expected start of text" )
2137 #~ raise ParseException( instring, loc, "Expected start of text" )
2138 exc = self.myException
2138 exc = self.myException
2139 exc.loc = loc
2139 exc.loc = loc
2140 exc.pstr = instring
2140 exc.pstr = instring
2141 raise exc
2141 raise exc
2142 return loc, []
2142 return loc, []
2143
2143
2144 class StringEnd(_PositionToken):
2144 class StringEnd(_PositionToken):
2145 """Matches if current position is at the end of the parse string"""
2145 """Matches if current position is at the end of the parse string"""
2146 def __init__( self ):
2146 def __init__( self ):
2147 super(StringEnd,self).__init__()
2147 super(StringEnd,self).__init__()
2148 self.errmsg = "Expected end of text"
2148 self.errmsg = "Expected end of text"
2149 #self.myException.msg = self.errmsg
2149 #self.myException.msg = self.errmsg
2150
2150
2151 def parseImpl( self, instring, loc, doActions=True ):
2151 def parseImpl( self, instring, loc, doActions=True ):
2152 if loc < len(instring):
2152 if loc < len(instring):
2153 #~ raise ParseException( instring, loc, "Expected end of text" )
2153 #~ raise ParseException( instring, loc, "Expected end of text" )
2154 exc = self.myException
2154 exc = self.myException
2155 exc.loc = loc
2155 exc.loc = loc
2156 exc.pstr = instring
2156 exc.pstr = instring
2157 raise exc
2157 raise exc
2158 elif loc == len(instring):
2158 elif loc == len(instring):
2159 return loc+1, []
2159 return loc+1, []
2160 elif loc > len(instring):
2160 elif loc > len(instring):
2161 return loc, []
2161 return loc, []
2162 else:
2162 else:
2163 exc = self.myException
2163 exc = self.myException
2164 exc.loc = loc
2164 exc.loc = loc
2165 exc.pstr = instring
2165 exc.pstr = instring
2166 raise exc
2166 raise exc
2167
2167
2168 class WordStart(_PositionToken):
2168 class WordStart(_PositionToken):
2169 """Matches if the current position is at the beginning of a Word, and
2169 """Matches if the current position is at the beginning of a Word, and
2170 is not preceded by any character in a given set of wordChars
2170 is not preceded by any character in a given set of wordChars
2171 (default=printables). To emulate the \b behavior of regular expressions,
2171 (default=printables). To emulate the \b behavior of regular expressions,
2172 use WordStart(alphanums). WordStart will also match at the beginning of
2172 use WordStart(alphanums). WordStart will also match at the beginning of
2173 the string being parsed, or at the beginning of a line.
2173 the string being parsed, or at the beginning of a line.
2174 """
2174 """
2175 def __init__(self, wordChars = printables):
2175 def __init__(self, wordChars = printables):
2176 super(WordStart,self).__init__()
2176 super(WordStart,self).__init__()
2177 self.wordChars = _str2dict(wordChars)
2177 self.wordChars = _str2dict(wordChars)
2178 self.errmsg = "Not at the start of a word"
2178 self.errmsg = "Not at the start of a word"
2179
2179
2180 def parseImpl(self, instring, loc, doActions=True ):
2180 def parseImpl(self, instring, loc, doActions=True ):
2181 if loc != 0:
2181 if loc != 0:
2182 if (instring[loc-1] in self.wordChars or
2182 if (instring[loc-1] in self.wordChars or
2183 instring[loc] not in self.wordChars):
2183 instring[loc] not in self.wordChars):
2184 exc = self.myException
2184 exc = self.myException
2185 exc.loc = loc
2185 exc.loc = loc
2186 exc.pstr = instring
2186 exc.pstr = instring
2187 raise exc
2187 raise exc
2188 return loc, []
2188 return loc, []
2189
2189
2190 class WordEnd(_PositionToken):
2190 class WordEnd(_PositionToken):
2191 """Matches if the current position is at the end of a Word, and
2191 """Matches if the current position is at the end of a Word, and
2192 is not followed by any character in a given set of wordChars
2192 is not followed by any character in a given set of wordChars
2193 (default=printables). To emulate the \b behavior of regular expressions,
2193 (default=printables). To emulate the \b behavior of regular expressions,
2194 use WordEnd(alphanums). WordEnd will also match at the end of
2194 use WordEnd(alphanums). WordEnd will also match at the end of
2195 the string being parsed, or at the end of a line.
2195 the string being parsed, or at the end of a line.
2196 """
2196 """
2197 def __init__(self, wordChars = printables):
2197 def __init__(self, wordChars = printables):
2198 super(WordEnd,self).__init__()
2198 super(WordEnd,self).__init__()
2199 self.wordChars = _str2dict(wordChars)
2199 self.wordChars = _str2dict(wordChars)
2200 self.skipWhitespace = False
2200 self.skipWhitespace = False
2201 self.errmsg = "Not at the end of a word"
2201 self.errmsg = "Not at the end of a word"
2202
2202
2203 def parseImpl(self, instring, loc, doActions=True ):
2203 def parseImpl(self, instring, loc, doActions=True ):
2204 instrlen = len(instring)
2204 instrlen = len(instring)
2205 if instrlen>0 and loc<instrlen:
2205 if instrlen>0 and loc<instrlen:
2206 if (instring[loc] in self.wordChars or
2206 if (instring[loc] in self.wordChars or
2207 instring[loc-1] not in self.wordChars):
2207 instring[loc-1] not in self.wordChars):
2208 #~ raise ParseException( instring, loc, "Expected end of word" )
2208 #~ raise ParseException( instring, loc, "Expected end of word" )
2209 exc = self.myException
2209 exc = self.myException
2210 exc.loc = loc
2210 exc.loc = loc
2211 exc.pstr = instring
2211 exc.pstr = instring
2212 raise exc
2212 raise exc
2213 return loc, []
2213 return loc, []
2214
2214
2215
2215
2216 class ParseExpression(ParserElement):
2216 class ParseExpression(ParserElement):
2217 """Abstract subclass of ParserElement, for combining and post-processing parsed tokens."""
2217 """Abstract subclass of ParserElement, for combining and post-processing parsed tokens."""
2218 def __init__( self, exprs, savelist = False ):
2218 def __init__( self, exprs, savelist = False ):
2219 super(ParseExpression,self).__init__(savelist)
2219 super(ParseExpression,self).__init__(savelist)
2220 if isinstance( exprs, list ):
2220 if isinstance( exprs, list ):
2221 self.exprs = exprs
2221 self.exprs = exprs
2222 elif isinstance( exprs, basestring ):
2222 elif isinstance( exprs, basestring ):
2223 self.exprs = [ Literal( exprs ) ]
2223 self.exprs = [ Literal( exprs ) ]
2224 else:
2224 else:
2225 try:
2225 try:
2226 self.exprs = list( exprs )
2226 self.exprs = list( exprs )
2227 except TypeError:
2227 except TypeError:
2228 self.exprs = [ exprs ]
2228 self.exprs = [ exprs ]
2229 self.callPreparse = False
2229 self.callPreparse = False
2230
2230
2231 def __getitem__( self, i ):
2231 def __getitem__( self, i ):
2232 return self.exprs[i]
2232 return self.exprs[i]
2233
2233
2234 def append( self, other ):
2234 def append( self, other ):
2235 self.exprs.append( other )
2235 self.exprs.append( other )
2236 self.strRepr = None
2236 self.strRepr = None
2237 return self
2237 return self
2238
2238
2239 def leaveWhitespace( self ):
2239 def leaveWhitespace( self ):
2240 """Extends leaveWhitespace defined in base class, and also invokes leaveWhitespace on
2240 """Extends leaveWhitespace defined in base class, and also invokes leaveWhitespace on
2241 all contained expressions."""
2241 all contained expressions."""
2242 self.skipWhitespace = False
2242 self.skipWhitespace = False
2243 self.exprs = [ e.copy() for e in self.exprs ]
2243 self.exprs = [ e.copy() for e in self.exprs ]
2244 for e in self.exprs:
2244 for e in self.exprs:
2245 e.leaveWhitespace()
2245 e.leaveWhitespace()
2246 return self
2246 return self
2247
2247
2248 def ignore( self, other ):
2248 def ignore( self, other ):
2249 if isinstance( other, Suppress ):
2249 if isinstance( other, Suppress ):
2250 if other not in self.ignoreExprs:
2250 if other not in self.ignoreExprs:
2251 super( ParseExpression, self).ignore( other )
2251 super( ParseExpression, self).ignore( other )
2252 for e in self.exprs:
2252 for e in self.exprs:
2253 e.ignore( self.ignoreExprs[-1] )
2253 e.ignore( self.ignoreExprs[-1] )
2254 else:
2254 else:
2255 super( ParseExpression, self).ignore( other )
2255 super( ParseExpression, self).ignore( other )
2256 for e in self.exprs:
2256 for e in self.exprs:
2257 e.ignore( self.ignoreExprs[-1] )
2257 e.ignore( self.ignoreExprs[-1] )
2258 return self
2258 return self
2259
2259
2260 def __str__( self ):
2260 def __str__( self ):
2261 try:
2261 try:
2262 return super(ParseExpression,self).__str__()
2262 return super(ParseExpression,self).__str__()
2263 except:
2263 except:
2264 pass
2264 pass
2265
2265
2266 if self.strRepr is None:
2266 if self.strRepr is None:
2267 self.strRepr = "%s:(%s)" % ( self.__class__.__name__, _ustr(self.exprs) )
2267 self.strRepr = "%s:(%s)" % ( self.__class__.__name__, _ustr(self.exprs) )
2268 return self.strRepr
2268 return self.strRepr
2269
2269
2270 def streamline( self ):
2270 def streamline( self ):
2271 super(ParseExpression,self).streamline()
2271 super(ParseExpression,self).streamline()
2272
2272
2273 for e in self.exprs:
2273 for e in self.exprs:
2274 e.streamline()
2274 e.streamline()
2275
2275
2276 # collapse nested And's of the form And( And( And( a,b), c), d) to And( a,b,c,d )
2276 # collapse nested And's of the form And( And( And( a,b), c), d) to And( a,b,c,d )
2277 # but only if there are no parse actions or resultsNames on the nested And's
2277 # but only if there are no parse actions or resultsNames on the nested And's
2278 # (likewise for Or's and MatchFirst's)
2278 # (likewise for Or's and MatchFirst's)
2279 if ( len(self.exprs) == 2 ):
2279 if ( len(self.exprs) == 2 ):
2280 other = self.exprs[0]
2280 other = self.exprs[0]
2281 if ( isinstance( other, self.__class__ ) and
2281 if ( isinstance( other, self.__class__ ) and
2282 not(other.parseAction) and
2282 not(other.parseAction) and
2283 other.resultsName is None and
2283 other.resultsName is None and
2284 not other.debug ):
2284 not other.debug ):
2285 self.exprs = other.exprs[:] + [ self.exprs[1] ]
2285 self.exprs = other.exprs[:] + [ self.exprs[1] ]
2286 self.strRepr = None
2286 self.strRepr = None
2287 self.mayReturnEmpty |= other.mayReturnEmpty
2287 self.mayReturnEmpty |= other.mayReturnEmpty
2288 self.mayIndexError |= other.mayIndexError
2288 self.mayIndexError |= other.mayIndexError
2289
2289
2290 other = self.exprs[-1]
2290 other = self.exprs[-1]
2291 if ( isinstance( other, self.__class__ ) and
2291 if ( isinstance( other, self.__class__ ) and
2292 not(other.parseAction) and
2292 not(other.parseAction) and
2293 other.resultsName is None and
2293 other.resultsName is None and
2294 not other.debug ):
2294 not other.debug ):
2295 self.exprs = self.exprs[:-1] + other.exprs[:]
2295 self.exprs = self.exprs[:-1] + other.exprs[:]
2296 self.strRepr = None
2296 self.strRepr = None
2297 self.mayReturnEmpty |= other.mayReturnEmpty
2297 self.mayReturnEmpty |= other.mayReturnEmpty
2298 self.mayIndexError |= other.mayIndexError
2298 self.mayIndexError |= other.mayIndexError
2299
2299
2300 return self
2300 return self
2301
2301
2302 def setResultsName( self, name, listAllMatches=False ):
2302 def setResultsName( self, name, listAllMatches=False ):
2303 ret = super(ParseExpression,self).setResultsName(name,listAllMatches)
2303 ret = super(ParseExpression,self).setResultsName(name,listAllMatches)
2304 return ret
2304 return ret
2305
2305
2306 def validate( self, validateTrace=[] ):
2306 def validate( self, validateTrace=[] ):
2307 tmp = validateTrace[:]+[self]
2307 tmp = validateTrace[:]+[self]
2308 for e in self.exprs:
2308 for e in self.exprs:
2309 e.validate(tmp)
2309 e.validate(tmp)
2310 self.checkRecursion( [] )
2310 self.checkRecursion( [] )
2311
2311
2312 class And(ParseExpression):
2312 class And(ParseExpression):
2313 """Requires all given ParseExpressions to be found in the given order.
2313 """Requires all given ParseExpressions to be found in the given order.
2314 Expressions may be separated by whitespace.
2314 Expressions may be separated by whitespace.
2315 May be constructed using the '+' operator.
2315 May be constructed using the '+' operator.
2316 """
2316 """
2317
2317
2318 class _ErrorStop(Empty):
2318 class _ErrorStop(Empty):
2319 def __init__(self, *args, **kwargs):
2319 def __init__(self, *args, **kwargs):
2320 super(Empty,self).__init__(*args, **kwargs)
2320 super(Empty,self).__init__(*args, **kwargs)
2321 self.leaveWhitespace()
2321 self.leaveWhitespace()
2322
2322
2323 def __init__( self, exprs, savelist = True ):
2323 def __init__( self, exprs, savelist = True ):
2324 super(And,self).__init__(exprs, savelist)
2324 super(And,self).__init__(exprs, savelist)
2325 self.mayReturnEmpty = True
2325 self.mayReturnEmpty = True
2326 for e in self.exprs:
2326 for e in self.exprs:
2327 if not e.mayReturnEmpty:
2327 if not e.mayReturnEmpty:
2328 self.mayReturnEmpty = False
2328 self.mayReturnEmpty = False
2329 break
2329 break
2330 self.setWhitespaceChars( exprs[0].whiteChars )
2330 self.setWhitespaceChars( exprs[0].whiteChars )
2331 self.skipWhitespace = exprs[0].skipWhitespace
2331 self.skipWhitespace = exprs[0].skipWhitespace
2332 self.callPreparse = True
2332 self.callPreparse = True
2333
2333
2334 def parseImpl( self, instring, loc, doActions=True ):
2334 def parseImpl( self, instring, loc, doActions=True ):
2335 # pass False as last arg to _parse for first element, since we already
2335 # pass False as last arg to _parse for first element, since we already
2336 # pre-parsed the string as part of our And pre-parsing
2336 # pre-parsed the string as part of our And pre-parsing
2337 loc, resultlist = self.exprs[0]._parse( instring, loc, doActions, callPreParse=False )
2337 loc, resultlist = self.exprs[0]._parse( instring, loc, doActions, callPreParse=False )
2338 errorStop = False
2338 errorStop = False
2339 for e in self.exprs[1:]:
2339 for e in self.exprs[1:]:
2340 if isinstance(e, And._ErrorStop):
2340 if isinstance(e, And._ErrorStop):
2341 errorStop = True
2341 errorStop = True
2342 continue
2342 continue
2343 if errorStop:
2343 if errorStop:
2344 try:
2344 try:
2345 loc, exprtokens = e._parse( instring, loc, doActions )
2345 loc, exprtokens = e._parse( instring, loc, doActions )
2346 except ParseSyntaxException:
2346 except ParseSyntaxException:
2347 raise
2347 raise
2348 except ParseBaseException, pe:
2348 except ParseBaseException, pe:
2349 raise ParseSyntaxException(pe)
2349 raise ParseSyntaxException(pe)
2350 except IndexError, ie:
2350 except IndexError, ie:
2351 raise ParseSyntaxException( ParseException(instring, len(instring), self.errmsg, self) )
2351 raise ParseSyntaxException( ParseException(instring, len(instring), self.errmsg, self) )
2352 else:
2352 else:
2353 loc, exprtokens = e._parse( instring, loc, doActions )
2353 loc, exprtokens = e._parse( instring, loc, doActions )
2354 if exprtokens or exprtokens.keys():
2354 if exprtokens or exprtokens.keys():
2355 resultlist += exprtokens
2355 resultlist += exprtokens
2356 return loc, resultlist
2356 return loc, resultlist
2357
2357
2358 def __iadd__(self, other ):
2358 def __iadd__(self, other ):
2359 if isinstance( other, basestring ):
2359 if isinstance( other, basestring ):
2360 other = Literal( other )
2360 other = Literal( other )
2361 return self.append( other ) #And( [ self, other ] )
2361 return self.append( other ) #And( [ self, other ] )
2362
2362
2363 def checkRecursion( self, parseElementList ):
2363 def checkRecursion( self, parseElementList ):
2364 subRecCheckList = parseElementList[:] + [ self ]
2364 subRecCheckList = parseElementList[:] + [ self ]
2365 for e in self.exprs:
2365 for e in self.exprs:
2366 e.checkRecursion( subRecCheckList )
2366 e.checkRecursion( subRecCheckList )
2367 if not e.mayReturnEmpty:
2367 if not e.mayReturnEmpty:
2368 break
2368 break
2369
2369
2370 def __str__( self ):
2370 def __str__( self ):
2371 if hasattr(self,"name"):
2371 if hasattr(self,"name"):
2372 return self.name
2372 return self.name
2373
2373
2374 if self.strRepr is None:
2374 if self.strRepr is None:
2375 self.strRepr = "{" + " ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2375 self.strRepr = "{" + " ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2376
2376
2377 return self.strRepr
2377 return self.strRepr
2378
2378
2379
2379
2380 class Or(ParseExpression):
2380 class Or(ParseExpression):
2381 """Requires that at least one ParseExpression is found.
2381 """Requires that at least one ParseExpression is found.
2382 If two expressions match, the expression that matches the longest string will be used.
2382 If two expressions match, the expression that matches the longest string will be used.
2383 May be constructed using the '^' operator.
2383 May be constructed using the '^' operator.
2384 """
2384 """
2385 def __init__( self, exprs, savelist = False ):
2385 def __init__( self, exprs, savelist = False ):
2386 super(Or,self).__init__(exprs, savelist)
2386 super(Or,self).__init__(exprs, savelist)
2387 self.mayReturnEmpty = False
2387 self.mayReturnEmpty = False
2388 for e in self.exprs:
2388 for e in self.exprs:
2389 if e.mayReturnEmpty:
2389 if e.mayReturnEmpty:
2390 self.mayReturnEmpty = True
2390 self.mayReturnEmpty = True
2391 break
2391 break
2392
2392
2393 def parseImpl( self, instring, loc, doActions=True ):
2393 def parseImpl( self, instring, loc, doActions=True ):
2394 maxExcLoc = -1
2394 maxExcLoc = -1
2395 maxMatchLoc = -1
2395 maxMatchLoc = -1
2396 maxException = None
2396 maxException = None
2397 for e in self.exprs:
2397 for e in self.exprs:
2398 try:
2398 try:
2399 loc2 = e.tryParse( instring, loc )
2399 loc2 = e.tryParse( instring, loc )
2400 except ParseException, err:
2400 except ParseException, err:
2401 if err.loc > maxExcLoc:
2401 if err.loc > maxExcLoc:
2402 maxException = err
2402 maxException = err
2403 maxExcLoc = err.loc
2403 maxExcLoc = err.loc
2404 except IndexError:
2404 except IndexError:
2405 if len(instring) > maxExcLoc:
2405 if len(instring) > maxExcLoc:
2406 maxException = ParseException(instring,len(instring),e.errmsg,self)
2406 maxException = ParseException(instring,len(instring),e.errmsg,self)
2407 maxExcLoc = len(instring)
2407 maxExcLoc = len(instring)
2408 else:
2408 else:
2409 if loc2 > maxMatchLoc:
2409 if loc2 > maxMatchLoc:
2410 maxMatchLoc = loc2
2410 maxMatchLoc = loc2
2411 maxMatchExp = e
2411 maxMatchExp = e
2412
2412
2413 if maxMatchLoc < 0:
2413 if maxMatchLoc < 0:
2414 if maxException is not None:
2414 if maxException is not None:
2415 raise maxException
2415 raise maxException
2416 else:
2416 else:
2417 raise ParseException(instring, loc, "no defined alternatives to match", self)
2417 raise ParseException(instring, loc, "no defined alternatives to match", self)
2418
2418
2419 return maxMatchExp._parse( instring, loc, doActions )
2419 return maxMatchExp._parse( instring, loc, doActions )
2420
2420
2421 def __ixor__(self, other ):
2421 def __ixor__(self, other ):
2422 if isinstance( other, basestring ):
2422 if isinstance( other, basestring ):
2423 other = Literal( other )
2423 other = Literal( other )
2424 return self.append( other ) #Or( [ self, other ] )
2424 return self.append( other ) #Or( [ self, other ] )
2425
2425
2426 def __str__( self ):
2426 def __str__( self ):
2427 if hasattr(self,"name"):
2427 if hasattr(self,"name"):
2428 return self.name
2428 return self.name
2429
2429
2430 if self.strRepr is None:
2430 if self.strRepr is None:
2431 self.strRepr = "{" + " ^ ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2431 self.strRepr = "{" + " ^ ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2432
2432
2433 return self.strRepr
2433 return self.strRepr
2434
2434
2435 def checkRecursion( self, parseElementList ):
2435 def checkRecursion( self, parseElementList ):
2436 subRecCheckList = parseElementList[:] + [ self ]
2436 subRecCheckList = parseElementList[:] + [ self ]
2437 for e in self.exprs:
2437 for e in self.exprs:
2438 e.checkRecursion( subRecCheckList )
2438 e.checkRecursion( subRecCheckList )
2439
2439
2440
2440
2441 class MatchFirst(ParseExpression):
2441 class MatchFirst(ParseExpression):
2442 """Requires that at least one ParseExpression is found.
2442 """Requires that at least one ParseExpression is found.
2443 If two expressions match, the first one listed is the one that will match.
2443 If two expressions match, the first one listed is the one that will match.
2444 May be constructed using the '|' operator.
2444 May be constructed using the '|' operator.
2445 """
2445 """
2446 def __init__( self, exprs, savelist = False ):
2446 def __init__( self, exprs, savelist = False ):
2447 super(MatchFirst,self).__init__(exprs, savelist)
2447 super(MatchFirst,self).__init__(exprs, savelist)
2448 if exprs:
2448 if exprs:
2449 self.mayReturnEmpty = False
2449 self.mayReturnEmpty = False
2450 for e in self.exprs:
2450 for e in self.exprs:
2451 if e.mayReturnEmpty:
2451 if e.mayReturnEmpty:
2452 self.mayReturnEmpty = True
2452 self.mayReturnEmpty = True
2453 break
2453 break
2454 else:
2454 else:
2455 self.mayReturnEmpty = True
2455 self.mayReturnEmpty = True
2456
2456
2457 def parseImpl( self, instring, loc, doActions=True ):
2457 def parseImpl( self, instring, loc, doActions=True ):
2458 maxExcLoc = -1
2458 maxExcLoc = -1
2459 maxException = None
2459 maxException = None
2460 for e in self.exprs:
2460 for e in self.exprs:
2461 try:
2461 try:
2462 ret = e._parse( instring, loc, doActions )
2462 ret = e._parse( instring, loc, doActions )
2463 return ret
2463 return ret
2464 except ParseException, err:
2464 except ParseException, err:
2465 if err.loc > maxExcLoc:
2465 if err.loc > maxExcLoc:
2466 maxException = err
2466 maxException = err
2467 maxExcLoc = err.loc
2467 maxExcLoc = err.loc
2468 except IndexError:
2468 except IndexError:
2469 if len(instring) > maxExcLoc:
2469 if len(instring) > maxExcLoc:
2470 maxException = ParseException(instring,len(instring),e.errmsg,self)
2470 maxException = ParseException(instring,len(instring),e.errmsg,self)
2471 maxExcLoc = len(instring)
2471 maxExcLoc = len(instring)
2472
2472
2473 # only got here if no expression matched, raise exception for match that made it the furthest
2473 # only got here if no expression matched, raise exception for match that made it the furthest
2474 else:
2474 else:
2475 if maxException is not None:
2475 if maxException is not None:
2476 raise maxException
2476 raise maxException
2477 else:
2477 else:
2478 raise ParseException(instring, loc, "no defined alternatives to match", self)
2478 raise ParseException(instring, loc, "no defined alternatives to match", self)
2479
2479
2480 def __ior__(self, other ):
2480 def __ior__(self, other ):
2481 if isinstance( other, basestring ):
2481 if isinstance( other, basestring ):
2482 other = Literal( other )
2482 other = Literal( other )
2483 return self.append( other ) #MatchFirst( [ self, other ] )
2483 return self.append( other ) #MatchFirst( [ self, other ] )
2484
2484
2485 def __str__( self ):
2485 def __str__( self ):
2486 if hasattr(self,"name"):
2486 if hasattr(self,"name"):
2487 return self.name
2487 return self.name
2488
2488
2489 if self.strRepr is None:
2489 if self.strRepr is None:
2490 self.strRepr = "{" + " | ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2490 self.strRepr = "{" + " | ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2491
2491
2492 return self.strRepr
2492 return self.strRepr
2493
2493
2494 def checkRecursion( self, parseElementList ):
2494 def checkRecursion( self, parseElementList ):
2495 subRecCheckList = parseElementList[:] + [ self ]
2495 subRecCheckList = parseElementList[:] + [ self ]
2496 for e in self.exprs:
2496 for e in self.exprs:
2497 e.checkRecursion( subRecCheckList )
2497 e.checkRecursion( subRecCheckList )
2498
2498
2499
2499
2500 class Each(ParseExpression):
2500 class Each(ParseExpression):
2501 """Requires all given ParseExpressions to be found, but in any order.
2501 """Requires all given ParseExpressions to be found, but in any order.
2502 Expressions may be separated by whitespace.
2502 Expressions may be separated by whitespace.
2503 May be constructed using the '&' operator.
2503 May be constructed using the '&' operator.
2504 """
2504 """
2505 def __init__( self, exprs, savelist = True ):
2505 def __init__( self, exprs, savelist = True ):
2506 super(Each,self).__init__(exprs, savelist)
2506 super(Each,self).__init__(exprs, savelist)
2507 self.mayReturnEmpty = True
2507 self.mayReturnEmpty = True
2508 for e in self.exprs:
2508 for e in self.exprs:
2509 if not e.mayReturnEmpty:
2509 if not e.mayReturnEmpty:
2510 self.mayReturnEmpty = False
2510 self.mayReturnEmpty = False
2511 break
2511 break
2512 self.skipWhitespace = True
2512 self.skipWhitespace = True
2513 self.initExprGroups = True
2513 self.initExprGroups = True
2514
2514
2515 def parseImpl( self, instring, loc, doActions=True ):
2515 def parseImpl( self, instring, loc, doActions=True ):
2516 if self.initExprGroups:
2516 if self.initExprGroups:
2517 self.optionals = [ e.expr for e in self.exprs if isinstance(e,Optional) ]
2517 self.optionals = [ e.expr for e in self.exprs if isinstance(e,Optional) ]
2518 self.multioptionals = [ e.expr for e in self.exprs if isinstance(e,ZeroOrMore) ]
2518 self.multioptionals = [ e.expr for e in self.exprs if isinstance(e,ZeroOrMore) ]
2519 self.multirequired = [ e.expr for e in self.exprs if isinstance(e,OneOrMore) ]
2519 self.multirequired = [ e.expr for e in self.exprs if isinstance(e,OneOrMore) ]
2520 self.required = [ e for e in self.exprs if not isinstance(e,(Optional,ZeroOrMore,OneOrMore)) ]
2520 self.required = [ e for e in self.exprs if not isinstance(e,(Optional,ZeroOrMore,OneOrMore)) ]
2521 self.required += self.multirequired
2521 self.required += self.multirequired
2522 self.initExprGroups = False
2522 self.initExprGroups = False
2523 tmpLoc = loc
2523 tmpLoc = loc
2524 tmpReqd = self.required[:]
2524 tmpReqd = self.required[:]
2525 tmpOpt = self.optionals[:]
2525 tmpOpt = self.optionals[:]
2526 matchOrder = []
2526 matchOrder = []
2527
2527
2528 keepMatching = True
2528 keepMatching = True
2529 while keepMatching:
2529 while keepMatching:
2530 tmpExprs = tmpReqd + tmpOpt + self.multioptionals + self.multirequired
2530 tmpExprs = tmpReqd + tmpOpt + self.multioptionals + self.multirequired
2531 failed = []
2531 failed = []
2532 for e in tmpExprs:
2532 for e in tmpExprs:
2533 try:
2533 try:
2534 tmpLoc = e.tryParse( instring, tmpLoc )
2534 tmpLoc = e.tryParse( instring, tmpLoc )
2535 except ParseException:
2535 except ParseException:
2536 failed.append(e)
2536 failed.append(e)
2537 else:
2537 else:
2538 matchOrder.append(e)
2538 matchOrder.append(e)
2539 if e in tmpReqd:
2539 if e in tmpReqd:
2540 tmpReqd.remove(e)
2540 tmpReqd.remove(e)
2541 elif e in tmpOpt:
2541 elif e in tmpOpt:
2542 tmpOpt.remove(e)
2542 tmpOpt.remove(e)
2543 if len(failed) == len(tmpExprs):
2543 if len(failed) == len(tmpExprs):
2544 keepMatching = False
2544 keepMatching = False
2545
2545
2546 if tmpReqd:
2546 if tmpReqd:
2547 missing = ", ".join( [ _ustr(e) for e in tmpReqd ] )
2547 missing = ", ".join( [ _ustr(e) for e in tmpReqd ] )
2548 raise ParseException(instring,loc,"Missing one or more required elements (%s)" % missing )
2548 raise ParseException(instring,loc,"Missing one or more required elements (%s)" % missing )
2549
2549
2550 # add any unmatched Optionals, in case they have default values defined
2550 # add any unmatched Optionals, in case they have default values defined
2551 matchOrder += list(e for e in self.exprs if isinstance(e,Optional) and e.expr in tmpOpt)
2551 matchOrder += list(e for e in self.exprs if isinstance(e,Optional) and e.expr in tmpOpt)
2552
2552
2553 resultlist = []
2553 resultlist = []
2554 for e in matchOrder:
2554 for e in matchOrder:
2555 loc,results = e._parse(instring,loc,doActions)
2555 loc,results = e._parse(instring,loc,doActions)
2556 resultlist.append(results)
2556 resultlist.append(results)
2557
2557
2558 finalResults = ParseResults([])
2558 finalResults = ParseResults([])
2559 for r in resultlist:
2559 for r in resultlist:
2560 dups = {}
2560 dups = {}
2561 for k in r.keys():
2561 for k in r.keys():
2562 if k in finalResults.keys():
2562 if k in finalResults.keys():
2563 tmp = ParseResults(finalResults[k])
2563 tmp = ParseResults(finalResults[k])
2564 tmp += ParseResults(r[k])
2564 tmp += ParseResults(r[k])
2565 dups[k] = tmp
2565 dups[k] = tmp
2566 finalResults += ParseResults(r)
2566 finalResults += ParseResults(r)
2567 for k,v in dups.iteritems():
2567 for k,v in dups.iteritems():
2568 finalResults[k] = v
2568 finalResults[k] = v
2569 return loc, finalResults
2569 return loc, finalResults
2570
2570
2571 def __str__( self ):
2571 def __str__( self ):
2572 if hasattr(self,"name"):
2572 if hasattr(self,"name"):
2573 return self.name
2573 return self.name
2574
2574
2575 if self.strRepr is None:
2575 if self.strRepr is None:
2576 self.strRepr = "{" + " & ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2576 self.strRepr = "{" + " & ".join( [ _ustr(e) for e in self.exprs ] ) + "}"
2577
2577
2578 return self.strRepr
2578 return self.strRepr
2579
2579
2580 def checkRecursion( self, parseElementList ):
2580 def checkRecursion( self, parseElementList ):
2581 subRecCheckList = parseElementList[:] + [ self ]
2581 subRecCheckList = parseElementList[:] + [ self ]
2582 for e in self.exprs:
2582 for e in self.exprs:
2583 e.checkRecursion( subRecCheckList )
2583 e.checkRecursion( subRecCheckList )
2584
2584
2585
2585
2586 class ParseElementEnhance(ParserElement):
2586 class ParseElementEnhance(ParserElement):
2587 """Abstract subclass of ParserElement, for combining and post-processing parsed tokens."""
2587 """Abstract subclass of ParserElement, for combining and post-processing parsed tokens."""
2588 def __init__( self, expr, savelist=False ):
2588 def __init__( self, expr, savelist=False ):
2589 super(ParseElementEnhance,self).__init__(savelist)
2589 super(ParseElementEnhance,self).__init__(savelist)
2590 if isinstance( expr, basestring ):
2590 if isinstance( expr, basestring ):
2591 expr = Literal(expr)
2591 expr = Literal(expr)
2592 self.expr = expr
2592 self.expr = expr
2593 self.strRepr = None
2593 self.strRepr = None
2594 if expr is not None:
2594 if expr is not None:
2595 self.mayIndexError = expr.mayIndexError
2595 self.mayIndexError = expr.mayIndexError
2596 self.mayReturnEmpty = expr.mayReturnEmpty
2596 self.mayReturnEmpty = expr.mayReturnEmpty
2597 self.setWhitespaceChars( expr.whiteChars )
2597 self.setWhitespaceChars( expr.whiteChars )
2598 self.skipWhitespace = expr.skipWhitespace
2598 self.skipWhitespace = expr.skipWhitespace
2599 self.saveAsList = expr.saveAsList
2599 self.saveAsList = expr.saveAsList
2600 self.callPreparse = expr.callPreparse
2600 self.callPreparse = expr.callPreparse
2601 self.ignoreExprs.extend(expr.ignoreExprs)
2601 self.ignoreExprs.extend(expr.ignoreExprs)
2602
2602
2603 def parseImpl( self, instring, loc, doActions=True ):
2603 def parseImpl( self, instring, loc, doActions=True ):
2604 if self.expr is not None:
2604 if self.expr is not None:
2605 return self.expr._parse( instring, loc, doActions, callPreParse=False )
2605 return self.expr._parse( instring, loc, doActions, callPreParse=False )
2606 else:
2606 else:
2607 raise ParseException("",loc,self.errmsg,self)
2607 raise ParseException("",loc,self.errmsg,self)
2608
2608
2609 def leaveWhitespace( self ):
2609 def leaveWhitespace( self ):
2610 self.skipWhitespace = False
2610 self.skipWhitespace = False
2611 self.expr = self.expr.copy()
2611 self.expr = self.expr.copy()
2612 if self.expr is not None:
2612 if self.expr is not None:
2613 self.expr.leaveWhitespace()
2613 self.expr.leaveWhitespace()
2614 return self
2614 return self
2615
2615
2616 def ignore( self, other ):
2616 def ignore( self, other ):
2617 if isinstance( other, Suppress ):
2617 if isinstance( other, Suppress ):
2618 if other not in self.ignoreExprs:
2618 if other not in self.ignoreExprs:
2619 super( ParseElementEnhance, self).ignore( other )
2619 super( ParseElementEnhance, self).ignore( other )
2620 if self.expr is not None:
2620 if self.expr is not None:
2621 self.expr.ignore( self.ignoreExprs[-1] )
2621 self.expr.ignore( self.ignoreExprs[-1] )
2622 else:
2622 else:
2623 super( ParseElementEnhance, self).ignore( other )
2623 super( ParseElementEnhance, self).ignore( other )
2624 if self.expr is not None:
2624 if self.expr is not None:
2625 self.expr.ignore( self.ignoreExprs[-1] )
2625 self.expr.ignore( self.ignoreExprs[-1] )
2626 return self
2626 return self
2627
2627
2628 def streamline( self ):
2628 def streamline( self ):
2629 super(ParseElementEnhance,self).streamline()
2629 super(ParseElementEnhance,self).streamline()
2630 if self.expr is not None:
2630 if self.expr is not None:
2631 self.expr.streamline()
2631 self.expr.streamline()
2632 return self
2632 return self
2633
2633
2634 def checkRecursion( self, parseElementList ):
2634 def checkRecursion( self, parseElementList ):
2635 if self in parseElementList:
2635 if self in parseElementList:
2636 raise RecursiveGrammarException( parseElementList+[self] )
2636 raise RecursiveGrammarException( parseElementList+[self] )
2637 subRecCheckList = parseElementList[:] + [ self ]
2637 subRecCheckList = parseElementList[:] + [ self ]
2638 if self.expr is not None:
2638 if self.expr is not None:
2639 self.expr.checkRecursion( subRecCheckList )
2639 self.expr.checkRecursion( subRecCheckList )
2640
2640
2641 def validate( self, validateTrace=[] ):
2641 def validate( self, validateTrace=[] ):
2642 tmp = validateTrace[:]+[self]
2642 tmp = validateTrace[:]+[self]
2643 if self.expr is not None:
2643 if self.expr is not None:
2644 self.expr.validate(tmp)
2644 self.expr.validate(tmp)
2645 self.checkRecursion( [] )
2645 self.checkRecursion( [] )
2646
2646
2647 def __str__( self ):
2647 def __str__( self ):
2648 try:
2648 try:
2649 return super(ParseElementEnhance,self).__str__()
2649 return super(ParseElementEnhance,self).__str__()
2650 except:
2650 except:
2651 pass
2651 pass
2652
2652
2653 if self.strRepr is None and self.expr is not None:
2653 if self.strRepr is None and self.expr is not None:
2654 self.strRepr = "%s:(%s)" % ( self.__class__.__name__, _ustr(self.expr) )
2654 self.strRepr = "%s:(%s)" % ( self.__class__.__name__, _ustr(self.expr) )
2655 return self.strRepr
2655 return self.strRepr
2656
2656
2657
2657
2658 class FollowedBy(ParseElementEnhance):
2658 class FollowedBy(ParseElementEnhance):
2659 """Lookahead matching of the given parse expression. FollowedBy
2659 """Lookahead matching of the given parse expression. FollowedBy
2660 does *not* advance the parsing position within the input string, it only
2660 does *not* advance the parsing position within the input string, it only
2661 verifies that the specified parse expression matches at the current
2661 verifies that the specified parse expression matches at the current
2662 position. FollowedBy always returns a null token list."""
2662 position. FollowedBy always returns a null token list."""
2663 def __init__( self, expr ):
2663 def __init__( self, expr ):
2664 super(FollowedBy,self).__init__(expr)
2664 super(FollowedBy,self).__init__(expr)
2665 self.mayReturnEmpty = True
2665 self.mayReturnEmpty = True
2666
2666
2667 def parseImpl( self, instring, loc, doActions=True ):
2667 def parseImpl( self, instring, loc, doActions=True ):
2668 self.expr.tryParse( instring, loc )
2668 self.expr.tryParse( instring, loc )
2669 return loc, []
2669 return loc, []
2670
2670
2671
2671
2672 class NotAny(ParseElementEnhance):
2672 class NotAny(ParseElementEnhance):
2673 """Lookahead to disallow matching with the given parse expression. NotAny
2673 """Lookahead to disallow matching with the given parse expression. NotAny
2674 does *not* advance the parsing position within the input string, it only
2674 does *not* advance the parsing position within the input string, it only
2675 verifies that the specified parse expression does *not* match at the current
2675 verifies that the specified parse expression does *not* match at the current
2676 position. Also, NotAny does *not* skip over leading whitespace. NotAny
2676 position. Also, NotAny does *not* skip over leading whitespace. NotAny
2677 always returns a null token list. May be constructed using the '~' operator."""
2677 always returns a null token list. May be constructed using the '~' operator."""
2678 def __init__( self, expr ):
2678 def __init__( self, expr ):
2679 super(NotAny,self).__init__(expr)
2679 super(NotAny,self).__init__(expr)
2680 #~ self.leaveWhitespace()
2680 #~ self.leaveWhitespace()
2681 self.skipWhitespace = False # do NOT use self.leaveWhitespace(), don't want to propagate to exprs
2681 self.skipWhitespace = False # do NOT use self.leaveWhitespace(), don't want to propagate to exprs
2682 self.mayReturnEmpty = True
2682 self.mayReturnEmpty = True
2683 self.errmsg = "Found unwanted token, "+_ustr(self.expr)
2683 self.errmsg = "Found unwanted token, "+_ustr(self.expr)
2684 #self.myException = ParseException("",0,self.errmsg,self)
2684 #self.myException = ParseException("",0,self.errmsg,self)
2685
2685
2686 def parseImpl( self, instring, loc, doActions=True ):
2686 def parseImpl( self, instring, loc, doActions=True ):
2687 try:
2687 try:
2688 self.expr.tryParse( instring, loc )
2688 self.expr.tryParse( instring, loc )
2689 except (ParseException,IndexError):
2689 except (ParseException,IndexError):
2690 pass
2690 pass
2691 else:
2691 else:
2692 #~ raise ParseException(instring, loc, self.errmsg )
2692 #~ raise ParseException(instring, loc, self.errmsg )
2693 exc = self.myException
2693 exc = self.myException
2694 exc.loc = loc
2694 exc.loc = loc
2695 exc.pstr = instring
2695 exc.pstr = instring
2696 raise exc
2696 raise exc
2697 return loc, []
2697 return loc, []
2698
2698
2699 def __str__( self ):
2699 def __str__( self ):
2700 if hasattr(self,"name"):
2700 if hasattr(self,"name"):
2701 return self.name
2701 return self.name
2702
2702
2703 if self.strRepr is None:
2703 if self.strRepr is None:
2704 self.strRepr = "~{" + _ustr(self.expr) + "}"
2704 self.strRepr = "~{" + _ustr(self.expr) + "}"
2705
2705
2706 return self.strRepr
2706 return self.strRepr
2707
2707
2708
2708
2709 class ZeroOrMore(ParseElementEnhance):
2709 class ZeroOrMore(ParseElementEnhance):
2710 """Optional repetition of zero or more of the given expression."""
2710 """Optional repetition of zero or more of the given expression."""
2711 def __init__( self, expr ):
2711 def __init__( self, expr ):
2712 super(ZeroOrMore,self).__init__(expr)
2712 super(ZeroOrMore,self).__init__(expr)
2713 self.mayReturnEmpty = True
2713 self.mayReturnEmpty = True
2714
2714
2715 def parseImpl( self, instring, loc, doActions=True ):
2715 def parseImpl( self, instring, loc, doActions=True ):
2716 tokens = []
2716 tokens = []
2717 try:
2717 try:
2718 loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )
2718 loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )
2719 hasIgnoreExprs = ( len(self.ignoreExprs) > 0 )
2719 hasIgnoreExprs = ( len(self.ignoreExprs) > 0 )
2720 while 1:
2720 while 1:
2721 if hasIgnoreExprs:
2721 if hasIgnoreExprs:
2722 preloc = self._skipIgnorables( instring, loc )
2722 preloc = self._skipIgnorables( instring, loc )
2723 else:
2723 else:
2724 preloc = loc
2724 preloc = loc
2725 loc, tmptokens = self.expr._parse( instring, preloc, doActions )
2725 loc, tmptokens = self.expr._parse( instring, preloc, doActions )
2726 if tmptokens or tmptokens.keys():
2726 if tmptokens or tmptokens.keys():
2727 tokens += tmptokens
2727 tokens += tmptokens
2728 except (ParseException,IndexError):
2728 except (ParseException,IndexError):
2729 pass
2729 pass
2730
2730
2731 return loc, tokens
2731 return loc, tokens
2732
2732
2733 def __str__( self ):
2733 def __str__( self ):
2734 if hasattr(self,"name"):
2734 if hasattr(self,"name"):
2735 return self.name
2735 return self.name
2736
2736
2737 if self.strRepr is None:
2737 if self.strRepr is None:
2738 self.strRepr = "[" + _ustr(self.expr) + "]..."
2738 self.strRepr = "[" + _ustr(self.expr) + "]..."
2739
2739
2740 return self.strRepr
2740 return self.strRepr
2741
2741
2742 def setResultsName( self, name, listAllMatches=False ):
2742 def setResultsName( self, name, listAllMatches=False ):
2743 ret = super(ZeroOrMore,self).setResultsName(name,listAllMatches)
2743 ret = super(ZeroOrMore,self).setResultsName(name,listAllMatches)
2744 ret.saveAsList = True
2744 ret.saveAsList = True
2745 return ret
2745 return ret
2746
2746
2747
2747
2748 class OneOrMore(ParseElementEnhance):
2748 class OneOrMore(ParseElementEnhance):
2749 """Repetition of one or more of the given expression."""
2749 """Repetition of one or more of the given expression."""
2750 def parseImpl( self, instring, loc, doActions=True ):
2750 def parseImpl( self, instring, loc, doActions=True ):
2751 # must be at least one
2751 # must be at least one
2752 loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )
2752 loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )
2753 try:
2753 try:
2754 hasIgnoreExprs = ( len(self.ignoreExprs) > 0 )
2754 hasIgnoreExprs = ( len(self.ignoreExprs) > 0 )
2755 while 1:
2755 while 1:
2756 if hasIgnoreExprs:
2756 if hasIgnoreExprs:
2757 preloc = self._skipIgnorables( instring, loc )
2757 preloc = self._skipIgnorables( instring, loc )
2758 else:
2758 else:
2759 preloc = loc
2759 preloc = loc
2760 loc, tmptokens = self.expr._parse( instring, preloc, doActions )
2760 loc, tmptokens = self.expr._parse( instring, preloc, doActions )
2761 if tmptokens or tmptokens.keys():
2761 if tmptokens or tmptokens.keys():
2762 tokens += tmptokens
2762 tokens += tmptokens
2763 except (ParseException,IndexError):
2763 except (ParseException,IndexError):
2764 pass
2764 pass
2765
2765
2766 return loc, tokens
2766 return loc, tokens
2767
2767
2768 def __str__( self ):
2768 def __str__( self ):
2769 if hasattr(self,"name"):
2769 if hasattr(self,"name"):
2770 return self.name
2770 return self.name
2771
2771
2772 if self.strRepr is None:
2772 if self.strRepr is None:
2773 self.strRepr = "{" + _ustr(self.expr) + "}..."
2773 self.strRepr = "{" + _ustr(self.expr) + "}..."
2774
2774
2775 return self.strRepr
2775 return self.strRepr
2776
2776
2777 def setResultsName( self, name, listAllMatches=False ):
2777 def setResultsName( self, name, listAllMatches=False ):
2778 ret = super(OneOrMore,self).setResultsName(name,listAllMatches)
2778 ret = super(OneOrMore,self).setResultsName(name,listAllMatches)
2779 ret.saveAsList = True
2779 ret.saveAsList = True
2780 return ret
2780 return ret
2781
2781
2782 class _NullToken(object):
2782 class _NullToken(object):
2783 def __bool__(self):
2783 def __bool__(self):
2784 return False
2784 return False
2785 __nonzero__ = __bool__
2785 __nonzero__ = __bool__
2786 def __str__(self):
2786 def __str__(self):
2787 return ""
2787 return ""
2788
2788
2789 _optionalNotMatched = _NullToken()
2789 _optionalNotMatched = _NullToken()
2790 class Optional(ParseElementEnhance):
2790 class Optional(ParseElementEnhance):
2791 """Optional matching of the given expression.
2791 """Optional matching of the given expression.
2792 A default return string can also be specified, if the optional expression
2792 A default return string can also be specified, if the optional expression
2793 is not found.
2793 is not found.
2794 """
2794 """
2795 def __init__( self, exprs, default=_optionalNotMatched ):
2795 def __init__( self, exprs, default=_optionalNotMatched ):
2796 super(Optional,self).__init__( exprs, savelist=False )
2796 super(Optional,self).__init__( exprs, savelist=False )
2797 self.defaultValue = default
2797 self.defaultValue = default
2798 self.mayReturnEmpty = True
2798 self.mayReturnEmpty = True
2799
2799
2800 def parseImpl( self, instring, loc, doActions=True ):
2800 def parseImpl( self, instring, loc, doActions=True ):
2801 try:
2801 try:
2802 loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )
2802 loc, tokens = self.expr._parse( instring, loc, doActions, callPreParse=False )
2803 except (ParseException,IndexError):
2803 except (ParseException,IndexError):
2804 if self.defaultValue is not _optionalNotMatched:
2804 if self.defaultValue is not _optionalNotMatched:
2805 if self.expr.resultsName:
2805 if self.expr.resultsName:
2806 tokens = ParseResults([ self.defaultValue ])
2806 tokens = ParseResults([ self.defaultValue ])
2807 tokens[self.expr.resultsName] = self.defaultValue
2807 tokens[self.expr.resultsName] = self.defaultValue
2808 else:
2808 else:
2809 tokens = [ self.defaultValue ]
2809 tokens = [ self.defaultValue ]
2810 else:
2810 else:
2811 tokens = []
2811 tokens = []
2812 return loc, tokens
2812 return loc, tokens
2813
2813
2814 def __str__( self ):
2814 def __str__( self ):
2815 if hasattr(self,"name"):
2815 if hasattr(self,"name"):
2816 return self.name
2816 return self.name
2817
2817
2818 if self.strRepr is None:
2818 if self.strRepr is None:
2819 self.strRepr = "[" + _ustr(self.expr) + "]"
2819 self.strRepr = "[" + _ustr(self.expr) + "]"
2820
2820
2821 return self.strRepr
2821 return self.strRepr
2822
2822
2823
2823
2824 class SkipTo(ParseElementEnhance):
2824 class SkipTo(ParseElementEnhance):
2825 """Token for skipping over all undefined text until the matched expression is found.
2825 """Token for skipping over all undefined text until the matched expression is found.
2826 If include is set to true, the matched expression is also parsed (the skipped text
2826 If include is set to true, the matched expression is also parsed (the skipped text
2827 and matched expression are returned as a 2-element list). The ignore
2827 and matched expression are returned as a 2-element list). The ignore
2828 argument is used to define grammars (typically quoted strings and comments) that
2828 argument is used to define grammars (typically quoted strings and comments) that
2829 might contain false matches.
2829 might contain false matches.
2830 """
2830 """
2831 def __init__( self, other, include=False, ignore=None, failOn=None ):
2831 def __init__( self, other, include=False, ignore=None, failOn=None ):
2832 super( SkipTo, self ).__init__( other )
2832 super( SkipTo, self ).__init__( other )
2833 self.ignoreExpr = ignore
2833 self.ignoreExpr = ignore
2834 self.mayReturnEmpty = True
2834 self.mayReturnEmpty = True
2835 self.mayIndexError = False
2835 self.mayIndexError = False
2836 self.includeMatch = include
2836 self.includeMatch = include
2837 self.asList = False
2837 self.asList = False
2838 if failOn is not None and isinstance(failOn, basestring):
2838 if failOn is not None and isinstance(failOn, basestring):
2839 self.failOn = Literal(failOn)
2839 self.failOn = Literal(failOn)
2840 else:
2840 else:
2841 self.failOn = failOn
2841 self.failOn = failOn
2842 self.errmsg = "No match found for "+_ustr(self.expr)
2842 self.errmsg = "No match found for "+_ustr(self.expr)
2843 #self.myException = ParseException("",0,self.errmsg,self)
2843 #self.myException = ParseException("",0,self.errmsg,self)
2844
2844
2845 def parseImpl( self, instring, loc, doActions=True ):
2845 def parseImpl( self, instring, loc, doActions=True ):
2846 startLoc = loc
2846 startLoc = loc
2847 instrlen = len(instring)
2847 instrlen = len(instring)
2848 expr = self.expr
2848 expr = self.expr
2849 failParse = False
2849 failParse = False
2850 while loc <= instrlen:
2850 while loc <= instrlen:
2851 try:
2851 try:
2852 if self.failOn:
2852 if self.failOn:
2853 try:
2853 try:
2854 self.failOn.tryParse(instring, loc)
2854 self.failOn.tryParse(instring, loc)
2855 except ParseBaseException:
2855 except ParseBaseException:
2856 pass
2856 pass
2857 else:
2857 else:
2858 failParse = True
2858 failParse = True
2859 raise ParseException(instring, loc, "Found expression " + str(self.failOn))
2859 raise ParseException(instring, loc, "Found expression " + str(self.failOn))
2860 failParse = False
2860 failParse = False
2861 if self.ignoreExpr is not None:
2861 if self.ignoreExpr is not None:
2862 while 1:
2862 while 1:
2863 try:
2863 try:
2864 loc = self.ignoreExpr.tryParse(instring,loc)
2864 loc = self.ignoreExpr.tryParse(instring,loc)
2865 print "found ignoreExpr, advance to", loc
2865 print "found ignoreExpr, advance to", loc
2866 except ParseBaseException:
2866 except ParseBaseException:
2867 break
2867 break
2868 expr._parse( instring, loc, doActions=False, callPreParse=False )
2868 expr._parse( instring, loc, doActions=False, callPreParse=False )
2869 skipText = instring[startLoc:loc]
2869 skipText = instring[startLoc:loc]
2870 if self.includeMatch:
2870 if self.includeMatch:
2871 loc,mat = expr._parse(instring,loc,doActions,callPreParse=False)
2871 loc,mat = expr._parse(instring,loc,doActions,callPreParse=False)
2872 if mat:
2872 if mat:
2873 skipRes = ParseResults( skipText )
2873 skipRes = ParseResults( skipText )
2874 skipRes += mat
2874 skipRes += mat
2875 return loc, [ skipRes ]
2875 return loc, [ skipRes ]
2876 else:
2876 else:
2877 return loc, [ skipText ]
2877 return loc, [ skipText ]
2878 else:
2878 else:
2879 return loc, [ skipText ]
2879 return loc, [ skipText ]
2880 except (ParseException,IndexError):
2880 except (ParseException,IndexError):
2881 if failParse:
2881 if failParse:
2882 raise
2882 raise
2883 else:
2883 else:
2884 loc += 1
2884 loc += 1
2885 exc = self.myException
2885 exc = self.myException
2886 exc.loc = loc
2886 exc.loc = loc
2887 exc.pstr = instring
2887 exc.pstr = instring
2888 raise exc
2888 raise exc
2889
2889
2890 class Forward(ParseElementEnhance):
2890 class Forward(ParseElementEnhance):
2891 """Forward declaration of an expression to be defined later -
2891 """Forward declaration of an expression to be defined later -
2892 used for recursive grammars, such as algebraic infix notation.
2892 used for recursive grammars, such as algebraic infix notation.
2893 When the expression is known, it is assigned to the Forward variable using the '<<' operator.
2893 When the expression is known, it is assigned to the Forward variable using the '<<' operator.
2894
2894
2895 Note: take care when assigning to Forward not to overlook precedence of operators.
2895 Note: take care when assigning to Forward not to overlook precedence of operators.
2896 Specifically, '|' has a lower precedence than '<<', so that::
2896 Specifically, '|' has a lower precedence than '<<', so that::
2897 fwdExpr << a | b | c
2897 fwdExpr << a | b | c
2898 will actually be evaluated as::
2898 will actually be evaluated as::
2899 (fwdExpr << a) | b | c
2899 (fwdExpr << a) | b | c
2900 thereby leaving b and c out as parseable alternatives. It is recommended that you
2900 thereby leaving b and c out as parseable alternatives. It is recommended that you
2901 explicitly group the values inserted into the Forward::
2901 explicitly group the values inserted into the Forward::
2902 fwdExpr << (a | b | c)
2902 fwdExpr << (a | b | c)
2903 """
2903 """
2904 def __init__( self, other=None ):
2904 def __init__( self, other=None ):
2905 super(Forward,self).__init__( other, savelist=False )
2905 super(Forward,self).__init__( other, savelist=False )
2906
2906
2907 def __lshift__( self, other ):
2907 def __lshift__( self, other ):
2908 if isinstance( other, basestring ):
2908 if isinstance( other, basestring ):
2909 other = Literal(other)
2909 other = Literal(other)
2910 self.expr = other
2910 self.expr = other
2911 self.mayReturnEmpty = other.mayReturnEmpty
2911 self.mayReturnEmpty = other.mayReturnEmpty
2912 self.strRepr = None
2912 self.strRepr = None
2913 self.mayIndexError = self.expr.mayIndexError
2913 self.mayIndexError = self.expr.mayIndexError
2914 self.mayReturnEmpty = self.expr.mayReturnEmpty
2914 self.mayReturnEmpty = self.expr.mayReturnEmpty
2915 self.setWhitespaceChars( self.expr.whiteChars )
2915 self.setWhitespaceChars( self.expr.whiteChars )
2916 self.skipWhitespace = self.expr.skipWhitespace
2916 self.skipWhitespace = self.expr.skipWhitespace
2917 self.saveAsList = self.expr.saveAsList
2917 self.saveAsList = self.expr.saveAsList
2918 self.ignoreExprs.extend(self.expr.ignoreExprs)
2918 self.ignoreExprs.extend(self.expr.ignoreExprs)
2919 return None
2919 return None
2920
2920
2921 def leaveWhitespace( self ):
2921 def leaveWhitespace( self ):
2922 self.skipWhitespace = False
2922 self.skipWhitespace = False
2923 return self
2923 return self
2924
2924
2925 def streamline( self ):
2925 def streamline( self ):
2926 if not self.streamlined:
2926 if not self.streamlined:
2927 self.streamlined = True
2927 self.streamlined = True
2928 if self.expr is not None:
2928 if self.expr is not None:
2929 self.expr.streamline()
2929 self.expr.streamline()
2930 return self
2930 return self
2931
2931
2932 def validate( self, validateTrace=[] ):
2932 def validate( self, validateTrace=[] ):
2933 if self not in validateTrace:
2933 if self not in validateTrace:
2934 tmp = validateTrace[:]+[self]
2934 tmp = validateTrace[:]+[self]
2935 if self.expr is not None:
2935 if self.expr is not None:
2936 self.expr.validate(tmp)
2936 self.expr.validate(tmp)
2937 self.checkRecursion([])
2937 self.checkRecursion([])
2938
2938
2939 def __str__( self ):
2939 def __str__( self ):
2940 if hasattr(self,"name"):
2940 if hasattr(self,"name"):
2941 return self.name
2941 return self.name
2942
2942
2943 self._revertClass = self.__class__
2943 self._revertClass = self.__class__
2944 self.__class__ = _ForwardNoRecurse
2944 self.__class__ = _ForwardNoRecurse
2945 try:
2945 try:
2946 if self.expr is not None:
2946 if self.expr is not None:
2947 retString = _ustr(self.expr)
2947 retString = _ustr(self.expr)
2948 else:
2948 else:
2949 retString = "None"
2949 retString = "None"
2950 finally:
2950 finally:
2951 self.__class__ = self._revertClass
2951 self.__class__ = self._revertClass
2952 return self.__class__.__name__ + ": " + retString
2952 return self.__class__.__name__ + ": " + retString
2953
2953
2954 def copy(self):
2954 def copy(self):
2955 if self.expr is not None:
2955 if self.expr is not None:
2956 return super(Forward,self).copy()
2956 return super(Forward,self).copy()
2957 else:
2957 else:
2958 ret = Forward()
2958 ret = Forward()
2959 ret << self
2959 ret << self
2960 return ret
2960 return ret
2961
2961
2962 class _ForwardNoRecurse(Forward):
2962 class _ForwardNoRecurse(Forward):
2963 def __str__( self ):
2963 def __str__( self ):
2964 return "..."
2964 return "..."
2965
2965
2966 class TokenConverter(ParseElementEnhance):
2966 class TokenConverter(ParseElementEnhance):
2967 """Abstract subclass of ParseExpression, for converting parsed results."""
2967 """Abstract subclass of ParseExpression, for converting parsed results."""
2968 def __init__( self, expr, savelist=False ):
2968 def __init__( self, expr, savelist=False ):
2969 super(TokenConverter,self).__init__( expr )#, savelist )
2969 super(TokenConverter,self).__init__( expr )#, savelist )
2970 self.saveAsList = False
2970 self.saveAsList = False
2971
2971
2972 class Upcase(TokenConverter):
2972 class Upcase(TokenConverter):
2973 """Converter to upper case all matching tokens."""
2973 """Converter to upper case all matching tokens."""
2974 def __init__(self, *args):
2974 def __init__(self, *args):
2975 super(Upcase,self).__init__(*args)
2975 super(Upcase,self).__init__(*args)
2976 warnings.warn("Upcase class is deprecated, use upcaseTokens parse action instead",
2976 warnings.warn("Upcase class is deprecated, use upcaseTokens parse action instead",
2977 DeprecationWarning,stacklevel=2)
2977 DeprecationWarning,stacklevel=2)
2978
2978
2979 def postParse( self, instring, loc, tokenlist ):
2979 def postParse( self, instring, loc, tokenlist ):
2980 return list(map( string.upper, tokenlist ))
2980 return list(map( string.upper, tokenlist ))
2981
2981
2982
2982
2983 class Combine(TokenConverter):
2983 class Combine(TokenConverter):
2984 """Converter to concatenate all matching tokens to a single string.
2984 """Converter to concatenate all matching tokens to a single string.
2985 By default, the matching patterns must also be contiguous in the input string;
2985 By default, the matching patterns must also be contiguous in the input string;
2986 this can be disabled by specifying 'adjacent=False' in the constructor.
2986 this can be disabled by specifying 'adjacent=False' in the constructor.
2987 """
2987 """
2988 def __init__( self, expr, joinString="", adjacent=True ):
2988 def __init__( self, expr, joinString="", adjacent=True ):
2989 super(Combine,self).__init__( expr )
2989 super(Combine,self).__init__( expr )
2990 # suppress whitespace-stripping in contained parse expressions, but re-enable it on the Combine itself
2990 # suppress whitespace-stripping in contained parse expressions, but re-enable it on the Combine itself
2991 if adjacent:
2991 if adjacent:
2992 self.leaveWhitespace()
2992 self.leaveWhitespace()
2993 self.adjacent = adjacent
2993 self.adjacent = adjacent
2994 self.skipWhitespace = True
2994 self.skipWhitespace = True
2995 self.joinString = joinString
2995 self.joinString = joinString
2996
2996
2997 def ignore( self, other ):
2997 def ignore( self, other ):
2998 if self.adjacent:
2998 if self.adjacent:
2999 ParserElement.ignore(self, other)
2999 ParserElement.ignore(self, other)
3000 else:
3000 else:
3001 super( Combine, self).ignore( other )
3001 super( Combine, self).ignore( other )
3002 return self
3002 return self
3003
3003
3004 def postParse( self, instring, loc, tokenlist ):
3004 def postParse( self, instring, loc, tokenlist ):
3005 retToks = tokenlist.copy()
3005 retToks = tokenlist.copy()
3006 del retToks[:]
3006 del retToks[:]
3007 retToks += ParseResults([ "".join(tokenlist._asStringList(self.joinString)) ], modal=self.modalResults)
3007 retToks += ParseResults([ "".join(tokenlist._asStringList(self.joinString)) ], modal=self.modalResults)
3008
3008
3009 if self.resultsName and len(retToks.keys())>0:
3009 if self.resultsName and len(retToks.keys())>0:
3010 return [ retToks ]
3010 return [ retToks ]
3011 else:
3011 else:
3012 return retToks
3012 return retToks
3013
3013
3014 class Group(TokenConverter):
3014 class Group(TokenConverter):
3015 """Converter to return the matched tokens as a list - useful for returning tokens of ZeroOrMore and OneOrMore expressions."""
3015 """Converter to return the matched tokens as a list - useful for returning tokens of ZeroOrMore and OneOrMore expressions."""
3016 def __init__( self, expr ):
3016 def __init__( self, expr ):
3017 super(Group,self).__init__( expr )
3017 super(Group,self).__init__( expr )
3018 self.saveAsList = True
3018 self.saveAsList = True
3019
3019
3020 def postParse( self, instring, loc, tokenlist ):
3020 def postParse( self, instring, loc, tokenlist ):
3021 return [ tokenlist ]
3021 return [ tokenlist ]
3022
3022
3023 class Dict(TokenConverter):
3023 class Dict(TokenConverter):
3024 """Converter to return a repetitive expression as a list, but also as a dictionary.
3024 """Converter to return a repetitive expression as a list, but also as a dictionary.
3025 Each element can also be referenced using the first token in the expression as its key.
3025 Each element can also be referenced using the first token in the expression as its key.
3026 Useful for tabular report scraping when the first column can be used as a item key.
3026 Useful for tabular report scraping when the first column can be used as a item key.
3027 """
3027 """
3028 def __init__( self, exprs ):
3028 def __init__( self, exprs ):
3029 super(Dict,self).__init__( exprs )
3029 super(Dict,self).__init__( exprs )
3030 self.saveAsList = True
3030 self.saveAsList = True
3031
3031
3032 def postParse( self, instring, loc, tokenlist ):
3032 def postParse( self, instring, loc, tokenlist ):
3033 for i,tok in enumerate(tokenlist):
3033 for i,tok in enumerate(tokenlist):
3034 if len(tok) == 0:
3034 if len(tok) == 0:
3035 continue
3035 continue
3036 ikey = tok[0]
3036 ikey = tok[0]
3037 if isinstance(ikey,int):
3037 if isinstance(ikey,int):
3038 ikey = _ustr(tok[0]).strip()
3038 ikey = _ustr(tok[0]).strip()
3039 if len(tok)==1:
3039 if len(tok)==1:
3040 tokenlist[ikey] = _ParseResultsWithOffset("",i)
3040 tokenlist[ikey] = _ParseResultsWithOffset("",i)
3041 elif len(tok)==2 and not isinstance(tok[1],ParseResults):
3041 elif len(tok)==2 and not isinstance(tok[1],ParseResults):
3042 tokenlist[ikey] = _ParseResultsWithOffset(tok[1],i)
3042 tokenlist[ikey] = _ParseResultsWithOffset(tok[1],i)
3043 else:
3043 else:
3044 dictvalue = tok.copy() #ParseResults(i)
3044 dictvalue = tok.copy() #ParseResults(i)
3045 del dictvalue[0]
3045 del dictvalue[0]
3046 if len(dictvalue)!= 1 or (isinstance(dictvalue,ParseResults) and dictvalue.keys()):
3046 if len(dictvalue)!= 1 or (isinstance(dictvalue,ParseResults) and dictvalue.keys()):
3047 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue,i)
3047 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue,i)
3048 else:
3048 else:
3049 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue[0],i)
3049 tokenlist[ikey] = _ParseResultsWithOffset(dictvalue[0],i)
3050
3050
3051 if self.resultsName:
3051 if self.resultsName:
3052 return [ tokenlist ]
3052 return [ tokenlist ]
3053 else:
3053 else:
3054 return tokenlist
3054 return tokenlist
3055
3055
3056
3056
3057 class Suppress(TokenConverter):
3057 class Suppress(TokenConverter):
3058 """Converter for ignoring the results of a parsed expression."""
3058 """Converter for ignoring the results of a parsed expression."""
3059 def postParse( self, instring, loc, tokenlist ):
3059 def postParse( self, instring, loc, tokenlist ):
3060 return []
3060 return []
3061
3061
3062 def suppress( self ):
3062 def suppress( self ):
3063 return self
3063 return self
3064
3064
3065
3065
3066 class OnlyOnce(object):
3066 class OnlyOnce(object):
3067 """Wrapper for parse actions, to ensure they are only called once."""
3067 """Wrapper for parse actions, to ensure they are only called once."""
3068 def __init__(self, methodCall):
3068 def __init__(self, methodCall):
3069 self.callable = ParserElement._normalizeParseActionArgs(methodCall)
3069 self.callable = ParserElement._normalizeParseActionArgs(methodCall)
3070 self.called = False
3070 self.called = False
3071 def __call__(self,s,l,t):
3071 def __call__(self,s,l,t):
3072 if not self.called:
3072 if not self.called:
3073 results = self.callable(s,l,t)
3073 results = self.callable(s,l,t)
3074 self.called = True
3074 self.called = True
3075 return results
3075 return results
3076 raise ParseException(s,l,"")
3076 raise ParseException(s,l,"")
3077 def reset(self):
3077 def reset(self):
3078 self.called = False
3078 self.called = False
3079
3079
3080 def traceParseAction(f):
3080 def traceParseAction(f):
3081 """Decorator for debugging parse actions."""
3081 """Decorator for debugging parse actions."""
3082 f = ParserElement._normalizeParseActionArgs(f)
3082 f = ParserElement._normalizeParseActionArgs(f)
3083 def z(*paArgs):
3083 def z(*paArgs):
3084 thisFunc = f.func_name
3084 thisFunc = f.func_name
3085 s,l,t = paArgs[-3:]
3085 s,l,t = paArgs[-3:]
3086 if len(paArgs)>3:
3086 if len(paArgs)>3:
3087 thisFunc = paArgs[0].__class__.__name__ + '.' + thisFunc
3087 thisFunc = paArgs[0].__class__.__name__ + '.' + thisFunc
3088 sys.stderr.write( ">>entering %s(line: '%s', %d, %s)\n" % (thisFunc,line(l,s),l,t) )
3088 sys.stderr.write( ">>entering %s(line: '%s', %d, %s)\n" % (thisFunc,line(l,s),l,t) )
3089 try:
3089 try:
3090 ret = f(*paArgs)
3090 ret = f(*paArgs)
3091 except Exception, exc:
3091 except Exception, exc:
3092 sys.stderr.write( "<<leaving %s (exception: %s)\n" % (thisFunc,exc) )
3092 sys.stderr.write( "<<leaving %s (exception: %s)\n" % (thisFunc,exc) )
3093 raise
3093 raise
3094 sys.stderr.write( "<<leaving %s (ret: %s)\n" % (thisFunc,ret) )
3094 sys.stderr.write( "<<leaving %s (ret: %s)\n" % (thisFunc,ret) )
3095 return ret
3095 return ret
3096 try:
3096 try:
3097 z.__name__ = f.__name__
3097 z.__name__ = f.__name__
3098 except AttributeError:
3098 except AttributeError:
3099 pass
3099 pass
3100 return z
3100 return z
3101
3101
3102 #
3102 #
3103 # global helpers
3103 # global helpers
3104 #
3104 #
3105 def delimitedList( expr, delim=",", combine=False ):
3105 def delimitedList( expr, delim=",", combine=False ):
3106 """Helper to define a delimited list of expressions - the delimiter defaults to ','.
3106 """Helper to define a delimited list of expressions - the delimiter defaults to ','.
3107 By default, the list elements and delimiters can have intervening whitespace, and
3107 By default, the list elements and delimiters can have intervening whitespace, and
3108 comments, but this can be overridden by passing 'combine=True' in the constructor.
3108 comments, but this can be overridden by passing 'combine=True' in the constructor.
3109 If combine is set to True, the matching tokens are returned as a single token
3109 If combine is set to True, the matching tokens are returned as a single token
3110 string, with the delimiters included; otherwise, the matching tokens are returned
3110 string, with the delimiters included; otherwise, the matching tokens are returned
3111 as a list of tokens, with the delimiters suppressed.
3111 as a list of tokens, with the delimiters suppressed.
3112 """
3112 """
3113 dlName = _ustr(expr)+" ["+_ustr(delim)+" "+_ustr(expr)+"]..."
3113 dlName = _ustr(expr)+" ["+_ustr(delim)+" "+_ustr(expr)+"]..."
3114 if combine:
3114 if combine:
3115 return Combine( expr + ZeroOrMore( delim + expr ) ).setName(dlName)
3115 return Combine( expr + ZeroOrMore( delim + expr ) ).setName(dlName)
3116 else:
3116 else:
3117 return ( expr + ZeroOrMore( Suppress( delim ) + expr ) ).setName(dlName)
3117 return ( expr + ZeroOrMore( Suppress( delim ) + expr ) ).setName(dlName)
3118
3118
3119 def countedArray( expr ):
3119 def countedArray( expr ):
3120 """Helper to define a counted list of expressions.
3120 """Helper to define a counted list of expressions.
3121 This helper defines a pattern of the form::
3121 This helper defines a pattern of the form::
3122 integer expr expr expr...
3122 integer expr expr expr...
3123 where the leading integer tells how many expr expressions follow.
3123 where the leading integer tells how many expr expressions follow.
3124 The matched tokens returns the array of expr tokens as a list - the leading count token is suppressed.
3124 The matched tokens returns the array of expr tokens as a list - the leading count token is suppressed.
3125 """
3125 """
3126 arrayExpr = Forward()
3126 arrayExpr = Forward()
3127 def countFieldParseAction(s,l,t):
3127 def countFieldParseAction(s,l,t):
3128 n = int(t[0])
3128 n = int(t[0])
3129 arrayExpr << (n and Group(And([expr]*n)) or Group(empty))
3129 arrayExpr << (n and Group(And([expr]*n)) or Group(empty))
3130 return []
3130 return []
3131 return ( Word(nums).setName("arrayLen").setParseAction(countFieldParseAction, callDuringTry=True) + arrayExpr )
3131 return ( Word(nums).setName("arrayLen").setParseAction(countFieldParseAction, callDuringTry=True) + arrayExpr )
3132
3132
3133 def _flatten(L):
3133 def _flatten(L):
3134 if type(L) is not list: return [L]
3134 if type(L) is not list: return [L]
3135 if L == []: return L
3135 if L == []: return L
3136 return _flatten(L[0]) + _flatten(L[1:])
3136 return _flatten(L[0]) + _flatten(L[1:])
3137
3137
3138 def matchPreviousLiteral(expr):
3138 def matchPreviousLiteral(expr):
3139 """Helper to define an expression that is indirectly defined from
3139 """Helper to define an expression that is indirectly defined from
3140 the tokens matched in a previous expression, that is, it looks
3140 the tokens matched in a previous expression, that is, it looks
3141 for a 'repeat' of a previous expression. For example::
3141 for a 'repeat' of a previous expression. For example::
3142 first = Word(nums)
3142 first = Word(nums)
3143 second = matchPreviousLiteral(first)
3143 second = matchPreviousLiteral(first)
3144 matchExpr = first + ":" + second
3144 matchExpr = first + ":" + second
3145 will match "1:1", but not "1:2". Because this matches a
3145 will match "1:1", but not "1:2". Because this matches a
3146 previous literal, will also match the leading "1:1" in "1:10".
3146 previous literal, will also match the leading "1:1" in "1:10".
3147 If this is not desired, use matchPreviousExpr.
3147 If this is not desired, use matchPreviousExpr.
3148 Do *not* use with packrat parsing enabled.
3148 Do *not* use with packrat parsing enabled.
3149 """
3149 """
3150 rep = Forward()
3150 rep = Forward()
3151 def copyTokenToRepeater(s,l,t):
3151 def copyTokenToRepeater(s,l,t):
3152 if t:
3152 if t:
3153 if len(t) == 1:
3153 if len(t) == 1:
3154 rep << t[0]
3154 rep << t[0]
3155 else:
3155 else:
3156 # flatten t tokens
3156 # flatten t tokens
3157 tflat = _flatten(t.asList())
3157 tflat = _flatten(t.asList())
3158 rep << And( [ Literal(tt) for tt in tflat ] )
3158 rep << And( [ Literal(tt) for tt in tflat ] )
3159 else:
3159 else:
3160 rep << Empty()
3160 rep << Empty()
3161 expr.addParseAction(copyTokenToRepeater, callDuringTry=True)
3161 expr.addParseAction(copyTokenToRepeater, callDuringTry=True)
3162 return rep
3162 return rep
3163
3163
3164 def matchPreviousExpr(expr):
3164 def matchPreviousExpr(expr):
3165 """Helper to define an expression that is indirectly defined from
3165 """Helper to define an expression that is indirectly defined from
3166 the tokens matched in a previous expression, that is, it looks
3166 the tokens matched in a previous expression, that is, it looks
3167 for a 'repeat' of a previous expression. For example::
3167 for a 'repeat' of a previous expression. For example::
3168 first = Word(nums)
3168 first = Word(nums)
3169 second = matchPreviousExpr(first)
3169 second = matchPreviousExpr(first)
3170 matchExpr = first + ":" + second
3170 matchExpr = first + ":" + second
3171 will match "1:1", but not "1:2". Because this matches by
3171 will match "1:1", but not "1:2". Because this matches by
3172 expressions, will *not* match the leading "1:1" in "1:10";
3172 expressions, will *not* match the leading "1:1" in "1:10";
3173 the expressions are evaluated first, and then compared, so
3173 the expressions are evaluated first, and then compared, so
3174 "1" is compared with "10".
3174 "1" is compared with "10".
3175 Do *not* use with packrat parsing enabled.
3175 Do *not* use with packrat parsing enabled.
3176 """
3176 """
3177 rep = Forward()
3177 rep = Forward()
3178 e2 = expr.copy()
3178 e2 = expr.copy()
3179 rep << e2
3179 rep << e2
3180 def copyTokenToRepeater(s,l,t):
3180 def copyTokenToRepeater(s,l,t):
3181 matchTokens = _flatten(t.asList())
3181 matchTokens = _flatten(t.asList())
3182 def mustMatchTheseTokens(s,l,t):
3182 def mustMatchTheseTokens(s,l,t):
3183 theseTokens = _flatten(t.asList())
3183 theseTokens = _flatten(t.asList())
3184 if theseTokens != matchTokens:
3184 if theseTokens != matchTokens:
3185 raise ParseException("",0,"")
3185 raise ParseException("",0,"")
3186 rep.setParseAction( mustMatchTheseTokens, callDuringTry=True )
3186 rep.setParseAction( mustMatchTheseTokens, callDuringTry=True )
3187 expr.addParseAction(copyTokenToRepeater, callDuringTry=True)
3187 expr.addParseAction(copyTokenToRepeater, callDuringTry=True)
3188 return rep
3188 return rep
3189
3189
3190 def _escapeRegexRangeChars(s):
3190 def _escapeRegexRangeChars(s):
3191 #~ escape these chars: ^-]
3191 #~ escape these chars: ^-]
3192 for c in r"\^-]":
3192 for c in r"\^-]":
3193 s = s.replace(c,_bslash+c)
3193 s = s.replace(c,_bslash+c)
3194 s = s.replace("\n",r"\n")
3194 s = s.replace("\n",r"\n")
3195 s = s.replace("\t",r"\t")
3195 s = s.replace("\t",r"\t")
3196 return _ustr(s)
3196 return _ustr(s)
3197
3197
3198 def oneOf( strs, caseless=False, useRegex=True ):
3198 def oneOf( strs, caseless=False, useRegex=True ):
3199 """Helper to quickly define a set of alternative Literals, and makes sure to do
3199 """Helper to quickly define a set of alternative Literals, and makes sure to do
3200 longest-first testing when there is a conflict, regardless of the input order,
3200 longest-first testing when there is a conflict, regardless of the input order,
3201 but returns a MatchFirst for best performance.
3201 but returns a MatchFirst for best performance.
3202
3202
3203 Parameters:
3203 Parameters:
3204 - strs - a string of space-delimited literals, or a list of string literals
3204 - strs - a string of space-delimited literals, or a list of string literals
3205 - caseless - (default=False) - treat all literals as caseless
3205 - caseless - (default=False) - treat all literals as caseless
3206 - useRegex - (default=True) - as an optimization, will generate a Regex
3206 - useRegex - (default=True) - as an optimization, will generate a Regex
3207 object; otherwise, will generate a MatchFirst object (if caseless=True, or
3207 object; otherwise, will generate a MatchFirst object (if caseless=True, or
3208 if creating a Regex raises an exception)
3208 if creating a Regex raises an exception)
3209 """
3209 """
3210 if caseless:
3210 if caseless:
3211 isequal = ( lambda a,b: a.upper() == b.upper() )
3211 isequal = ( lambda a,b: a.upper() == b.upper() )
3212 masks = ( lambda a,b: b.upper().startswith(a.upper()) )
3212 masks = ( lambda a,b: b.upper().startswith(a.upper()) )
3213 parseElementClass = CaselessLiteral
3213 parseElementClass = CaselessLiteral
3214 else:
3214 else:
3215 isequal = ( lambda a,b: a == b )
3215 isequal = ( lambda a,b: a == b )
3216 masks = ( lambda a,b: b.startswith(a) )
3216 masks = ( lambda a,b: b.startswith(a) )
3217 parseElementClass = Literal
3217 parseElementClass = Literal
3218
3218
3219 if isinstance(strs,(list,tuple)):
3219 if isinstance(strs,(list,tuple)):
3220 symbols = list(strs[:])
3220 symbols = list(strs[:])
3221 elif isinstance(strs,basestring):
3221 elif isinstance(strs,basestring):
3222 symbols = strs.split()
3222 symbols = strs.split()
3223 else:
3223 else:
3224 warnings.warn("Invalid argument to oneOf, expected string or list",
3224 warnings.warn("Invalid argument to oneOf, expected string or list",
3225 SyntaxWarning, stacklevel=2)
3225 SyntaxWarning, stacklevel=2)
3226
3226
3227 i = 0
3227 i = 0
3228 while i < len(symbols)-1:
3228 while i < len(symbols)-1:
3229 cur = symbols[i]
3229 cur = symbols[i]
3230 for j,other in enumerate(symbols[i+1:]):
3230 for j,other in enumerate(symbols[i+1:]):
3231 if ( isequal(other, cur) ):
3231 if ( isequal(other, cur) ):
3232 del symbols[i+j+1]
3232 del symbols[i+j+1]
3233 break
3233 break
3234 elif ( masks(cur, other) ):
3234 elif ( masks(cur, other) ):
3235 del symbols[i+j+1]
3235 del symbols[i+j+1]
3236 symbols.insert(i,other)
3236 symbols.insert(i,other)
3237 cur = other
3237 cur = other
3238 break
3238 break
3239 else:
3239 else:
3240 i += 1
3240 i += 1
3241
3241
3242 if not caseless and useRegex:
3242 if not caseless and useRegex:
3243 #~ print (strs,"->", "|".join( [ _escapeRegexChars(sym) for sym in symbols] ))
3243 #~ print (strs,"->", "|".join( [ _escapeRegexChars(sym) for sym in symbols] ))
3244 try:
3244 try:
3245 if len(symbols)==len("".join(symbols)):
3245 if len(symbols)==len("".join(symbols)):
3246 return Regex( "[%s]" % "".join( [ _escapeRegexRangeChars(sym) for sym in symbols] ) )
3246 return Regex( "[%s]" % "".join( [ _escapeRegexRangeChars(sym) for sym in symbols] ) )
3247 else:
3247 else:
3248 return Regex( "|".join( [ re.escape(sym) for sym in symbols] ) )
3248 return Regex( "|".join( [ re.escape(sym) for sym in symbols] ) )
3249 except:
3249 except:
3250 warnings.warn("Exception creating Regex for oneOf, building MatchFirst",
3250 warnings.warn("Exception creating Regex for oneOf, building MatchFirst",
3251 SyntaxWarning, stacklevel=2)
3251 SyntaxWarning, stacklevel=2)
3252
3252
3253
3253
3254 # last resort, just use MatchFirst
3254 # last resort, just use MatchFirst
3255 return MatchFirst( [ parseElementClass(sym) for sym in symbols ] )
3255 return MatchFirst( [ parseElementClass(sym) for sym in symbols ] )
3256
3256
3257 def dictOf( key, value ):
3257 def dictOf( key, value ):
3258 """Helper to easily and clearly define a dictionary by specifying the respective patterns
3258 """Helper to easily and clearly define a dictionary by specifying the respective patterns
3259 for the key and value. Takes care of defining the Dict, ZeroOrMore, and Group tokens
3259 for the key and value. Takes care of defining the Dict, ZeroOrMore, and Group tokens
3260 in the proper order. The key pattern can include delimiting markers or punctuation,
3260 in the proper order. The key pattern can include delimiting markers or punctuation,
3261 as long as they are suppressed, thereby leaving the significant key text. The value
3261 as long as they are suppressed, thereby leaving the significant key text. The value
3262 pattern can include named results, so that the Dict results can include named token
3262 pattern can include named results, so that the Dict results can include named token
3263 fields.
3263 fields.
3264 """
3264 """
3265 return Dict( ZeroOrMore( Group ( key + value ) ) )
3265 return Dict( ZeroOrMore( Group ( key + value ) ) )
3266
3266
3267 def originalTextFor(expr, asString=True):
3267 def originalTextFor(expr, asString=True):
3268 """Helper to return the original, untokenized text for a given expression. Useful to
3268 """Helper to return the original, untokenized text for a given expression. Useful to
3269 restore the parsed fields of an HTML start tag into the raw tag text itself, or to
3269 restore the parsed fields of an HTML start tag into the raw tag text itself, or to
3270 revert separate tokens with intervening whitespace back to the original matching
3270 revert separate tokens with intervening whitespace back to the original matching
3271 input text. Simpler to use than the parse action keepOriginalText, and does not
3271 input text. Simpler to use than the parse action keepOriginalText, and does not
3272 require the inspect module to chase up the call stack. By default, returns a
3272 require the inspect module to chase up the call stack. By default, returns a
3273 string containing the original parsed text.
3273 string containing the original parsed text.
3274
3274
3275 If the optional asString argument is passed as False, then the return value is a
3275 If the optional asString argument is passed as False, then the return value is a
3276 ParseResults containing any results names that were originally matched, and a
3276 ParseResults containing any results names that were originally matched, and a
3277 single token containing the original matched text from the input string. So if
3277 single token containing the original matched text from the input string. So if
3278 the expression passed to originalTextFor contains expressions with defined
3278 the expression passed to originalTextFor contains expressions with defined
3279 results names, you must set asString to False if you want to preserve those
3279 results names, you must set asString to False if you want to preserve those
3280 results name values."""
3280 results name values."""
3281 locMarker = Empty().setParseAction(lambda s,loc,t: loc)
3281 locMarker = Empty().setParseAction(lambda s,loc,t: loc)
3282 matchExpr = locMarker("_original_start") + expr + locMarker("_original_end")
3282 matchExpr = locMarker("_original_start") + expr + locMarker("_original_end")
3283 if asString:
3283 if asString:
3284 extractText = lambda s,l,t: s[t._original_start:t._original_end]
3284 extractText = lambda s,l,t: s[t._original_start:t._original_end]
3285 else:
3285 else:
3286 def extractText(s,l,t):
3286 def extractText(s,l,t):
3287 del t[:]
3287 del t[:]
3288 t.insert(0, s[t._original_start:t._original_end])
3288 t.insert(0, s[t._original_start:t._original_end])
3289 del t["_original_start"]
3289 del t["_original_start"]
3290 del t["_original_end"]
3290 del t["_original_end"]
3291 matchExpr.setParseAction(extractText)
3291 matchExpr.setParseAction(extractText)
3292 return matchExpr
3292 return matchExpr
3293
3293
3294 # convenience constants for positional expressions
3294 # convenience constants for positional expressions
3295 empty = Empty().setName("empty")
3295 empty = Empty().setName("empty")
3296 lineStart = LineStart().setName("lineStart")
3296 lineStart = LineStart().setName("lineStart")
3297 lineEnd = LineEnd().setName("lineEnd")
3297 lineEnd = LineEnd().setName("lineEnd")
3298 stringStart = StringStart().setName("stringStart")
3298 stringStart = StringStart().setName("stringStart")
3299 stringEnd = StringEnd().setName("stringEnd")
3299 stringEnd = StringEnd().setName("stringEnd")
3300
3300
3301 _escapedPunc = Word( _bslash, r"\[]-*.$+^?()~ ", exact=2 ).setParseAction(lambda s,l,t:t[0][1])
3301 _escapedPunc = Word( _bslash, r"\[]-*.$+^?()~ ", exact=2 ).setParseAction(lambda s,l,t:t[0][1])
3302 _printables_less_backslash = "".join([ c for c in printables if c not in r"\]" ])
3302 _printables_less_backslash = "".join([ c for c in printables if c not in r"\]" ])
3303 _escapedHexChar = Combine( Suppress(_bslash + "0x") + Word(hexnums) ).setParseAction(lambda s,l,t:unichr(int(t[0],16)))
3303 _escapedHexChar = Combine( Suppress(_bslash + "0x") + Word(hexnums) ).setParseAction(lambda s,l,t:unichr(int(t[0],16)))
3304 _escapedOctChar = Combine( Suppress(_bslash) + Word("0","01234567") ).setParseAction(lambda s,l,t:unichr(int(t[0],8)))
3304 _escapedOctChar = Combine( Suppress(_bslash) + Word("0","01234567") ).setParseAction(lambda s,l,t:unichr(int(t[0],8)))
3305 _singleChar = _escapedPunc | _escapedHexChar | _escapedOctChar | Word(_printables_less_backslash,exact=1)
3305 _singleChar = _escapedPunc | _escapedHexChar | _escapedOctChar | Word(_printables_less_backslash,exact=1)
3306 _charRange = Group(_singleChar + Suppress("-") + _singleChar)
3306 _charRange = Group(_singleChar + Suppress("-") + _singleChar)
3307 _reBracketExpr = Literal("[") + Optional("^").setResultsName("negate") + Group( OneOrMore( _charRange | _singleChar ) ).setResultsName("body") + "]"
3307 _reBracketExpr = Literal("[") + Optional("^").setResultsName("negate") + Group( OneOrMore( _charRange | _singleChar ) ).setResultsName("body") + "]"
3308
3308
3309 _expanded = lambda p: (isinstance(p,ParseResults) and ''.join([ unichr(c) for c in range(ord(p[0]),ord(p[1])+1) ]) or p)
3309 _expanded = lambda p: (isinstance(p,ParseResults) and ''.join([ unichr(c) for c in range(ord(p[0]),ord(p[1])+1) ]) or p)
3310
3310
3311 def srange(s):
3311 def srange(s):
3312 r"""Helper to easily define string ranges for use in Word construction. Borrows
3312 r"""Helper to easily define string ranges for use in Word construction. Borrows
3313 syntax from regexp '[]' string range definitions::
3313 syntax from regexp '[]' string range definitions::
3314 srange("[0-9]") -> "0123456789"
3314 srange("[0-9]") -> "0123456789"
3315 srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz"
3315 srange("[a-z]") -> "abcdefghijklmnopqrstuvwxyz"
3316 srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_"
3316 srange("[a-z$_]") -> "abcdefghijklmnopqrstuvwxyz$_"
3317 The input string must be enclosed in []'s, and the returned string is the expanded
3317 The input string must be enclosed in []'s, and the returned string is the expanded
3318 character set joined into a single string.
3318 character set joined into a single string.
3319 The values enclosed in the []'s may be::
3319 The values enclosed in the []'s may be::
3320 a single character
3320 a single character
3321 an escaped character with a leading backslash (such as \- or \])
3321 an escaped character with a leading backslash (such as \- or \])
3322 an escaped hex character with a leading '\0x' (\0x21, which is a '!' character)
3322 an escaped hex character with a leading '\0x' (\0x21, which is a '!' character)
3323 an escaped octal character with a leading '\0' (\041, which is a '!' character)
3323 an escaped octal character with a leading '\0' (\041, which is a '!' character)
3324 a range of any of the above, separated by a dash ('a-z', etc.)
3324 a range of any of the above, separated by a dash ('a-z', etc.)
3325 any combination of the above ('aeiouy', 'a-zA-Z0-9_$', etc.)
3325 any combination of the above ('aeiouy', 'a-zA-Z0-9_$', etc.)
3326 """
3326 """
3327 try:
3327 try:
3328 return "".join([_expanded(part) for part in _reBracketExpr.parseString(s).body])
3328 return "".join([_expanded(part) for part in _reBracketExpr.parseString(s).body])
3329 except:
3329 except:
3330 return ""
3330 return ""
3331
3331
3332 def matchOnlyAtCol(n):
3332 def matchOnlyAtCol(n):
3333 """Helper method for defining parse actions that require matching at a specific
3333 """Helper method for defining parse actions that require matching at a specific
3334 column in the input text.
3334 column in the input text.
3335 """
3335 """
3336 def verifyCol(strg,locn,toks):
3336 def verifyCol(strg,locn,toks):
3337 if col(locn,strg) != n:
3337 if col(locn,strg) != n:
3338 raise ParseException(strg,locn,"matched token not at column %d" % n)
3338 raise ParseException(strg,locn,"matched token not at column %d" % n)
3339 return verifyCol
3339 return verifyCol
3340
3340
3341 def replaceWith(replStr):
3341 def replaceWith(replStr):
3342 """Helper method for common parse actions that simply return a literal value. Especially
3342 """Helper method for common parse actions that simply return a literal value. Especially
3343 useful when used with transformString().
3343 useful when used with transformString().
3344 """
3344 """
3345 def _replFunc(*args):
3345 def _replFunc(*args):
3346 return [replStr]
3346 return [replStr]
3347 return _replFunc
3347 return _replFunc
3348
3348
3349 def removeQuotes(s,l,t):
3349 def removeQuotes(s,l,t):
3350 """Helper parse action for removing quotation marks from parsed quoted strings.
3350 """Helper parse action for removing quotation marks from parsed quoted strings.
3351 To use, add this parse action to quoted string using::
3351 To use, add this parse action to quoted string using::
3352 quotedString.setParseAction( removeQuotes )
3352 quotedString.setParseAction( removeQuotes )
3353 """
3353 """
3354 return t[0][1:-1]
3354 return t[0][1:-1]
3355
3355
3356 def upcaseTokens(s,l,t):
3356 def upcaseTokens(s,l,t):
3357 """Helper parse action to convert tokens to upper case."""
3357 """Helper parse action to convert tokens to upper case."""
3358 return [ tt.upper() for tt in map(_ustr,t) ]
3358 return [ tt.upper() for tt in map(_ustr,t) ]
3359
3359
3360 def downcaseTokens(s,l,t):
3360 def downcaseTokens(s,l,t):
3361 """Helper parse action to convert tokens to lower case."""
3361 """Helper parse action to convert tokens to lower case."""
3362 return [ tt.lower() for tt in map(_ustr,t) ]
3362 return [ tt.lower() for tt in map(_ustr,t) ]
3363
3363
3364 def keepOriginalText(s,startLoc,t):
3364 def keepOriginalText(s,startLoc,t):
3365 """Helper parse action to preserve original parsed text,
3365 """Helper parse action to preserve original parsed text,
3366 overriding any nested parse actions."""
3366 overriding any nested parse actions."""
3367 try:
3367 try:
3368 endloc = getTokensEndLoc()
3368 endloc = getTokensEndLoc()
3369 except ParseException:
3369 except ParseException:
3370 raise ParseFatalException("incorrect usage of keepOriginalText - may only be called as a parse action")
3370 raise ParseFatalException("incorrect usage of keepOriginalText - may only be called as a parse action")
3371 del t[:]
3371 del t[:]
3372 t += ParseResults(s[startLoc:endloc])
3372 t += ParseResults(s[startLoc:endloc])
3373 return t
3373 return t
3374
3374
3375 def getTokensEndLoc():
3375 def getTokensEndLoc():
3376 """Method to be called from within a parse action to determine the end
3376 """Method to be called from within a parse action to determine the end
3377 location of the parsed tokens."""
3377 location of the parsed tokens."""
3378 import inspect
3378 import inspect
3379 fstack = inspect.stack()
3379 fstack = inspect.stack()
3380 try:
3380 try:
3381 # search up the stack (through intervening argument normalizers) for correct calling routine
3381 # search up the stack (through intervening argument normalizers) for correct calling routine
3382 for f in fstack[2:]:
3382 for f in fstack[2:]:
3383 if f[3] == "_parseNoCache":
3383 if f[3] == "_parseNoCache":
3384 endloc = f[0].f_locals["loc"]
3384 endloc = f[0].f_locals["loc"]
3385 return endloc
3385 return endloc
3386 else:
3386 else:
3387 raise ParseFatalException("incorrect usage of getTokensEndLoc - may only be called from within a parse action")
3387 raise ParseFatalException("incorrect usage of getTokensEndLoc - may only be called from within a parse action")
3388 finally:
3388 finally:
3389 del fstack
3389 del fstack
3390
3390
3391 def _makeTags(tagStr, xml):
3391 def _makeTags(tagStr, xml):
3392 """Internal helper to construct opening and closing tag expressions, given a tag name"""
3392 """Internal helper to construct opening and closing tag expressions, given a tag name"""
3393 if isinstance(tagStr,basestring):
3393 if isinstance(tagStr,basestring):
3394 resname = tagStr
3394 resname = tagStr
3395 tagStr = Keyword(tagStr, caseless=not xml)
3395 tagStr = Keyword(tagStr, caseless=not xml)
3396 else:
3396 else:
3397 resname = tagStr.name
3397 resname = tagStr.name
3398
3398
3399 tagAttrName = Word(alphas,alphanums+"_-:")
3399 tagAttrName = Word(alphas,alphanums+"_-:")
3400 if (xml):
3400 if (xml):
3401 tagAttrValue = dblQuotedString.copy().setParseAction( removeQuotes )
3401 tagAttrValue = dblQuotedString.copy().setParseAction( removeQuotes )
3402 openTag = Suppress("<") + tagStr + \
3402 openTag = Suppress("<") + tagStr + \
3403 Dict(ZeroOrMore(Group( tagAttrName + Suppress("=") + tagAttrValue ))) + \
3403 Dict(ZeroOrMore(Group( tagAttrName + Suppress("=") + tagAttrValue ))) + \
3404 Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">")
3404 Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">")
3405 else:
3405 else:
3406 printablesLessRAbrack = "".join( [ c for c in printables if c not in ">" ] )
3406 printablesLessRAbrack = "".join( [ c for c in printables if c not in ">" ] )
3407 tagAttrValue = quotedString.copy().setParseAction( removeQuotes ) | Word(printablesLessRAbrack)
3407 tagAttrValue = quotedString.copy().setParseAction( removeQuotes ) | Word(printablesLessRAbrack)
3408 openTag = Suppress("<") + tagStr + \
3408 openTag = Suppress("<") + tagStr + \
3409 Dict(ZeroOrMore(Group( tagAttrName.setParseAction(downcaseTokens) + \
3409 Dict(ZeroOrMore(Group( tagAttrName.setParseAction(downcaseTokens) + \
3410 Optional( Suppress("=") + tagAttrValue ) ))) + \
3410 Optional( Suppress("=") + tagAttrValue ) ))) + \
3411 Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">")
3411 Optional("/",default=[False]).setResultsName("empty").setParseAction(lambda s,l,t:t[0]=='/') + Suppress(">")
3412 closeTag = Combine(_L("</") + tagStr + ">")
3412 closeTag = Combine(_L("</") + tagStr + ">")
3413
3413
3414 openTag = openTag.setResultsName("start"+"".join(resname.replace(":"," ").title().split())).setName("<%s>" % tagStr)
3414 openTag = openTag.setResultsName("start"+"".join(resname.replace(":"," ").title().split())).setName("<%s>" % tagStr)
3415 closeTag = closeTag.setResultsName("end"+"".join(resname.replace(":"," ").title().split())).setName("</%s>" % tagStr)
3415 closeTag = closeTag.setResultsName("end"+"".join(resname.replace(":"," ").title().split())).setName("</%s>" % tagStr)
3416
3416
3417 return openTag, closeTag
3417 return openTag, closeTag
3418
3418
3419 def makeHTMLTags(tagStr):
3419 def makeHTMLTags(tagStr):
3420 """Helper to construct opening and closing tag expressions for HTML, given a tag name"""
3420 """Helper to construct opening and closing tag expressions for HTML, given a tag name"""
3421 return _makeTags( tagStr, False )
3421 return _makeTags( tagStr, False )
3422
3422
3423 def makeXMLTags(tagStr):
3423 def makeXMLTags(tagStr):
3424 """Helper to construct opening and closing tag expressions for XML, given a tag name"""
3424 """Helper to construct opening and closing tag expressions for XML, given a tag name"""
3425 return _makeTags( tagStr, True )
3425 return _makeTags( tagStr, True )
3426
3426
3427 def withAttribute(*args,**attrDict):
3427 def withAttribute(*args,**attrDict):
3428 """Helper to create a validating parse action to be used with start tags created
3428 """Helper to create a validating parse action to be used with start tags created
3429 with makeXMLTags or makeHTMLTags. Use withAttribute to qualify a starting tag
3429 with makeXMLTags or makeHTMLTags. Use withAttribute to qualify a starting tag
3430 with a required attribute value, to avoid false matches on common tags such as
3430 with a required attribute value, to avoid false matches on common tags such as
3431 <TD> or <DIV>.
3431 <TD> or <DIV>.
3432
3432
3433 Call withAttribute with a series of attribute names and values. Specify the list
3433 Call withAttribute with a series of attribute names and values. Specify the list
3434 of filter attributes names and values as:
3434 of filter attributes names and values as:
3435 - keyword arguments, as in (class="Customer",align="right"), or
3435 - keyword arguments, as in (class="Customer",align="right"), or
3436 - a list of name-value tuples, as in ( ("ns1:class", "Customer"), ("ns2:align","right") )
3436 - a list of name-value tuples, as in ( ("ns1:class", "Customer"), ("ns2:align","right") )
3437 For attribute names with a namespace prefix, you must use the second form. Attribute
3437 For attribute names with a namespace prefix, you must use the second form. Attribute
3438 names are matched insensitive to upper/lower case.
3438 names are matched insensitive to upper/lower case.
3439
3439
3440 To verify that the attribute exists, but without specifying a value, pass
3440 To verify that the attribute exists, but without specifying a value, pass
3441 withAttribute.ANY_VALUE as the value.
3441 withAttribute.ANY_VALUE as the value.
3442 """
3442 """
3443 if args:
3443 if args:
3444 attrs = args[:]
3444 attrs = args[:]
3445 else:
3445 else:
3446 attrs = attrDict.iteritems()
3446 attrs = attrDict.iteritems()
3447 attrs = [(k,v) for k,v in attrs]
3447 attrs = [(k,v) for k,v in attrs]
3448 def pa(s,l,tokens):
3448 def pa(s,l,tokens):
3449 for attrName,attrValue in attrs:
3449 for attrName,attrValue in attrs:
3450 if attrName not in tokens:
3450 if attrName not in tokens:
3451 raise ParseException(s,l,"no matching attribute " + attrName)
3451 raise ParseException(s,l,"no matching attribute " + attrName)
3452 if attrValue != withAttribute.ANY_VALUE and tokens[attrName] != attrValue:
3452 if attrValue != withAttribute.ANY_VALUE and tokens[attrName] != attrValue:
3453 raise ParseException(s,l,"attribute '%s' has value '%s', must be '%s'" %
3453 raise ParseException(s,l,"attribute '%s' has value '%s', must be '%s'" %
3454 (attrName, tokens[attrName], attrValue))
3454 (attrName, tokens[attrName], attrValue))
3455 return pa
3455 return pa
3456 withAttribute.ANY_VALUE = object()
3456 withAttribute.ANY_VALUE = object()
3457
3457
3458 opAssoc = _Constants()
3458 opAssoc = _Constants()
3459 opAssoc.LEFT = object()
3459 opAssoc.LEFT = object()
3460 opAssoc.RIGHT = object()
3460 opAssoc.RIGHT = object()
3461
3461
3462 def operatorPrecedence( baseExpr, opList ):
3462 def operatorPrecedence( baseExpr, opList ):
3463 """Helper method for constructing grammars of expressions made up of
3463 """Helper method for constructing grammars of expressions made up of
3464 operators working in a precedence hierarchy. Operators may be unary or
3464 operators working in a precedence hierarchy. Operators may be unary or
3465 binary, left- or right-associative. Parse actions can also be attached
3465 binary, left- or right-associative. Parse actions can also be attached
3466 to operator expressions.
3466 to operator expressions.
3467
3467
3468 Parameters:
3468 Parameters:
3469 - baseExpr - expression representing the most basic element for the nested
3469 - baseExpr - expression representing the most basic element for the nested
3470 - opList - list of tuples, one for each operator precedence level in the
3470 - opList - list of tuples, one for each operator precedence level in the
3471 expression grammar; each tuple is of the form
3471 expression grammar; each tuple is of the form
3472 (opExpr, numTerms, rightLeftAssoc, parseAction), where:
3472 (opExpr, numTerms, rightLeftAssoc, parseAction), where:
3473 - opExpr is the pyparsing expression for the operator;
3473 - opExpr is the pyparsing expression for the operator;
3474 may also be a string, which will be converted to a Literal;
3474 may also be a string, which will be converted to a Literal;
3475 if numTerms is 3, opExpr is a tuple of two expressions, for the
3475 if numTerms is 3, opExpr is a tuple of two expressions, for the
3476 two operators separating the 3 terms
3476 two operators separating the 3 terms
3477 - numTerms is the number of terms for this operator (must
3477 - numTerms is the number of terms for this operator (must
3478 be 1, 2, or 3)
3478 be 1, 2, or 3)
3479 - rightLeftAssoc is the indicator whether the operator is
3479 - rightLeftAssoc is the indicator whether the operator is
3480 right or left associative, using the pyparsing-defined
3480 right or left associative, using the pyparsing-defined
3481 constants opAssoc.RIGHT and opAssoc.LEFT.
3481 constants opAssoc.RIGHT and opAssoc.LEFT.
3482 - parseAction is the parse action to be associated with
3482 - parseAction is the parse action to be associated with
3483 expressions matching this operator expression (the
3483 expressions matching this operator expression (the
3484 parse action tuple member may be omitted)
3484 parse action tuple member may be omitted)
3485 """
3485 """
3486 ret = Forward()
3486 ret = Forward()
3487 lastExpr = baseExpr | ( Suppress('(') + ret + Suppress(')') )
3487 lastExpr = baseExpr | ( Suppress('(') + ret + Suppress(')') )
3488 for i,operDef in enumerate(opList):
3488 for i,operDef in enumerate(opList):
3489 opExpr,arity,rightLeftAssoc,pa = (operDef + (None,))[:4]
3489 opExpr,arity,rightLeftAssoc,pa = (operDef + (None,))[:4]
3490 if arity == 3:
3490 if arity == 3:
3491 if opExpr is None or len(opExpr) != 2:
3491 if opExpr is None or len(opExpr) != 2:
3492 raise ValueError("if numterms=3, opExpr must be a tuple or list of two expressions")
3492 raise ValueError("if numterms=3, opExpr must be a tuple or list of two expressions")
3493 opExpr1, opExpr2 = opExpr
3493 opExpr1, opExpr2 = opExpr
3494 thisExpr = Forward()#.setName("expr%d" % i)
3494 thisExpr = Forward()#.setName("expr%d" % i)
3495 if rightLeftAssoc == opAssoc.LEFT:
3495 if rightLeftAssoc == opAssoc.LEFT:
3496 if arity == 1:
3496 if arity == 1:
3497 matchExpr = FollowedBy(lastExpr + opExpr) + Group( lastExpr + OneOrMore( opExpr ) )
3497 matchExpr = FollowedBy(lastExpr + opExpr) + Group( lastExpr + OneOrMore( opExpr ) )
3498 elif arity == 2:
3498 elif arity == 2:
3499 if opExpr is not None:
3499 if opExpr is not None:
3500 matchExpr = FollowedBy(lastExpr + opExpr + lastExpr) + Group( lastExpr + OneOrMore( opExpr + lastExpr ) )
3500 matchExpr = FollowedBy(lastExpr + opExpr + lastExpr) + Group( lastExpr + OneOrMore( opExpr + lastExpr ) )
3501 else:
3501 else:
3502 matchExpr = FollowedBy(lastExpr+lastExpr) + Group( lastExpr + OneOrMore(lastExpr) )
3502 matchExpr = FollowedBy(lastExpr+lastExpr) + Group( lastExpr + OneOrMore(lastExpr) )
3503 elif arity == 3:
3503 elif arity == 3:
3504 matchExpr = FollowedBy(lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr) + \
3504 matchExpr = FollowedBy(lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr) + \
3505 Group( lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr )
3505 Group( lastExpr + opExpr1 + lastExpr + opExpr2 + lastExpr )
3506 else:
3506 else:
3507 raise ValueError("operator must be unary (1), binary (2), or ternary (3)")
3507 raise ValueError("operator must be unary (1), binary (2), or ternary (3)")
3508 elif rightLeftAssoc == opAssoc.RIGHT:
3508 elif rightLeftAssoc == opAssoc.RIGHT:
3509 if arity == 1:
3509 if arity == 1:
3510 # try to avoid LR with this extra test
3510 # try to avoid LR with this extra test
3511 if not isinstance(opExpr, Optional):
3511 if not isinstance(opExpr, Optional):
3512 opExpr = Optional(opExpr)
3512 opExpr = Optional(opExpr)
3513 matchExpr = FollowedBy(opExpr.expr + thisExpr) + Group( opExpr + thisExpr )
3513 matchExpr = FollowedBy(opExpr.expr + thisExpr) + Group( opExpr + thisExpr )
3514 elif arity == 2:
3514 elif arity == 2:
3515 if opExpr is not None:
3515 if opExpr is not None:
3516 matchExpr = FollowedBy(lastExpr + opExpr + thisExpr) + Group( lastExpr + OneOrMore( opExpr + thisExpr ) )
3516 matchExpr = FollowedBy(lastExpr + opExpr + thisExpr) + Group( lastExpr + OneOrMore( opExpr + thisExpr ) )
3517 else:
3517 else:
3518 matchExpr = FollowedBy(lastExpr + thisExpr) + Group( lastExpr + OneOrMore( thisExpr ) )
3518 matchExpr = FollowedBy(lastExpr + thisExpr) + Group( lastExpr + OneOrMore( thisExpr ) )
3519 elif arity == 3:
3519 elif arity == 3:
3520 matchExpr = FollowedBy(lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr) + \
3520 matchExpr = FollowedBy(lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr) + \
3521 Group( lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr )
3521 Group( lastExpr + opExpr1 + thisExpr + opExpr2 + thisExpr )
3522 else:
3522 else:
3523 raise ValueError("operator must be unary (1), binary (2), or ternary (3)")
3523 raise ValueError("operator must be unary (1), binary (2), or ternary (3)")
3524 else:
3524 else:
3525 raise ValueError("operator must indicate right or left associativity")
3525 raise ValueError("operator must indicate right or left associativity")
3526 if pa:
3526 if pa:
3527 matchExpr.setParseAction( pa )
3527 matchExpr.setParseAction( pa )
3528 thisExpr << ( matchExpr | lastExpr )
3528 thisExpr << ( matchExpr | lastExpr )
3529 lastExpr = thisExpr
3529 lastExpr = thisExpr
3530 ret << lastExpr
3530 ret << lastExpr
3531 return ret
3531 return ret
3532
3532
3533 dblQuotedString = Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*"').setName("string enclosed in double quotes")
3533 dblQuotedString = Regex(r'"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*"').setName("string enclosed in double quotes")
3534 sglQuotedString = Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*'").setName("string enclosed in single quotes")
3534 sglQuotedString = Regex(r"'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*'").setName("string enclosed in single quotes")
3535 quotedString = Regex(r'''(?:"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*")|(?:'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*')''').setName("quotedString using single or double quotes")
3535 quotedString = Regex(r'''(?:"(?:[^"\n\r\\]|(?:"")|(?:\\x[0-9a-fA-F]+)|(?:\\.))*")|(?:'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*')''').setName("quotedString using single or double quotes")
3536 unicodeString = Combine(_L('u') + quotedString.copy())
3536 unicodeString = Combine(_L('u') + quotedString.copy())
3537
3537
3538 def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString):
3538 def nestedExpr(opener="(", closer=")", content=None, ignoreExpr=quotedString):
3539 """Helper method for defining nested lists enclosed in opening and closing
3539 """Helper method for defining nested lists enclosed in opening and closing
3540 delimiters ("(" and ")" are the default).
3540 delimiters ("(" and ")" are the default).
3541
3541
3542 Parameters:
3542 Parameters:
3543 - opener - opening character for a nested list (default="("); can also be a pyparsing expression
3543 - opener - opening character for a nested list (default="("); can also be a pyparsing expression
3544 - closer - closing character for a nested list (default=")"); can also be a pyparsing expression
3544 - closer - closing character for a nested list (default=")"); can also be a pyparsing expression
3545 - content - expression for items within the nested lists (default=None)
3545 - content - expression for items within the nested lists (default=None)
3546 - ignoreExpr - expression for ignoring opening and closing delimiters (default=quotedString)
3546 - ignoreExpr - expression for ignoring opening and closing delimiters (default=quotedString)
3547
3547
3548 If an expression is not provided for the content argument, the nested
3548 If an expression is not provided for the content argument, the nested
3549 expression will capture all whitespace-delimited content between delimiters
3549 expression will capture all whitespace-delimited content between delimiters
3550 as a list of separate values.
3550 as a list of separate values.
3551
3551
3552 Use the ignoreExpr argument to define expressions that may contain
3552 Use the ignoreExpr argument to define expressions that may contain
3553 opening or closing characters that should not be treated as opening
3553 opening or closing characters that should not be treated as opening
3554 or closing characters for nesting, such as quotedString or a comment
3554 or closing characters for nesting, such as quotedString or a comment
3555 expression. Specify multiple expressions using an Or or MatchFirst.
3555 expression. Specify multiple expressions using an Or or MatchFirst.
3556 The default is quotedString, but if no expressions are to be ignored,
3556 The default is quotedString, but if no expressions are to be ignored,
3557 then pass None for this argument.
3557 then pass None for this argument.
3558 """
3558 """
3559 if opener == closer:
3559 if opener == closer:
3560 raise ValueError("opening and closing strings cannot be the same")
3560 raise ValueError("opening and closing strings cannot be the same")
3561 if content is None:
3561 if content is None:
3562 if isinstance(opener,basestring) and isinstance(closer,basestring):
3562 if isinstance(opener,basestring) and isinstance(closer,basestring):
3563 if len(opener) == 1 and len(closer)==1:
3563 if len(opener) == 1 and len(closer)==1:
3564 if ignoreExpr is not None:
3564 if ignoreExpr is not None:
3565 content = (Combine(OneOrMore(~ignoreExpr +
3565 content = (Combine(OneOrMore(~ignoreExpr +
3566 CharsNotIn(opener+closer+ParserElement.DEFAULT_WHITE_CHARS,exact=1))
3566 CharsNotIn(opener+closer+ParserElement.DEFAULT_WHITE_CHARS,exact=1))
3567 ).setParseAction(lambda t:t[0].strip()))
3567 ).setParseAction(lambda t:t[0].strip()))
3568 else:
3568 else:
3569 content = (empty+CharsNotIn(opener+closer+ParserElement.DEFAULT_WHITE_CHARS
3569 content = (empty+CharsNotIn(opener+closer+ParserElement.DEFAULT_WHITE_CHARS
3570 ).setParseAction(lambda t:t[0].strip()))
3570 ).setParseAction(lambda t:t[0].strip()))
3571 else:
3571 else:
3572 if ignoreExpr is not None:
3572 if ignoreExpr is not None:
3573 content = (Combine(OneOrMore(~ignoreExpr +
3573 content = (Combine(OneOrMore(~ignoreExpr +
3574 ~Literal(opener) + ~Literal(closer) +
3574 ~Literal(opener) + ~Literal(closer) +
3575 CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS,exact=1))
3575 CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS,exact=1))
3576 ).setParseAction(lambda t:t[0].strip()))
3576 ).setParseAction(lambda t:t[0].strip()))
3577 else:
3577 else:
3578 content = (Combine(OneOrMore(~Literal(opener) + ~Literal(closer) +
3578 content = (Combine(OneOrMore(~Literal(opener) + ~Literal(closer) +
3579 CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS,exact=1))
3579 CharsNotIn(ParserElement.DEFAULT_WHITE_CHARS,exact=1))
3580 ).setParseAction(lambda t:t[0].strip()))
3580 ).setParseAction(lambda t:t[0].strip()))
3581 else:
3581 else:
3582 raise ValueError("opening and closing arguments must be strings if no content expression is given")
3582 raise ValueError("opening and closing arguments must be strings if no content expression is given")
3583 ret = Forward()
3583 ret = Forward()
3584 if ignoreExpr is not None:
3584 if ignoreExpr is not None:
3585 ret << Group( Suppress(opener) + ZeroOrMore( ignoreExpr | ret | content ) + Suppress(closer) )
3585 ret << Group( Suppress(opener) + ZeroOrMore( ignoreExpr | ret | content ) + Suppress(closer) )
3586 else:
3586 else:
3587 ret << Group( Suppress(opener) + ZeroOrMore( ret | content ) + Suppress(closer) )
3587 ret << Group( Suppress(opener) + ZeroOrMore( ret | content ) + Suppress(closer) )
3588 return ret
3588 return ret
3589
3589
3590 def indentedBlock(blockStatementExpr, indentStack, indent=True):
3590 def indentedBlock(blockStatementExpr, indentStack, indent=True):
3591 """Helper method for defining space-delimited indentation blocks, such as
3591 """Helper method for defining space-delimited indentation blocks, such as
3592 those used to define block statements in Python source code.
3592 those used to define block statements in Python source code.
3593
3593
3594 Parameters:
3594 Parameters:
3595 - blockStatementExpr - expression defining syntax of statement that
3595 - blockStatementExpr - expression defining syntax of statement that
3596 is repeated within the indented block
3596 is repeated within the indented block
3597 - indentStack - list created by caller to manage indentation stack
3597 - indentStack - list created by caller to manage indentation stack
3598 (multiple statementWithIndentedBlock expressions within a single grammar
3598 (multiple statementWithIndentedBlock expressions within a single grammar
3599 should share a common indentStack)
3599 should share a common indentStack)
3600 - indent - boolean indicating whether block must be indented beyond the
3600 - indent - boolean indicating whether block must be indented beyond the
3601 the current level; set to False for block of left-most statements
3601 the current level; set to False for block of left-most statements
3602 (default=True)
3602 (default=True)
3603
3603
3604 A valid block must contain at least one blockStatement.
3604 A valid block must contain at least one blockStatement.
3605 """
3605 """
3606 def checkPeerIndent(s,l,t):
3606 def checkPeerIndent(s,l,t):
3607 if l >= len(s): return
3607 if l >= len(s): return
3608 curCol = col(l,s)
3608 curCol = col(l,s)
3609 if curCol != indentStack[-1]:
3609 if curCol != indentStack[-1]:
3610 if curCol > indentStack[-1]:
3610 if curCol > indentStack[-1]:
3611 raise ParseFatalException(s,l,"illegal nesting")
3611 raise ParseFatalException(s,l,"illegal nesting")
3612 raise ParseException(s,l,"not a peer entry")
3612 raise ParseException(s,l,"not a peer entry")
3613
3613
3614 def checkSubIndent(s,l,t):
3614 def checkSubIndent(s,l,t):
3615 curCol = col(l,s)
3615 curCol = col(l,s)
3616 if curCol > indentStack[-1]:
3616 if curCol > indentStack[-1]:
3617 indentStack.append( curCol )
3617 indentStack.append( curCol )
3618 else:
3618 else:
3619 raise ParseException(s,l,"not a subentry")
3619 raise ParseException(s,l,"not a subentry")
3620
3620
3621 def checkUnindent(s,l,t):
3621 def checkUnindent(s,l,t):
3622 if l >= len(s): return
3622 if l >= len(s): return
3623 curCol = col(l,s)
3623 curCol = col(l,s)
3624 if not(indentStack and curCol < indentStack[-1] and curCol <= indentStack[-2]):
3624 if not(indentStack and curCol < indentStack[-1] and curCol <= indentStack[-2]):
3625 raise ParseException(s,l,"not an unindent")
3625 raise ParseException(s,l,"not an unindent")
3626 indentStack.pop()
3626 indentStack.pop()
3627
3627
3628 NL = OneOrMore(LineEnd().setWhitespaceChars("\t ").suppress())
3628 NL = OneOrMore(LineEnd().setWhitespaceChars("\t ").suppress())
3629 INDENT = Empty() + Empty().setParseAction(checkSubIndent)
3629 INDENT = Empty() + Empty().setParseAction(checkSubIndent)
3630 PEER = Empty().setParseAction(checkPeerIndent)
3630 PEER = Empty().setParseAction(checkPeerIndent)
3631 UNDENT = Empty().setParseAction(checkUnindent)
3631 UNDENT = Empty().setParseAction(checkUnindent)
3632 if indent:
3632 if indent:
3633 smExpr = Group( Optional(NL) +
3633 smExpr = Group( Optional(NL) +
3634 FollowedBy(blockStatementExpr) +
3634 FollowedBy(blockStatementExpr) +
3635 INDENT + (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) + UNDENT)
3635 INDENT + (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) + UNDENT)
3636 else:
3636 else:
3637 smExpr = Group( Optional(NL) +
3637 smExpr = Group( Optional(NL) +
3638 (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) )
3638 (OneOrMore( PEER + Group(blockStatementExpr) + Optional(NL) )) )
3639 blockStatementExpr.ignore(_bslash + LineEnd())
3639 blockStatementExpr.ignore(_bslash + LineEnd())
3640 return smExpr
3640 return smExpr
3641
3641
3642 alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]")
3642 alphas8bit = srange(r"[\0xc0-\0xd6\0xd8-\0xf6\0xf8-\0xff]")
3643 punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]")
3643 punc8bit = srange(r"[\0xa1-\0xbf\0xd7\0xf7]")
3644
3644
3645 anyOpenTag,anyCloseTag = makeHTMLTags(Word(alphas,alphanums+"_:"))
3645 anyOpenTag,anyCloseTag = makeHTMLTags(Word(alphas,alphanums+"_:"))
3646 commonHTMLEntity = Combine(_L("&") + oneOf("gt lt amp nbsp quot").setResultsName("entity") +";").streamline()
3646 commonHTMLEntity = Combine(_L("&") + oneOf("gt lt amp nbsp quot").setResultsName("entity") +";").streamline()
3647 _htmlEntityMap = dict(zip("gt lt amp nbsp quot".split(),'><& "'))
3647 _htmlEntityMap = dict(zip("gt lt amp nbsp quot".split(),'><& "'))
3648 replaceHTMLEntity = lambda t : t.entity in _htmlEntityMap and _htmlEntityMap[t.entity] or None
3648 replaceHTMLEntity = lambda t : t.entity in _htmlEntityMap and _htmlEntityMap[t.entity] or None
3649
3649
3650 # it's easy to get these comment structures wrong - they're very common, so may as well make them available
3650 # it's easy to get these comment structures wrong - they're very common, so may as well make them available
3651 cStyleComment = Regex(r"/\*(?:[^*]*\*+)+?/").setName("C style comment")
3651 cStyleComment = Regex(r"/\*(?:[^*]*\*+)+?/").setName("C style comment")
3652
3652
3653 htmlComment = Regex(r"<!--[\s\S]*?-->")
3653 htmlComment = Regex(r"<!--[\s\S]*?-->")
3654 restOfLine = Regex(r".*").leaveWhitespace()
3654 restOfLine = Regex(r".*").leaveWhitespace()
3655 dblSlashComment = Regex(r"\/\/(\\\n|.)*").setName("// comment")
3655 dblSlashComment = Regex(r"\/\/(\\\n|.)*").setName("// comment")
3656 cppStyleComment = Regex(r"/(?:\*(?:[^*]*\*+)+?/|/[^\n]*(?:\n[^\n]*)*?(?:(?<!\\)|\Z))").setName("C++ style comment")
3656 cppStyleComment = Regex(r"/(?:\*(?:[^*]*\*+)+?/|/[^\n]*(?:\n[^\n]*)*?(?:(?<!\\)|\Z))").setName("C++ style comment")
3657
3657
3658 javaStyleComment = cppStyleComment
3658 javaStyleComment = cppStyleComment
3659 pythonStyleComment = Regex(r"#.*").setName("Python style comment")
3659 pythonStyleComment = Regex(r"#.*").setName("Python style comment")
3660 _noncomma = "".join( [ c for c in printables if c != "," ] )
3660 _noncomma = "".join( [ c for c in printables if c != "," ] )
3661 _commasepitem = Combine(OneOrMore(Word(_noncomma) +
3661 _commasepitem = Combine(OneOrMore(Word(_noncomma) +
3662 Optional( Word(" \t") +
3662 Optional( Word(" \t") +
3663 ~Literal(",") + ~LineEnd() ) ) ).streamline().setName("commaItem")
3663 ~Literal(",") + ~LineEnd() ) ) ).streamline().setName("commaItem")
3664 commaSeparatedList = delimitedList( Optional( quotedString | _commasepitem, default="") ).setName("commaSeparatedList")
3664 commaSeparatedList = delimitedList( Optional( quotedString | _commasepitem, default="") ).setName("commaSeparatedList")
3665
3665
3666
3666
3667 if __name__ == "__main__":
3667 if __name__ == "__main__":
3668
3668
3669 def test( teststring ):
3669 def test( teststring ):
3670 try:
3670 try:
3671 tokens = simpleSQL.parseString( teststring )
3671 tokens = simpleSQL.parseString( teststring )
3672 tokenlist = tokens.asList()
3672 tokenlist = tokens.asList()
3673 print (teststring + "->" + str(tokenlist))
3673 print (teststring + "->" + str(tokenlist))
3674 print ("tokens = " + str(tokens))
3674 print ("tokens = " + str(tokens))
3675 print ("tokens.columns = " + str(tokens.columns))
3675 print ("tokens.columns = " + str(tokens.columns))
3676 print ("tokens.tables = " + str(tokens.tables))
3676 print ("tokens.tables = " + str(tokens.tables))
3677 print (tokens.asXML("SQL",True))
3677 print (tokens.asXML("SQL",True))
3678 except ParseBaseException,err:
3678 except ParseBaseException,err:
3679 print (teststring + "->")
3679 print (teststring + "->")
3680 print (err.line)
3680 print (err.line)
3681 print (" "*(err.column-1) + "^")
3681 print (" "*(err.column-1) + "^")
3682 print (err)
3682 print (err)
3683 print()
3683 print()
3684
3684
3685 selectToken = CaselessLiteral( "select" )
3685 selectToken = CaselessLiteral( "select" )
3686 fromToken = CaselessLiteral( "from" )
3686 fromToken = CaselessLiteral( "from" )
3687
3687
3688 ident = Word( alphas, alphanums + "_$" )
3688 ident = Word( alphas, alphanums + "_$" )
3689 columnName = delimitedList( ident, ".", combine=True ).setParseAction( upcaseTokens )
3689 columnName = delimitedList( ident, ".", combine=True ).setParseAction( upcaseTokens )
3690 columnNameList = Group( delimitedList( columnName ) )#.setName("columns")
3690 columnNameList = Group( delimitedList( columnName ) )#.setName("columns")
3691 tableName = delimitedList( ident, ".", combine=True ).setParseAction( upcaseTokens )
3691 tableName = delimitedList( ident, ".", combine=True ).setParseAction( upcaseTokens )
3692 tableNameList = Group( delimitedList( tableName ) )#.setName("tables")
3692 tableNameList = Group( delimitedList( tableName ) )#.setName("tables")
3693 simpleSQL = ( selectToken + \
3693 simpleSQL = ( selectToken + \
3694 ( '*' | columnNameList ).setResultsName( "columns" ) + \
3694 ( '*' | columnNameList ).setResultsName( "columns" ) + \
3695 fromToken + \
3695 fromToken + \
3696 tableNameList.setResultsName( "tables" ) )
3696 tableNameList.setResultsName( "tables" ) )
3697
3697
3698 test( "SELECT * from XYZZY, ABC" )
3698 test( "SELECT * from XYZZY, ABC" )
3699 test( "select * from SYS.XYZZY" )
3699 test( "select * from SYS.XYZZY" )
3700 test( "Select A from Sys.dual" )
3700 test( "Select A from Sys.dual" )
3701 test( "Select AA,BB,CC from Sys.dual" )
3701 test( "Select AA,BB,CC from Sys.dual" )
3702 test( "Select A, B, C from Sys.dual" )
3702 test( "Select A, B, C from Sys.dual" )
3703 test( "Select A, B, C from Sys.dual" )
3703 test( "Select A, B, C from Sys.dual" )
3704 test( "Xelect A, B, C from Sys.dual" )
3704 test( "Xelect A, B, C from Sys.dual" )
3705 test( "Select A, B, C frox Sys.dual" )
3705 test( "Select A, B, C frox Sys.dual" )
3706 test( "Select" )
3706 test( "Select" )
3707 test( "Select ^^^ frox Sys.dual" )
3707 test( "Select ^^^ frox Sys.dual" )
3708 test( "Select A, B, C from Sys.dual, Table2 " )
3708 test( "Select A, B, C from Sys.dual, Table2 " )
@@ -1,391 +1,391 b''
1 """Generic testing tools.
1 """Generic testing tools.
2
2
3 In particular, this module exposes a set of top-level assert* functions that
3 In particular, this module exposes a set of top-level assert* functions that
4 can be used in place of nose.tools.assert* in method generators (the ones in
4 can be used in place of nose.tools.assert* in method generators (the ones in
5 nose can not, at least as of nose 0.10.4).
5 nose can not, at least as of nose 0.10.4).
6
6
7
7
8 Authors
8 Authors
9 -------
9 -------
10 - Fernando Perez <Fernando.Perez@berkeley.edu>
10 - Fernando Perez <Fernando.Perez@berkeley.edu>
11 """
11 """
12
12
13 from __future__ import absolute_import
13 from __future__ import absolute_import
14
14
15 #-----------------------------------------------------------------------------
15 #-----------------------------------------------------------------------------
16 # Copyright (C) 2009-2011 The IPython Development Team
16 # Copyright (C) 2009-2011 The IPython Development Team
17 #
17 #
18 # Distributed under the terms of the BSD License. The full license is in
18 # Distributed under the terms of the BSD License. The full license is in
19 # the file COPYING, distributed as part of this software.
19 # the file COPYING, distributed as part of this software.
20 #-----------------------------------------------------------------------------
20 #-----------------------------------------------------------------------------
21
21
22 #-----------------------------------------------------------------------------
22 #-----------------------------------------------------------------------------
23 # Imports
23 # Imports
24 #-----------------------------------------------------------------------------
24 #-----------------------------------------------------------------------------
25
25
26 import os
26 import os
27 import re
27 import re
28 import sys
28 import sys
29 import tempfile
29 import tempfile
30
30
31 from contextlib import contextmanager
31 from contextlib import contextmanager
32 from io import StringIO
32 from io import StringIO
33
33
34 try:
34 try:
35 # These tools are used by parts of the runtime, so we make the nose
35 # These tools are used by parts of the runtime, so we make the nose
36 # dependency optional at this point. Nose is a hard dependency to run the
36 # dependency optional at this point. Nose is a hard dependency to run the
37 # test suite, but NOT to use ipython itself.
37 # test suite, but NOT to use ipython itself.
38 import nose.tools as nt
38 import nose.tools as nt
39 has_nose = True
39 has_nose = True
40 except ImportError:
40 except ImportError:
41 has_nose = False
41 has_nose = False
42
42
43 from IPython.config.loader import Config
43 from IPython.config.loader import Config
44 from IPython.utils.process import find_cmd, getoutputerror
44 from IPython.utils.process import find_cmd, getoutputerror
45 from IPython.utils.text import list_strings, getdefaultencoding
45 from IPython.utils.text import list_strings
46 from IPython.utils.io import temp_pyfile, Tee
46 from IPython.utils.io import temp_pyfile, Tee
47 from IPython.utils import py3compat
47 from IPython.utils import py3compat
48
48
49 from . import decorators as dec
49 from . import decorators as dec
50 from . import skipdoctest
50 from . import skipdoctest
51
51
52 #-----------------------------------------------------------------------------
52 #-----------------------------------------------------------------------------
53 # Globals
53 # Globals
54 #-----------------------------------------------------------------------------
54 #-----------------------------------------------------------------------------
55
55
56 # Make a bunch of nose.tools assert wrappers that can be used in test
56 # Make a bunch of nose.tools assert wrappers that can be used in test
57 # generators. This will expose an assert* function for each one in nose.tools.
57 # generators. This will expose an assert* function for each one in nose.tools.
58
58
59 _tpl = """
59 _tpl = """
60 def %(name)s(*a,**kw):
60 def %(name)s(*a,**kw):
61 return nt.%(name)s(*a,**kw)
61 return nt.%(name)s(*a,**kw)
62 """
62 """
63
63
64 if has_nose:
64 if has_nose:
65 for _x in [a for a in dir(nt) if a.startswith('assert')]:
65 for _x in [a for a in dir(nt) if a.startswith('assert')]:
66 exec _tpl % dict(name=_x)
66 exec _tpl % dict(name=_x)
67
67
68 #-----------------------------------------------------------------------------
68 #-----------------------------------------------------------------------------
69 # Functions and classes
69 # Functions and classes
70 #-----------------------------------------------------------------------------
70 #-----------------------------------------------------------------------------
71
71
72 # The docstring for full_path doctests differently on win32 (different path
72 # The docstring for full_path doctests differently on win32 (different path
73 # separator) so just skip the doctest there. The example remains informative.
73 # separator) so just skip the doctest there. The example remains informative.
74 doctest_deco = skipdoctest.skip_doctest if sys.platform == 'win32' else dec.null_deco
74 doctest_deco = skipdoctest.skip_doctest if sys.platform == 'win32' else dec.null_deco
75
75
76 @doctest_deco
76 @doctest_deco
77 def full_path(startPath,files):
77 def full_path(startPath,files):
78 """Make full paths for all the listed files, based on startPath.
78 """Make full paths for all the listed files, based on startPath.
79
79
80 Only the base part of startPath is kept, since this routine is typically
80 Only the base part of startPath is kept, since this routine is typically
81 used with a script's __file__ variable as startPath. The base of startPath
81 used with a script's __file__ variable as startPath. The base of startPath
82 is then prepended to all the listed files, forming the output list.
82 is then prepended to all the listed files, forming the output list.
83
83
84 Parameters
84 Parameters
85 ----------
85 ----------
86 startPath : string
86 startPath : string
87 Initial path to use as the base for the results. This path is split
87 Initial path to use as the base for the results. This path is split
88 using os.path.split() and only its first component is kept.
88 using os.path.split() and only its first component is kept.
89
89
90 files : string or list
90 files : string or list
91 One or more files.
91 One or more files.
92
92
93 Examples
93 Examples
94 --------
94 --------
95
95
96 >>> full_path('/foo/bar.py',['a.txt','b.txt'])
96 >>> full_path('/foo/bar.py',['a.txt','b.txt'])
97 ['/foo/a.txt', '/foo/b.txt']
97 ['/foo/a.txt', '/foo/b.txt']
98
98
99 >>> full_path('/foo',['a.txt','b.txt'])
99 >>> full_path('/foo',['a.txt','b.txt'])
100 ['/a.txt', '/b.txt']
100 ['/a.txt', '/b.txt']
101
101
102 If a single file is given, the output is still a list:
102 If a single file is given, the output is still a list:
103 >>> full_path('/foo','a.txt')
103 >>> full_path('/foo','a.txt')
104 ['/a.txt']
104 ['/a.txt']
105 """
105 """
106
106
107 files = list_strings(files)
107 files = list_strings(files)
108 base = os.path.split(startPath)[0]
108 base = os.path.split(startPath)[0]
109 return [ os.path.join(base,f) for f in files ]
109 return [ os.path.join(base,f) for f in files ]
110
110
111
111
112 def parse_test_output(txt):
112 def parse_test_output(txt):
113 """Parse the output of a test run and return errors, failures.
113 """Parse the output of a test run and return errors, failures.
114
114
115 Parameters
115 Parameters
116 ----------
116 ----------
117 txt : str
117 txt : str
118 Text output of a test run, assumed to contain a line of one of the
118 Text output of a test run, assumed to contain a line of one of the
119 following forms::
119 following forms::
120 'FAILED (errors=1)'
120 'FAILED (errors=1)'
121 'FAILED (failures=1)'
121 'FAILED (failures=1)'
122 'FAILED (errors=1, failures=1)'
122 'FAILED (errors=1, failures=1)'
123
123
124 Returns
124 Returns
125 -------
125 -------
126 nerr, nfail: number of errors and failures.
126 nerr, nfail: number of errors and failures.
127 """
127 """
128
128
129 err_m = re.search(r'^FAILED \(errors=(\d+)\)', txt, re.MULTILINE)
129 err_m = re.search(r'^FAILED \(errors=(\d+)\)', txt, re.MULTILINE)
130 if err_m:
130 if err_m:
131 nerr = int(err_m.group(1))
131 nerr = int(err_m.group(1))
132 nfail = 0
132 nfail = 0
133 return nerr, nfail
133 return nerr, nfail
134
134
135 fail_m = re.search(r'^FAILED \(failures=(\d+)\)', txt, re.MULTILINE)
135 fail_m = re.search(r'^FAILED \(failures=(\d+)\)', txt, re.MULTILINE)
136 if fail_m:
136 if fail_m:
137 nerr = 0
137 nerr = 0
138 nfail = int(fail_m.group(1))
138 nfail = int(fail_m.group(1))
139 return nerr, nfail
139 return nerr, nfail
140
140
141 both_m = re.search(r'^FAILED \(errors=(\d+), failures=(\d+)\)', txt,
141 both_m = re.search(r'^FAILED \(errors=(\d+), failures=(\d+)\)', txt,
142 re.MULTILINE)
142 re.MULTILINE)
143 if both_m:
143 if both_m:
144 nerr = int(both_m.group(1))
144 nerr = int(both_m.group(1))
145 nfail = int(both_m.group(2))
145 nfail = int(both_m.group(2))
146 return nerr, nfail
146 return nerr, nfail
147
147
148 # If the input didn't match any of these forms, assume no error/failures
148 # If the input didn't match any of these forms, assume no error/failures
149 return 0, 0
149 return 0, 0
150
150
151
151
152 # So nose doesn't think this is a test
152 # So nose doesn't think this is a test
153 parse_test_output.__test__ = False
153 parse_test_output.__test__ = False
154
154
155
155
156 def default_argv():
156 def default_argv():
157 """Return a valid default argv for creating testing instances of ipython"""
157 """Return a valid default argv for creating testing instances of ipython"""
158
158
159 return ['--quick', # so no config file is loaded
159 return ['--quick', # so no config file is loaded
160 # Other defaults to minimize side effects on stdout
160 # Other defaults to minimize side effects on stdout
161 '--colors=NoColor', '--no-term-title','--no-banner',
161 '--colors=NoColor', '--no-term-title','--no-banner',
162 '--autocall=0']
162 '--autocall=0']
163
163
164
164
165 def default_config():
165 def default_config():
166 """Return a config object with good defaults for testing."""
166 """Return a config object with good defaults for testing."""
167 config = Config()
167 config = Config()
168 config.TerminalInteractiveShell.colors = 'NoColor'
168 config.TerminalInteractiveShell.colors = 'NoColor'
169 config.TerminalTerminalInteractiveShell.term_title = False,
169 config.TerminalTerminalInteractiveShell.term_title = False,
170 config.TerminalInteractiveShell.autocall = 0
170 config.TerminalInteractiveShell.autocall = 0
171 config.HistoryManager.hist_file = tempfile.mktemp(u'test_hist.sqlite')
171 config.HistoryManager.hist_file = tempfile.mktemp(u'test_hist.sqlite')
172 config.HistoryManager.db_cache_size = 10000
172 config.HistoryManager.db_cache_size = 10000
173 return config
173 return config
174
174
175
175
176 def ipexec(fname, options=None):
176 def ipexec(fname, options=None):
177 """Utility to call 'ipython filename'.
177 """Utility to call 'ipython filename'.
178
178
179 Starts IPython witha minimal and safe configuration to make startup as fast
179 Starts IPython witha minimal and safe configuration to make startup as fast
180 as possible.
180 as possible.
181
181
182 Note that this starts IPython in a subprocess!
182 Note that this starts IPython in a subprocess!
183
183
184 Parameters
184 Parameters
185 ----------
185 ----------
186 fname : str
186 fname : str
187 Name of file to be executed (should have .py or .ipy extension).
187 Name of file to be executed (should have .py or .ipy extension).
188
188
189 options : optional, list
189 options : optional, list
190 Extra command-line flags to be passed to IPython.
190 Extra command-line flags to be passed to IPython.
191
191
192 Returns
192 Returns
193 -------
193 -------
194 (stdout, stderr) of ipython subprocess.
194 (stdout, stderr) of ipython subprocess.
195 """
195 """
196 if options is None: options = []
196 if options is None: options = []
197
197
198 # For these subprocess calls, eliminate all prompt printing so we only see
198 # For these subprocess calls, eliminate all prompt printing so we only see
199 # output from script execution
199 # output from script execution
200 prompt_opts = [ '--PromptManager.in_template=""',
200 prompt_opts = [ '--PromptManager.in_template=""',
201 '--PromptManager.in2_template=""',
201 '--PromptManager.in2_template=""',
202 '--PromptManager.out_template=""'
202 '--PromptManager.out_template=""'
203 ]
203 ]
204 cmdargs = ' '.join(default_argv() + prompt_opts + options)
204 cmdargs = ' '.join(default_argv() + prompt_opts + options)
205
205
206 _ip = get_ipython()
206 _ip = get_ipython()
207 test_dir = os.path.dirname(__file__)
207 test_dir = os.path.dirname(__file__)
208
208
209 ipython_cmd = find_cmd('ipython3' if py3compat.PY3 else 'ipython')
209 ipython_cmd = find_cmd('ipython3' if py3compat.PY3 else 'ipython')
210 # Absolute path for filename
210 # Absolute path for filename
211 full_fname = os.path.join(test_dir, fname)
211 full_fname = os.path.join(test_dir, fname)
212 full_cmd = '%s %s %s' % (ipython_cmd, cmdargs, full_fname)
212 full_cmd = '%s %s %s' % (ipython_cmd, cmdargs, full_fname)
213 #print >> sys.stderr, 'FULL CMD:', full_cmd # dbg
213 #print >> sys.stderr, 'FULL CMD:', full_cmd # dbg
214 out, err = getoutputerror(full_cmd)
214 out, err = getoutputerror(full_cmd)
215 # `import readline` causes 'ESC[?1034h' to be output sometimes,
215 # `import readline` causes 'ESC[?1034h' to be output sometimes,
216 # so strip that out before doing comparisons
216 # so strip that out before doing comparisons
217 if out:
217 if out:
218 out = re.sub(r'\x1b\[[^h]+h', '', out)
218 out = re.sub(r'\x1b\[[^h]+h', '', out)
219 return out, err
219 return out, err
220
220
221
221
222 def ipexec_validate(fname, expected_out, expected_err='',
222 def ipexec_validate(fname, expected_out, expected_err='',
223 options=None):
223 options=None):
224 """Utility to call 'ipython filename' and validate output/error.
224 """Utility to call 'ipython filename' and validate output/error.
225
225
226 This function raises an AssertionError if the validation fails.
226 This function raises an AssertionError if the validation fails.
227
227
228 Note that this starts IPython in a subprocess!
228 Note that this starts IPython in a subprocess!
229
229
230 Parameters
230 Parameters
231 ----------
231 ----------
232 fname : str
232 fname : str
233 Name of the file to be executed (should have .py or .ipy extension).
233 Name of the file to be executed (should have .py or .ipy extension).
234
234
235 expected_out : str
235 expected_out : str
236 Expected stdout of the process.
236 Expected stdout of the process.
237
237
238 expected_err : optional, str
238 expected_err : optional, str
239 Expected stderr of the process.
239 Expected stderr of the process.
240
240
241 options : optional, list
241 options : optional, list
242 Extra command-line flags to be passed to IPython.
242 Extra command-line flags to be passed to IPython.
243
243
244 Returns
244 Returns
245 -------
245 -------
246 None
246 None
247 """
247 """
248
248
249 import nose.tools as nt
249 import nose.tools as nt
250
250
251 out, err = ipexec(fname, options)
251 out, err = ipexec(fname, options)
252 #print 'OUT', out # dbg
252 #print 'OUT', out # dbg
253 #print 'ERR', err # dbg
253 #print 'ERR', err # dbg
254 # If there are any errors, we must check those befor stdout, as they may be
254 # If there are any errors, we must check those befor stdout, as they may be
255 # more informative than simply having an empty stdout.
255 # more informative than simply having an empty stdout.
256 if err:
256 if err:
257 if expected_err:
257 if expected_err:
258 nt.assert_equals(err.strip(), expected_err.strip())
258 nt.assert_equals(err.strip(), expected_err.strip())
259 else:
259 else:
260 raise ValueError('Running file %r produced error: %r' %
260 raise ValueError('Running file %r produced error: %r' %
261 (fname, err))
261 (fname, err))
262 # If no errors or output on stderr was expected, match stdout
262 # If no errors or output on stderr was expected, match stdout
263 nt.assert_equals(out.strip(), expected_out.strip())
263 nt.assert_equals(out.strip(), expected_out.strip())
264
264
265
265
266 class TempFileMixin(object):
266 class TempFileMixin(object):
267 """Utility class to create temporary Python/IPython files.
267 """Utility class to create temporary Python/IPython files.
268
268
269 Meant as a mixin class for test cases."""
269 Meant as a mixin class for test cases."""
270
270
271 def mktmp(self, src, ext='.py'):
271 def mktmp(self, src, ext='.py'):
272 """Make a valid python temp file."""
272 """Make a valid python temp file."""
273 fname, f = temp_pyfile(src, ext)
273 fname, f = temp_pyfile(src, ext)
274 self.tmpfile = f
274 self.tmpfile = f
275 self.fname = fname
275 self.fname = fname
276
276
277 def tearDown(self):
277 def tearDown(self):
278 if hasattr(self, 'tmpfile'):
278 if hasattr(self, 'tmpfile'):
279 # If the tmpfile wasn't made because of skipped tests, like in
279 # If the tmpfile wasn't made because of skipped tests, like in
280 # win32, there's nothing to cleanup.
280 # win32, there's nothing to cleanup.
281 self.tmpfile.close()
281 self.tmpfile.close()
282 try:
282 try:
283 os.unlink(self.fname)
283 os.unlink(self.fname)
284 except:
284 except:
285 # On Windows, even though we close the file, we still can't
285 # On Windows, even though we close the file, we still can't
286 # delete it. I have no clue why
286 # delete it. I have no clue why
287 pass
287 pass
288
288
289 pair_fail_msg = ("Testing {0}\n\n"
289 pair_fail_msg = ("Testing {0}\n\n"
290 "In:\n"
290 "In:\n"
291 " {1!r}\n"
291 " {1!r}\n"
292 "Expected:\n"
292 "Expected:\n"
293 " {2!r}\n"
293 " {2!r}\n"
294 "Got:\n"
294 "Got:\n"
295 " {3!r}\n")
295 " {3!r}\n")
296 def check_pairs(func, pairs):
296 def check_pairs(func, pairs):
297 """Utility function for the common case of checking a function with a
297 """Utility function for the common case of checking a function with a
298 sequence of input/output pairs.
298 sequence of input/output pairs.
299
299
300 Parameters
300 Parameters
301 ----------
301 ----------
302 func : callable
302 func : callable
303 The function to be tested. Should accept a single argument.
303 The function to be tested. Should accept a single argument.
304 pairs : iterable
304 pairs : iterable
305 A list of (input, expected_output) tuples.
305 A list of (input, expected_output) tuples.
306
306
307 Returns
307 Returns
308 -------
308 -------
309 None. Raises an AssertionError if any output does not match the expected
309 None. Raises an AssertionError if any output does not match the expected
310 value.
310 value.
311 """
311 """
312 name = getattr(func, "func_name", getattr(func, "__name__", "<unknown>"))
312 name = getattr(func, "func_name", getattr(func, "__name__", "<unknown>"))
313 for inp, expected in pairs:
313 for inp, expected in pairs:
314 out = func(inp)
314 out = func(inp)
315 assert out == expected, pair_fail_msg.format(name, inp, expected, out)
315 assert out == expected, pair_fail_msg.format(name, inp, expected, out)
316
316
317
317
318 if py3compat.PY3:
318 if py3compat.PY3:
319 MyStringIO = StringIO
319 MyStringIO = StringIO
320 else:
320 else:
321 # In Python 2, stdout/stderr can have either bytes or unicode written to them,
321 # In Python 2, stdout/stderr can have either bytes or unicode written to them,
322 # so we need a class that can handle both.
322 # so we need a class that can handle both.
323 class MyStringIO(StringIO):
323 class MyStringIO(StringIO):
324 def write(self, s):
324 def write(self, s):
325 s = py3compat.cast_unicode(s, encoding=getdefaultencoding())
325 s = py3compat.cast_unicode(s, encoding=py3compat.getdefaultencoding())
326 super(MyStringIO, self).write(s)
326 super(MyStringIO, self).write(s)
327
327
328 notprinted_msg = """Did not find {0!r} in printed output (on {1}):
328 notprinted_msg = """Did not find {0!r} in printed output (on {1}):
329 {2!r}"""
329 {2!r}"""
330
330
331 class AssertPrints(object):
331 class AssertPrints(object):
332 """Context manager for testing that code prints certain text.
332 """Context manager for testing that code prints certain text.
333
333
334 Examples
334 Examples
335 --------
335 --------
336 >>> with AssertPrints("abc", suppress=False):
336 >>> with AssertPrints("abc", suppress=False):
337 ... print "abcd"
337 ... print "abcd"
338 ... print "def"
338 ... print "def"
339 ...
339 ...
340 abcd
340 abcd
341 def
341 def
342 """
342 """
343 def __init__(self, s, channel='stdout', suppress=True):
343 def __init__(self, s, channel='stdout', suppress=True):
344 self.s = s
344 self.s = s
345 self.channel = channel
345 self.channel = channel
346 self.suppress = suppress
346 self.suppress = suppress
347
347
348 def __enter__(self):
348 def __enter__(self):
349 self.orig_stream = getattr(sys, self.channel)
349 self.orig_stream = getattr(sys, self.channel)
350 self.buffer = MyStringIO()
350 self.buffer = MyStringIO()
351 self.tee = Tee(self.buffer, channel=self.channel)
351 self.tee = Tee(self.buffer, channel=self.channel)
352 setattr(sys, self.channel, self.buffer if self.suppress else self.tee)
352 setattr(sys, self.channel, self.buffer if self.suppress else self.tee)
353
353
354 def __exit__(self, etype, value, traceback):
354 def __exit__(self, etype, value, traceback):
355 self.tee.flush()
355 self.tee.flush()
356 setattr(sys, self.channel, self.orig_stream)
356 setattr(sys, self.channel, self.orig_stream)
357 printed = self.buffer.getvalue()
357 printed = self.buffer.getvalue()
358 assert self.s in printed, notprinted_msg.format(self.s, self.channel, printed)
358 assert self.s in printed, notprinted_msg.format(self.s, self.channel, printed)
359 return False
359 return False
360
360
361 class AssertNotPrints(AssertPrints):
361 class AssertNotPrints(AssertPrints):
362 """Context manager for checking that certain output *isn't* produced.
362 """Context manager for checking that certain output *isn't* produced.
363
363
364 Counterpart of AssertPrints"""
364 Counterpart of AssertPrints"""
365 def __exit__(self, etype, value, traceback):
365 def __exit__(self, etype, value, traceback):
366 self.tee.flush()
366 self.tee.flush()
367 setattr(sys, self.channel, self.orig_stream)
367 setattr(sys, self.channel, self.orig_stream)
368 printed = self.buffer.getvalue()
368 printed = self.buffer.getvalue()
369 assert self.s not in printed, notprinted_msg.format(self.s, self.channel, printed)
369 assert self.s not in printed, notprinted_msg.format(self.s, self.channel, printed)
370 return False
370 return False
371
371
372 @contextmanager
372 @contextmanager
373 def mute_warn():
373 def mute_warn():
374 from IPython.utils import warn
374 from IPython.utils import warn
375 save_warn = warn.warn
375 save_warn = warn.warn
376 warn.warn = lambda *a, **kw: None
376 warn.warn = lambda *a, **kw: None
377 try:
377 try:
378 yield
378 yield
379 finally:
379 finally:
380 warn.warn = save_warn
380 warn.warn = save_warn
381
381
382 @contextmanager
382 @contextmanager
383 def make_tempfile(name):
383 def make_tempfile(name):
384 """ Create an empty, named, temporary file for the duration of the context.
384 """ Create an empty, named, temporary file for the duration of the context.
385 """
385 """
386 f = open(name, 'w')
386 f = open(name, 'w')
387 f.close()
387 f.close()
388 try:
388 try:
389 yield
389 yield
390 finally:
390 finally:
391 os.unlink(name)
391 os.unlink(name)
@@ -1,197 +1,197 b''
1 """Posix-specific implementation of process utilities.
1 """Posix-specific implementation of process utilities.
2
2
3 This file is only meant to be imported by process.py, not by end-users.
3 This file is only meant to be imported by process.py, not by end-users.
4 """
4 """
5
5
6 #-----------------------------------------------------------------------------
6 #-----------------------------------------------------------------------------
7 # Copyright (C) 2010-2011 The IPython Development Team
7 # Copyright (C) 2010-2011 The IPython Development Team
8 #
8 #
9 # Distributed under the terms of the BSD License. The full license is in
9 # Distributed under the terms of the BSD License. The full license is in
10 # the file COPYING, distributed as part of this software.
10 # the file COPYING, distributed as part of this software.
11 #-----------------------------------------------------------------------------
11 #-----------------------------------------------------------------------------
12
12
13 #-----------------------------------------------------------------------------
13 #-----------------------------------------------------------------------------
14 # Imports
14 # Imports
15 #-----------------------------------------------------------------------------
15 #-----------------------------------------------------------------------------
16 from __future__ import print_function
16 from __future__ import print_function
17
17
18 # Stdlib
18 # Stdlib
19 import subprocess as sp
19 import subprocess as sp
20 import sys
20 import sys
21
21
22 from IPython.external import pexpect
22 from IPython.external import pexpect
23
23
24 # Our own
24 # Our own
25 from .autoattr import auto_attr
25 from .autoattr import auto_attr
26 from ._process_common import getoutput, arg_split
26 from ._process_common import getoutput, arg_split
27 from IPython.utils import text
27 from IPython.utils import text
28 from IPython.utils import py3compat
28 from IPython.utils import py3compat
29
29
30 #-----------------------------------------------------------------------------
30 #-----------------------------------------------------------------------------
31 # Function definitions
31 # Function definitions
32 #-----------------------------------------------------------------------------
32 #-----------------------------------------------------------------------------
33
33
34 def _find_cmd(cmd):
34 def _find_cmd(cmd):
35 """Find the full path to a command using which."""
35 """Find the full path to a command using which."""
36
36
37 path = sp.Popen(['/usr/bin/env', 'which', cmd],
37 path = sp.Popen(['/usr/bin/env', 'which', cmd],
38 stdout=sp.PIPE).communicate()[0]
38 stdout=sp.PIPE).communicate()[0]
39 return py3compat.bytes_to_str(path)
39 return py3compat.bytes_to_str(path)
40
40
41
41
42 class ProcessHandler(object):
42 class ProcessHandler(object):
43 """Execute subprocesses under the control of pexpect.
43 """Execute subprocesses under the control of pexpect.
44 """
44 """
45 # Timeout in seconds to wait on each reading of the subprocess' output.
45 # Timeout in seconds to wait on each reading of the subprocess' output.
46 # This should not be set too low to avoid cpu overusage from our side,
46 # This should not be set too low to avoid cpu overusage from our side,
47 # since we read in a loop whose period is controlled by this timeout.
47 # since we read in a loop whose period is controlled by this timeout.
48 read_timeout = 0.05
48 read_timeout = 0.05
49
49
50 # Timeout to give a process if we receive SIGINT, between sending the
50 # Timeout to give a process if we receive SIGINT, between sending the
51 # SIGINT to the process and forcefully terminating it.
51 # SIGINT to the process and forcefully terminating it.
52 terminate_timeout = 0.2
52 terminate_timeout = 0.2
53
53
54 # File object where stdout and stderr of the subprocess will be written
54 # File object where stdout and stderr of the subprocess will be written
55 logfile = None
55 logfile = None
56
56
57 # Shell to call for subprocesses to execute
57 # Shell to call for subprocesses to execute
58 sh = None
58 sh = None
59
59
60 @auto_attr
60 @auto_attr
61 def sh(self):
61 def sh(self):
62 sh = pexpect.which('sh')
62 sh = pexpect.which('sh')
63 if sh is None:
63 if sh is None:
64 raise OSError('"sh" shell not found')
64 raise OSError('"sh" shell not found')
65 return sh
65 return sh
66
66
67 def __init__(self, logfile=None, read_timeout=None, terminate_timeout=None):
67 def __init__(self, logfile=None, read_timeout=None, terminate_timeout=None):
68 """Arguments are used for pexpect calls."""
68 """Arguments are used for pexpect calls."""
69 self.read_timeout = (ProcessHandler.read_timeout if read_timeout is
69 self.read_timeout = (ProcessHandler.read_timeout if read_timeout is
70 None else read_timeout)
70 None else read_timeout)
71 self.terminate_timeout = (ProcessHandler.terminate_timeout if
71 self.terminate_timeout = (ProcessHandler.terminate_timeout if
72 terminate_timeout is None else
72 terminate_timeout is None else
73 terminate_timeout)
73 terminate_timeout)
74 self.logfile = sys.stdout if logfile is None else logfile
74 self.logfile = sys.stdout if logfile is None else logfile
75
75
76 def getoutput(self, cmd):
76 def getoutput(self, cmd):
77 """Run a command and return its stdout/stderr as a string.
77 """Run a command and return its stdout/stderr as a string.
78
78
79 Parameters
79 Parameters
80 ----------
80 ----------
81 cmd : str
81 cmd : str
82 A command to be executed in the system shell.
82 A command to be executed in the system shell.
83
83
84 Returns
84 Returns
85 -------
85 -------
86 output : str
86 output : str
87 A string containing the combination of stdout and stderr from the
87 A string containing the combination of stdout and stderr from the
88 subprocess, in whatever order the subprocess originally wrote to its
88 subprocess, in whatever order the subprocess originally wrote to its
89 file descriptors (so the order of the information in this string is the
89 file descriptors (so the order of the information in this string is the
90 correct order as would be seen if running the command in a terminal).
90 correct order as would be seen if running the command in a terminal).
91 """
91 """
92 try:
92 try:
93 return pexpect.run(self.sh, args=['-c', cmd]).replace('\r\n', '\n')
93 return pexpect.run(self.sh, args=['-c', cmd]).replace('\r\n', '\n')
94 except KeyboardInterrupt:
94 except KeyboardInterrupt:
95 print('^C', file=sys.stderr, end='')
95 print('^C', file=sys.stderr, end='')
96
96
97 def getoutput_pexpect(self, cmd):
97 def getoutput_pexpect(self, cmd):
98 """Run a command and return its stdout/stderr as a string.
98 """Run a command and return its stdout/stderr as a string.
99
99
100 Parameters
100 Parameters
101 ----------
101 ----------
102 cmd : str
102 cmd : str
103 A command to be executed in the system shell.
103 A command to be executed in the system shell.
104
104
105 Returns
105 Returns
106 -------
106 -------
107 output : str
107 output : str
108 A string containing the combination of stdout and stderr from the
108 A string containing the combination of stdout and stderr from the
109 subprocess, in whatever order the subprocess originally wrote to its
109 subprocess, in whatever order the subprocess originally wrote to its
110 file descriptors (so the order of the information in this string is the
110 file descriptors (so the order of the information in this string is the
111 correct order as would be seen if running the command in a terminal).
111 correct order as would be seen if running the command in a terminal).
112 """
112 """
113 try:
113 try:
114 return pexpect.run(self.sh, args=['-c', cmd]).replace('\r\n', '\n')
114 return pexpect.run(self.sh, args=['-c', cmd]).replace('\r\n', '\n')
115 except KeyboardInterrupt:
115 except KeyboardInterrupt:
116 print('^C', file=sys.stderr, end='')
116 print('^C', file=sys.stderr, end='')
117
117
118 def system(self, cmd):
118 def system(self, cmd):
119 """Execute a command in a subshell.
119 """Execute a command in a subshell.
120
120
121 Parameters
121 Parameters
122 ----------
122 ----------
123 cmd : str
123 cmd : str
124 A command to be executed in the system shell.
124 A command to be executed in the system shell.
125
125
126 Returns
126 Returns
127 -------
127 -------
128 int : child's exitstatus
128 int : child's exitstatus
129 """
129 """
130 # Get likely encoding for the output.
130 # Get likely encoding for the output.
131 enc = text.getdefaultencoding()
131 enc = py3compat.getdefaultencoding()
132
132
133 # Patterns to match on the output, for pexpect. We read input and
133 # Patterns to match on the output, for pexpect. We read input and
134 # allow either a short timeout or EOF
134 # allow either a short timeout or EOF
135 patterns = [pexpect.TIMEOUT, pexpect.EOF]
135 patterns = [pexpect.TIMEOUT, pexpect.EOF]
136 # the index of the EOF pattern in the list.
136 # the index of the EOF pattern in the list.
137 # even though we know it's 1, this call means we don't have to worry if
137 # even though we know it's 1, this call means we don't have to worry if
138 # we change the above list, and forget to change this value:
138 # we change the above list, and forget to change this value:
139 EOF_index = patterns.index(pexpect.EOF)
139 EOF_index = patterns.index(pexpect.EOF)
140 # The size of the output stored so far in the process output buffer.
140 # The size of the output stored so far in the process output buffer.
141 # Since pexpect only appends to this buffer, each time we print we
141 # Since pexpect only appends to this buffer, each time we print we
142 # record how far we've printed, so that next time we only print *new*
142 # record how far we've printed, so that next time we only print *new*
143 # content from the buffer.
143 # content from the buffer.
144 out_size = 0
144 out_size = 0
145 try:
145 try:
146 # Since we're not really searching the buffer for text patterns, we
146 # Since we're not really searching the buffer for text patterns, we
147 # can set pexpect's search window to be tiny and it won't matter.
147 # can set pexpect's search window to be tiny and it won't matter.
148 # We only search for the 'patterns' timeout or EOF, which aren't in
148 # We only search for the 'patterns' timeout or EOF, which aren't in
149 # the text itself.
149 # the text itself.
150 #child = pexpect.spawn(pcmd, searchwindowsize=1)
150 #child = pexpect.spawn(pcmd, searchwindowsize=1)
151 if hasattr(pexpect, 'spawnb'):
151 if hasattr(pexpect, 'spawnb'):
152 child = pexpect.spawnb(self.sh, args=['-c', cmd]) # Pexpect-U
152 child = pexpect.spawnb(self.sh, args=['-c', cmd]) # Pexpect-U
153 else:
153 else:
154 child = pexpect.spawn(self.sh, args=['-c', cmd]) # Vanilla Pexpect
154 child = pexpect.spawn(self.sh, args=['-c', cmd]) # Vanilla Pexpect
155 flush = sys.stdout.flush
155 flush = sys.stdout.flush
156 while True:
156 while True:
157 # res is the index of the pattern that caused the match, so we
157 # res is the index of the pattern that caused the match, so we
158 # know whether we've finished (if we matched EOF) or not
158 # know whether we've finished (if we matched EOF) or not
159 res_idx = child.expect_list(patterns, self.read_timeout)
159 res_idx = child.expect_list(patterns, self.read_timeout)
160 print(child.before[out_size:].decode(enc, 'replace'), end='')
160 print(child.before[out_size:].decode(enc, 'replace'), end='')
161 flush()
161 flush()
162 if res_idx==EOF_index:
162 if res_idx==EOF_index:
163 break
163 break
164 # Update the pointer to what we've already printed
164 # Update the pointer to what we've already printed
165 out_size = len(child.before)
165 out_size = len(child.before)
166 except KeyboardInterrupt:
166 except KeyboardInterrupt:
167 # We need to send ^C to the process. The ascii code for '^C' is 3
167 # We need to send ^C to the process. The ascii code for '^C' is 3
168 # (the character is known as ETX for 'End of Text', see
168 # (the character is known as ETX for 'End of Text', see
169 # curses.ascii.ETX).
169 # curses.ascii.ETX).
170 child.sendline(chr(3))
170 child.sendline(chr(3))
171 # Read and print any more output the program might produce on its
171 # Read and print any more output the program might produce on its
172 # way out.
172 # way out.
173 try:
173 try:
174 out_size = len(child.before)
174 out_size = len(child.before)
175 child.expect_list(patterns, self.terminate_timeout)
175 child.expect_list(patterns, self.terminate_timeout)
176 print(child.before[out_size:].decode(enc, 'replace'), end='')
176 print(child.before[out_size:].decode(enc, 'replace'), end='')
177 sys.stdout.flush()
177 sys.stdout.flush()
178 except KeyboardInterrupt:
178 except KeyboardInterrupt:
179 # Impatient users tend to type it multiple times
179 # Impatient users tend to type it multiple times
180 pass
180 pass
181 finally:
181 finally:
182 # Ensure the subprocess really is terminated
182 # Ensure the subprocess really is terminated
183 child.terminate(force=True)
183 child.terminate(force=True)
184 # add isalive check, to ensure exitstatus is set:
184 # add isalive check, to ensure exitstatus is set:
185 child.isalive()
185 child.isalive()
186 return child.exitstatus
186 return child.exitstatus
187
187
188
188
189 # Make system() with a functional interface for outside use. Note that we use
189 # Make system() with a functional interface for outside use. Note that we use
190 # getoutput() from the _common utils, which is built on top of popen(). Using
190 # getoutput() from the _common utils, which is built on top of popen(). Using
191 # pexpect to get subprocess output produces difficult to parse output, since
191 # pexpect to get subprocess output produces difficult to parse output, since
192 # programs think they are talking to a tty and produce highly formatted output
192 # programs think they are talking to a tty and produce highly formatted output
193 # (ls is a good example) that makes them hard.
193 # (ls is a good example) that makes them hard.
194 system = ProcessHandler().system
194 system = ProcessHandler().system
195
195
196
196
197
197
@@ -1,184 +1,184 b''
1 """Windows-specific implementation of process utilities.
1 """Windows-specific implementation of process utilities.
2
2
3 This file is only meant to be imported by process.py, not by end-users.
3 This file is only meant to be imported by process.py, not by end-users.
4 """
4 """
5
5
6 #-----------------------------------------------------------------------------
6 #-----------------------------------------------------------------------------
7 # Copyright (C) 2010-2011 The IPython Development Team
7 # Copyright (C) 2010-2011 The IPython Development Team
8 #
8 #
9 # Distributed under the terms of the BSD License. The full license is in
9 # Distributed under the terms of the BSD License. The full license is in
10 # the file COPYING, distributed as part of this software.
10 # the file COPYING, distributed as part of this software.
11 #-----------------------------------------------------------------------------
11 #-----------------------------------------------------------------------------
12
12
13 #-----------------------------------------------------------------------------
13 #-----------------------------------------------------------------------------
14 # Imports
14 # Imports
15 #-----------------------------------------------------------------------------
15 #-----------------------------------------------------------------------------
16 from __future__ import print_function
16 from __future__ import print_function
17
17
18 # stdlib
18 # stdlib
19 import os
19 import os
20 import sys
20 import sys
21 import ctypes
21 import ctypes
22 import msvcrt
22 import msvcrt
23
23
24 from ctypes import c_int, POINTER
24 from ctypes import c_int, POINTER
25 from ctypes.wintypes import LPCWSTR, HLOCAL
25 from ctypes.wintypes import LPCWSTR, HLOCAL
26 from subprocess import STDOUT
26 from subprocess import STDOUT
27
27
28 # our own imports
28 # our own imports
29 from ._process_common import read_no_interrupt, process_handler, arg_split as py_arg_split
29 from ._process_common import read_no_interrupt, process_handler, arg_split as py_arg_split
30 from . import py3compat
30 from . import py3compat
31 from . import text
31 from . import text
32
32
33 #-----------------------------------------------------------------------------
33 #-----------------------------------------------------------------------------
34 # Function definitions
34 # Function definitions
35 #-----------------------------------------------------------------------------
35 #-----------------------------------------------------------------------------
36
36
37 class AvoidUNCPath(object):
37 class AvoidUNCPath(object):
38 """A context manager to protect command execution from UNC paths.
38 """A context manager to protect command execution from UNC paths.
39
39
40 In the Win32 API, commands can't be invoked with the cwd being a UNC path.
40 In the Win32 API, commands can't be invoked with the cwd being a UNC path.
41 This context manager temporarily changes directory to the 'C:' drive on
41 This context manager temporarily changes directory to the 'C:' drive on
42 entering, and restores the original working directory on exit.
42 entering, and restores the original working directory on exit.
43
43
44 The context manager returns the starting working directory *if* it made a
44 The context manager returns the starting working directory *if* it made a
45 change and None otherwise, so that users can apply the necessary adjustment
45 change and None otherwise, so that users can apply the necessary adjustment
46 to their system calls in the event of a change.
46 to their system calls in the event of a change.
47
47
48 Example
48 Example
49 -------
49 -------
50 ::
50 ::
51 cmd = 'dir'
51 cmd = 'dir'
52 with AvoidUNCPath() as path:
52 with AvoidUNCPath() as path:
53 if path is not None:
53 if path is not None:
54 cmd = '"pushd %s &&"%s' % (path, cmd)
54 cmd = '"pushd %s &&"%s' % (path, cmd)
55 os.system(cmd)
55 os.system(cmd)
56 """
56 """
57 def __enter__(self):
57 def __enter__(self):
58 self.path = os.getcwdu()
58 self.path = os.getcwdu()
59 self.is_unc_path = self.path.startswith(r"\\")
59 self.is_unc_path = self.path.startswith(r"\\")
60 if self.is_unc_path:
60 if self.is_unc_path:
61 # change to c drive (as cmd.exe cannot handle UNC addresses)
61 # change to c drive (as cmd.exe cannot handle UNC addresses)
62 os.chdir("C:")
62 os.chdir("C:")
63 return self.path
63 return self.path
64 else:
64 else:
65 # We return None to signal that there was no change in the working
65 # We return None to signal that there was no change in the working
66 # directory
66 # directory
67 return None
67 return None
68
68
69 def __exit__(self, exc_type, exc_value, traceback):
69 def __exit__(self, exc_type, exc_value, traceback):
70 if self.is_unc_path:
70 if self.is_unc_path:
71 os.chdir(self.path)
71 os.chdir(self.path)
72
72
73
73
74 def _find_cmd(cmd):
74 def _find_cmd(cmd):
75 """Find the full path to a .bat or .exe using the win32api module."""
75 """Find the full path to a .bat or .exe using the win32api module."""
76 try:
76 try:
77 from win32api import SearchPath
77 from win32api import SearchPath
78 except ImportError:
78 except ImportError:
79 raise ImportError('you need to have pywin32 installed for this to work')
79 raise ImportError('you need to have pywin32 installed for this to work')
80 else:
80 else:
81 PATH = os.environ['PATH']
81 PATH = os.environ['PATH']
82 extensions = ['.exe', '.com', '.bat', '.py']
82 extensions = ['.exe', '.com', '.bat', '.py']
83 path = None
83 path = None
84 for ext in extensions:
84 for ext in extensions:
85 try:
85 try:
86 path = SearchPath(PATH, cmd + ext)[0]
86 path = SearchPath(PATH, cmd + ext)[0]
87 except:
87 except:
88 pass
88 pass
89 if path is None:
89 if path is None:
90 raise OSError("command %r not found" % cmd)
90 raise OSError("command %r not found" % cmd)
91 else:
91 else:
92 return path
92 return path
93
93
94
94
95 def _system_body(p):
95 def _system_body(p):
96 """Callback for _system."""
96 """Callback for _system."""
97 enc = text.getdefaultencoding()
97 enc = py3compat.getdefaultencoding()
98 for line in read_no_interrupt(p.stdout).splitlines():
98 for line in read_no_interrupt(p.stdout).splitlines():
99 line = line.decode(enc, 'replace')
99 line = line.decode(enc, 'replace')
100 print(line, file=sys.stdout)
100 print(line, file=sys.stdout)
101 for line in read_no_interrupt(p.stderr).splitlines():
101 for line in read_no_interrupt(p.stderr).splitlines():
102 line = line.decode(enc, 'replace')
102 line = line.decode(enc, 'replace')
103 print(line, file=sys.stderr)
103 print(line, file=sys.stderr)
104
104
105 # Wait to finish for returncode
105 # Wait to finish for returncode
106 return p.wait()
106 return p.wait()
107
107
108
108
109 def system(cmd):
109 def system(cmd):
110 """Win32 version of os.system() that works with network shares.
110 """Win32 version of os.system() that works with network shares.
111
111
112 Note that this implementation returns None, as meant for use in IPython.
112 Note that this implementation returns None, as meant for use in IPython.
113
113
114 Parameters
114 Parameters
115 ----------
115 ----------
116 cmd : str
116 cmd : str
117 A command to be executed in the system shell.
117 A command to be executed in the system shell.
118
118
119 Returns
119 Returns
120 -------
120 -------
121 None : we explicitly do NOT return the subprocess status code, as this
121 None : we explicitly do NOT return the subprocess status code, as this
122 utility is meant to be used extensively in IPython, where any return value
122 utility is meant to be used extensively in IPython, where any return value
123 would trigger :func:`sys.displayhook` calls.
123 would trigger :func:`sys.displayhook` calls.
124 """
124 """
125 # The controller provides interactivity with both
125 # The controller provides interactivity with both
126 # stdin and stdout
126 # stdin and stdout
127 import _process_win32_controller
127 import _process_win32_controller
128 _process_win32_controller.system(cmd)
128 _process_win32_controller.system(cmd)
129
129
130
130
131 def getoutput(cmd):
131 def getoutput(cmd):
132 """Return standard output of executing cmd in a shell.
132 """Return standard output of executing cmd in a shell.
133
133
134 Accepts the same arguments as os.system().
134 Accepts the same arguments as os.system().
135
135
136 Parameters
136 Parameters
137 ----------
137 ----------
138 cmd : str
138 cmd : str
139 A command to be executed in the system shell.
139 A command to be executed in the system shell.
140
140
141 Returns
141 Returns
142 -------
142 -------
143 stdout : str
143 stdout : str
144 """
144 """
145
145
146 with AvoidUNCPath() as path:
146 with AvoidUNCPath() as path:
147 if path is not None:
147 if path is not None:
148 cmd = '"pushd %s &&"%s' % (path, cmd)
148 cmd = '"pushd %s &&"%s' % (path, cmd)
149 out = process_handler(cmd, lambda p: p.communicate()[0], STDOUT)
149 out = process_handler(cmd, lambda p: p.communicate()[0], STDOUT)
150
150
151 if out is None:
151 if out is None:
152 out = ''
152 out = ''
153 return out
153 return out
154
154
155 try:
155 try:
156 CommandLineToArgvW = ctypes.windll.shell32.CommandLineToArgvW
156 CommandLineToArgvW = ctypes.windll.shell32.CommandLineToArgvW
157 CommandLineToArgvW.arg_types = [LPCWSTR, POINTER(c_int)]
157 CommandLineToArgvW.arg_types = [LPCWSTR, POINTER(c_int)]
158 CommandLineToArgvW.res_types = [POINTER(LPCWSTR)]
158 CommandLineToArgvW.res_types = [POINTER(LPCWSTR)]
159 LocalFree = ctypes.windll.kernel32.LocalFree
159 LocalFree = ctypes.windll.kernel32.LocalFree
160 LocalFree.res_type = HLOCAL
160 LocalFree.res_type = HLOCAL
161 LocalFree.arg_types = [HLOCAL]
161 LocalFree.arg_types = [HLOCAL]
162
162
163 def arg_split(commandline, posix=False, strict=True):
163 def arg_split(commandline, posix=False, strict=True):
164 """Split a command line's arguments in a shell-like manner.
164 """Split a command line's arguments in a shell-like manner.
165
165
166 This is a special version for windows that use a ctypes call to CommandLineToArgvW
166 This is a special version for windows that use a ctypes call to CommandLineToArgvW
167 to do the argv splitting. The posix paramter is ignored.
167 to do the argv splitting. The posix paramter is ignored.
168
168
169 If strict=False, process_common.arg_split(...strict=False) is used instead.
169 If strict=False, process_common.arg_split(...strict=False) is used instead.
170 """
170 """
171 #CommandLineToArgvW returns path to executable if called with empty string.
171 #CommandLineToArgvW returns path to executable if called with empty string.
172 if commandline.strip() == "":
172 if commandline.strip() == "":
173 return []
173 return []
174 if not strict:
174 if not strict:
175 # not really a cl-arg, fallback on _process_common
175 # not really a cl-arg, fallback on _process_common
176 return py_arg_split(commandline, posix=posix, strict=strict)
176 return py_arg_split(commandline, posix=posix, strict=strict)
177 argvn = c_int()
177 argvn = c_int()
178 result_pointer = CommandLineToArgvW(py3compat.cast_unicode(commandline.lstrip()), ctypes.byref(argvn))
178 result_pointer = CommandLineToArgvW(py3compat.cast_unicode(commandline.lstrip()), ctypes.byref(argvn))
179 result_array_type = LPCWSTR * argvn.value
179 result_array_type = LPCWSTR * argvn.value
180 result = [arg for arg in result_array_type.from_address(result_pointer)]
180 result = [arg for arg in result_array_type.from_address(result_pointer)]
181 retval = LocalFree(result_pointer)
181 retval = LocalFree(result_pointer)
182 return result
182 return result
183 except AttributeError:
183 except AttributeError:
184 arg_split = py_arg_split
184 arg_split = py_arg_split
@@ -1,322 +1,323 b''
1 # encoding: utf-8
1 # encoding: utf-8
2 """
2 """
3 IO related utilities.
3 IO related utilities.
4 """
4 """
5
5
6 #-----------------------------------------------------------------------------
6 #-----------------------------------------------------------------------------
7 # Copyright (C) 2008-2011 The IPython Development Team
7 # Copyright (C) 2008-2011 The IPython Development Team
8 #
8 #
9 # Distributed under the terms of the BSD License. The full license is in
9 # Distributed under the terms of the BSD License. The full license is in
10 # the file COPYING, distributed as part of this software.
10 # the file COPYING, distributed as part of this software.
11 #-----------------------------------------------------------------------------
11 #-----------------------------------------------------------------------------
12 from __future__ import print_function
12 from __future__ import print_function
13
13
14 #-----------------------------------------------------------------------------
14 #-----------------------------------------------------------------------------
15 # Imports
15 # Imports
16 #-----------------------------------------------------------------------------
16 #-----------------------------------------------------------------------------
17 import os
17 import sys
18 import sys
18 import tempfile
19 import tempfile
19
20
20 #-----------------------------------------------------------------------------
21 #-----------------------------------------------------------------------------
21 # Code
22 # Code
22 #-----------------------------------------------------------------------------
23 #-----------------------------------------------------------------------------
23
24
24
25
25 class IOStream:
26 class IOStream:
26
27
27 def __init__(self,stream, fallback=None):
28 def __init__(self,stream, fallback=None):
28 if not hasattr(stream,'write') or not hasattr(stream,'flush'):
29 if not hasattr(stream,'write') or not hasattr(stream,'flush'):
29 if fallback is not None:
30 if fallback is not None:
30 stream = fallback
31 stream = fallback
31 else:
32 else:
32 raise ValueError("fallback required, but not specified")
33 raise ValueError("fallback required, but not specified")
33 self.stream = stream
34 self.stream = stream
34 self._swrite = stream.write
35 self._swrite = stream.write
35
36
36 # clone all methods not overridden:
37 # clone all methods not overridden:
37 def clone(meth):
38 def clone(meth):
38 return not hasattr(self, meth) and not meth.startswith('_')
39 return not hasattr(self, meth) and not meth.startswith('_')
39 for meth in filter(clone, dir(stream)):
40 for meth in filter(clone, dir(stream)):
40 setattr(self, meth, getattr(stream, meth))
41 setattr(self, meth, getattr(stream, meth))
41
42
42 def write(self,data):
43 def write(self,data):
43 try:
44 try:
44 self._swrite(data)
45 self._swrite(data)
45 except:
46 except:
46 try:
47 try:
47 # print handles some unicode issues which may trip a plain
48 # print handles some unicode issues which may trip a plain
48 # write() call. Emulate write() by using an empty end
49 # write() call. Emulate write() by using an empty end
49 # argument.
50 # argument.
50 print(data, end='', file=self.stream)
51 print(data, end='', file=self.stream)
51 except:
52 except:
52 # if we get here, something is seriously broken.
53 # if we get here, something is seriously broken.
53 print('ERROR - failed to write data to stream:', self.stream,
54 print('ERROR - failed to write data to stream:', self.stream,
54 file=sys.stderr)
55 file=sys.stderr)
55
56
56 def writelines(self, lines):
57 def writelines(self, lines):
57 if isinstance(lines, basestring):
58 if isinstance(lines, basestring):
58 lines = [lines]
59 lines = [lines]
59 for line in lines:
60 for line in lines:
60 self.write(line)
61 self.write(line)
61
62
62 # This class used to have a writeln method, but regular files and streams
63 # This class used to have a writeln method, but regular files and streams
63 # in Python don't have this method. We need to keep this completely
64 # in Python don't have this method. We need to keep this completely
64 # compatible so we removed it.
65 # compatible so we removed it.
65
66
66 @property
67 @property
67 def closed(self):
68 def closed(self):
68 return self.stream.closed
69 return self.stream.closed
69
70
70 def close(self):
71 def close(self):
71 pass
72 pass
72
73
73 # setup stdin/stdout/stderr to sys.stdin/sys.stdout/sys.stderr
74 # setup stdin/stdout/stderr to sys.stdin/sys.stdout/sys.stderr
74 devnull = open(os.devnull, 'a')
75 devnull = open(os.devnull, 'a')
75 stdin = IOStream(sys.stdin, fallback=devnull)
76 stdin = IOStream(sys.stdin, fallback=devnull)
76 stdout = IOStream(sys.stdout, fallback=devnull)
77 stdout = IOStream(sys.stdout, fallback=devnull)
77 stderr = IOStream(sys.stderr, fallback=devnull)
78 stderr = IOStream(sys.stderr, fallback=devnull)
78
79
79 class IOTerm:
80 class IOTerm:
80 """ Term holds the file or file-like objects for handling I/O operations.
81 """ Term holds the file or file-like objects for handling I/O operations.
81
82
82 These are normally just sys.stdin, sys.stdout and sys.stderr but for
83 These are normally just sys.stdin, sys.stdout and sys.stderr but for
83 Windows they can can replaced to allow editing the strings before they are
84 Windows they can can replaced to allow editing the strings before they are
84 displayed."""
85 displayed."""
85
86
86 # In the future, having IPython channel all its I/O operations through
87 # In the future, having IPython channel all its I/O operations through
87 # this class will make it easier to embed it into other environments which
88 # this class will make it easier to embed it into other environments which
88 # are not a normal terminal (such as a GUI-based shell)
89 # are not a normal terminal (such as a GUI-based shell)
89 def __init__(self, stdin=None, stdout=None, stderr=None):
90 def __init__(self, stdin=None, stdout=None, stderr=None):
90 mymodule = sys.modules[__name__]
91 mymodule = sys.modules[__name__]
91 self.stdin = IOStream(stdin, mymodule.stdin)
92 self.stdin = IOStream(stdin, mymodule.stdin)
92 self.stdout = IOStream(stdout, mymodule.stdout)
93 self.stdout = IOStream(stdout, mymodule.stdout)
93 self.stderr = IOStream(stderr, mymodule.stderr)
94 self.stderr = IOStream(stderr, mymodule.stderr)
94
95
95
96
96 class Tee(object):
97 class Tee(object):
97 """A class to duplicate an output stream to stdout/err.
98 """A class to duplicate an output stream to stdout/err.
98
99
99 This works in a manner very similar to the Unix 'tee' command.
100 This works in a manner very similar to the Unix 'tee' command.
100
101
101 When the object is closed or deleted, it closes the original file given to
102 When the object is closed or deleted, it closes the original file given to
102 it for duplication.
103 it for duplication.
103 """
104 """
104 # Inspired by:
105 # Inspired by:
105 # http://mail.python.org/pipermail/python-list/2007-May/442737.html
106 # http://mail.python.org/pipermail/python-list/2007-May/442737.html
106
107
107 def __init__(self, file_or_name, mode="w", channel='stdout'):
108 def __init__(self, file_or_name, mode="w", channel='stdout'):
108 """Construct a new Tee object.
109 """Construct a new Tee object.
109
110
110 Parameters
111 Parameters
111 ----------
112 ----------
112 file_or_name : filename or open filehandle (writable)
113 file_or_name : filename or open filehandle (writable)
113 File that will be duplicated
114 File that will be duplicated
114
115
115 mode : optional, valid mode for open().
116 mode : optional, valid mode for open().
116 If a filename was give, open with this mode.
117 If a filename was give, open with this mode.
117
118
118 channel : str, one of ['stdout', 'stderr']
119 channel : str, one of ['stdout', 'stderr']
119 """
120 """
120 if channel not in ['stdout', 'stderr']:
121 if channel not in ['stdout', 'stderr']:
121 raise ValueError('Invalid channel spec %s' % channel)
122 raise ValueError('Invalid channel spec %s' % channel)
122
123
123 if hasattr(file_or_name, 'write') and hasattr(file_or_name, 'seek'):
124 if hasattr(file_or_name, 'write') and hasattr(file_or_name, 'seek'):
124 self.file = file_or_name
125 self.file = file_or_name
125 else:
126 else:
126 self.file = open(file_or_name, mode)
127 self.file = open(file_or_name, mode)
127 self.channel = channel
128 self.channel = channel
128 self.ostream = getattr(sys, channel)
129 self.ostream = getattr(sys, channel)
129 setattr(sys, channel, self)
130 setattr(sys, channel, self)
130 self._closed = False
131 self._closed = False
131
132
132 def close(self):
133 def close(self):
133 """Close the file and restore the channel."""
134 """Close the file and restore the channel."""
134 self.flush()
135 self.flush()
135 setattr(sys, self.channel, self.ostream)
136 setattr(sys, self.channel, self.ostream)
136 self.file.close()
137 self.file.close()
137 self._closed = True
138 self._closed = True
138
139
139 def write(self, data):
140 def write(self, data):
140 """Write data to both channels."""
141 """Write data to both channels."""
141 self.file.write(data)
142 self.file.write(data)
142 self.ostream.write(data)
143 self.ostream.write(data)
143 self.ostream.flush()
144 self.ostream.flush()
144
145
145 def flush(self):
146 def flush(self):
146 """Flush both channels."""
147 """Flush both channels."""
147 self.file.flush()
148 self.file.flush()
148 self.ostream.flush()
149 self.ostream.flush()
149
150
150 def __del__(self):
151 def __del__(self):
151 if not self._closed:
152 if not self._closed:
152 self.close()
153 self.close()
153
154
154
155
155 def file_read(filename):
156 def file_read(filename):
156 """Read a file and close it. Returns the file source."""
157 """Read a file and close it. Returns the file source."""
157 fobj = open(filename,'r');
158 fobj = open(filename,'r');
158 source = fobj.read();
159 source = fobj.read();
159 fobj.close()
160 fobj.close()
160 return source
161 return source
161
162
162
163
163 def file_readlines(filename):
164 def file_readlines(filename):
164 """Read a file and close it. Returns the file source using readlines()."""
165 """Read a file and close it. Returns the file source using readlines()."""
165 fobj = open(filename,'r');
166 fobj = open(filename,'r');
166 lines = fobj.readlines();
167 lines = fobj.readlines();
167 fobj.close()
168 fobj.close()
168 return lines
169 return lines
169
170
170
171
171 def raw_input_multi(header='', ps1='==> ', ps2='..> ',terminate_str = '.'):
172 def raw_input_multi(header='', ps1='==> ', ps2='..> ',terminate_str = '.'):
172 """Take multiple lines of input.
173 """Take multiple lines of input.
173
174
174 A list with each line of input as a separate element is returned when a
175 A list with each line of input as a separate element is returned when a
175 termination string is entered (defaults to a single '.'). Input can also
176 termination string is entered (defaults to a single '.'). Input can also
176 terminate via EOF (^D in Unix, ^Z-RET in Windows).
177 terminate via EOF (^D in Unix, ^Z-RET in Windows).
177
178
178 Lines of input which end in \\ are joined into single entries (and a
179 Lines of input which end in \\ are joined into single entries (and a
179 secondary continuation prompt is issued as long as the user terminates
180 secondary continuation prompt is issued as long as the user terminates
180 lines with \\). This allows entering very long strings which are still
181 lines with \\). This allows entering very long strings which are still
181 meant to be treated as single entities.
182 meant to be treated as single entities.
182 """
183 """
183
184
184 try:
185 try:
185 if header:
186 if header:
186 header += '\n'
187 header += '\n'
187 lines = [raw_input(header + ps1)]
188 lines = [raw_input(header + ps1)]
188 except EOFError:
189 except EOFError:
189 return []
190 return []
190 terminate = [terminate_str]
191 terminate = [terminate_str]
191 try:
192 try:
192 while lines[-1:] != terminate:
193 while lines[-1:] != terminate:
193 new_line = raw_input(ps1)
194 new_line = raw_input(ps1)
194 while new_line.endswith('\\'):
195 while new_line.endswith('\\'):
195 new_line = new_line[:-1] + raw_input(ps2)
196 new_line = new_line[:-1] + raw_input(ps2)
196 lines.append(new_line)
197 lines.append(new_line)
197
198
198 return lines[:-1] # don't return the termination command
199 return lines[:-1] # don't return the termination command
199 except EOFError:
200 except EOFError:
200 print()
201 print()
201 return lines
202 return lines
202
203
203
204
204 def raw_input_ext(prompt='', ps2='... '):
205 def raw_input_ext(prompt='', ps2='... '):
205 """Similar to raw_input(), but accepts extended lines if input ends with \\."""
206 """Similar to raw_input(), but accepts extended lines if input ends with \\."""
206
207
207 line = raw_input(prompt)
208 line = raw_input(prompt)
208 while line.endswith('\\'):
209 while line.endswith('\\'):
209 line = line[:-1] + raw_input(ps2)
210 line = line[:-1] + raw_input(ps2)
210 return line
211 return line
211
212
212
213
213 def ask_yes_no(prompt,default=None):
214 def ask_yes_no(prompt,default=None):
214 """Asks a question and returns a boolean (y/n) answer.
215 """Asks a question and returns a boolean (y/n) answer.
215
216
216 If default is given (one of 'y','n'), it is used if the user input is
217 If default is given (one of 'y','n'), it is used if the user input is
217 empty. Otherwise the question is repeated until an answer is given.
218 empty. Otherwise the question is repeated until an answer is given.
218
219
219 An EOF is treated as the default answer. If there is no default, an
220 An EOF is treated as the default answer. If there is no default, an
220 exception is raised to prevent infinite loops.
221 exception is raised to prevent infinite loops.
221
222
222 Valid answers are: y/yes/n/no (match is not case sensitive)."""
223 Valid answers are: y/yes/n/no (match is not case sensitive)."""
223
224
224 answers = {'y':True,'n':False,'yes':True,'no':False}
225 answers = {'y':True,'n':False,'yes':True,'no':False}
225 ans = None
226 ans = None
226 while ans not in answers.keys():
227 while ans not in answers.keys():
227 try:
228 try:
228 ans = raw_input(prompt+' ').lower()
229 ans = raw_input(prompt+' ').lower()
229 if not ans: # response was an empty string
230 if not ans: # response was an empty string
230 ans = default
231 ans = default
231 except KeyboardInterrupt:
232 except KeyboardInterrupt:
232 pass
233 pass
233 except EOFError:
234 except EOFError:
234 if default in answers.keys():
235 if default in answers.keys():
235 ans = default
236 ans = default
236 print()
237 print()
237 else:
238 else:
238 raise
239 raise
239
240
240 return answers[ans]
241 return answers[ans]
241
242
242
243
243 class NLprinter:
244 class NLprinter:
244 """Print an arbitrarily nested list, indicating index numbers.
245 """Print an arbitrarily nested list, indicating index numbers.
245
246
246 An instance of this class called nlprint is available and callable as a
247 An instance of this class called nlprint is available and callable as a
247 function.
248 function.
248
249
249 nlprint(list,indent=' ',sep=': ') -> prints indenting each level by 'indent'
250 nlprint(list,indent=' ',sep=': ') -> prints indenting each level by 'indent'
250 and using 'sep' to separate the index from the value. """
251 and using 'sep' to separate the index from the value. """
251
252
252 def __init__(self):
253 def __init__(self):
253 self.depth = 0
254 self.depth = 0
254
255
255 def __call__(self,lst,pos='',**kw):
256 def __call__(self,lst,pos='',**kw):
256 """Prints the nested list numbering levels."""
257 """Prints the nested list numbering levels."""
257 kw.setdefault('indent',' ')
258 kw.setdefault('indent',' ')
258 kw.setdefault('sep',': ')
259 kw.setdefault('sep',': ')
259 kw.setdefault('start',0)
260 kw.setdefault('start',0)
260 kw.setdefault('stop',len(lst))
261 kw.setdefault('stop',len(lst))
261 # we need to remove start and stop from kw so they don't propagate
262 # we need to remove start and stop from kw so they don't propagate
262 # into a recursive call for a nested list.
263 # into a recursive call for a nested list.
263 start = kw['start']; del kw['start']
264 start = kw['start']; del kw['start']
264 stop = kw['stop']; del kw['stop']
265 stop = kw['stop']; del kw['stop']
265 if self.depth == 0 and 'header' in kw.keys():
266 if self.depth == 0 and 'header' in kw.keys():
266 print(kw['header'])
267 print(kw['header'])
267
268
268 for idx in range(start,stop):
269 for idx in range(start,stop):
269 elem = lst[idx]
270 elem = lst[idx]
270 newpos = pos + str(idx)
271 newpos = pos + str(idx)
271 if type(elem)==type([]):
272 if type(elem)==type([]):
272 self.depth += 1
273 self.depth += 1
273 self.__call__(elem, newpos+",", **kw)
274 self.__call__(elem, newpos+",", **kw)
274 self.depth -= 1
275 self.depth -= 1
275 else:
276 else:
276 print(kw['indent']*self.depth + newpos + kw["sep"] + repr(elem))
277 print(kw['indent']*self.depth + newpos + kw["sep"] + repr(elem))
277
278
278 nlprint = NLprinter()
279 nlprint = NLprinter()
279
280
280
281
281 def temp_pyfile(src, ext='.py'):
282 def temp_pyfile(src, ext='.py'):
282 """Make a temporary python file, return filename and filehandle.
283 """Make a temporary python file, return filename and filehandle.
283
284
284 Parameters
285 Parameters
285 ----------
286 ----------
286 src : string or list of strings (no need for ending newlines if list)
287 src : string or list of strings (no need for ending newlines if list)
287 Source code to be written to the file.
288 Source code to be written to the file.
288
289
289 ext : optional, string
290 ext : optional, string
290 Extension for the generated file.
291 Extension for the generated file.
291
292
292 Returns
293 Returns
293 -------
294 -------
294 (filename, open filehandle)
295 (filename, open filehandle)
295 It is the caller's responsibility to close the open file and unlink it.
296 It is the caller's responsibility to close the open file and unlink it.
296 """
297 """
297 fname = tempfile.mkstemp(ext)[1]
298 fname = tempfile.mkstemp(ext)[1]
298 f = open(fname,'w')
299 f = open(fname,'w')
299 f.write(src)
300 f.write(src)
300 f.flush()
301 f.flush()
301 return fname, f
302 return fname, f
302
303
303
304
304 def raw_print(*args, **kw):
305 def raw_print(*args, **kw):
305 """Raw print to sys.__stdout__, otherwise identical interface to print()."""
306 """Raw print to sys.__stdout__, otherwise identical interface to print()."""
306
307
307 print(*args, sep=kw.get('sep', ' '), end=kw.get('end', '\n'),
308 print(*args, sep=kw.get('sep', ' '), end=kw.get('end', '\n'),
308 file=sys.__stdout__)
309 file=sys.__stdout__)
309 sys.__stdout__.flush()
310 sys.__stdout__.flush()
310
311
311
312
312 def raw_print_err(*args, **kw):
313 def raw_print_err(*args, **kw):
313 """Raw print to sys.__stderr__, otherwise identical interface to print()."""
314 """Raw print to sys.__stderr__, otherwise identical interface to print()."""
314
315
315 print(*args, sep=kw.get('sep', ' '), end=kw.get('end', '\n'),
316 print(*args, sep=kw.get('sep', ' '), end=kw.get('end', '\n'),
316 file=sys.__stderr__)
317 file=sys.__stderr__)
317 sys.__stderr__.flush()
318 sys.__stderr__.flush()
318
319
319
320
320 # Short aliases for quick debugging, do NOT use these in production code.
321 # Short aliases for quick debugging, do NOT use these in production code.
321 rprint = raw_print
322 rprint = raw_print
322 rprinte = raw_print_err
323 rprinte = raw_print_err
@@ -1,165 +1,165 b''
1 """Utilities to manipulate JSON objects.
1 """Utilities to manipulate JSON objects.
2 """
2 """
3 #-----------------------------------------------------------------------------
3 #-----------------------------------------------------------------------------
4 # Copyright (C) 2010-2011 The IPython Development Team
4 # Copyright (C) 2010-2011 The IPython Development Team
5 #
5 #
6 # Distributed under the terms of the BSD License. The full license is in
6 # Distributed under the terms of the BSD License. The full license is in
7 # the file COPYING.txt, distributed as part of this software.
7 # the file COPYING.txt, distributed as part of this software.
8 #-----------------------------------------------------------------------------
8 #-----------------------------------------------------------------------------
9
9
10 #-----------------------------------------------------------------------------
10 #-----------------------------------------------------------------------------
11 # Imports
11 # Imports
12 #-----------------------------------------------------------------------------
12 #-----------------------------------------------------------------------------
13 # stdlib
13 # stdlib
14 import re
14 import re
15 import sys
15 import sys
16 import types
16 import types
17 from datetime import datetime
17 from datetime import datetime
18
18
19 from IPython.utils import py3compat
19 from IPython.utils import py3compat
20 from IPython.utils import text
20 from IPython.utils import text
21 next_attr_name = '__next__' if py3compat.PY3 else 'next'
21 next_attr_name = '__next__' if py3compat.PY3 else 'next'
22
22
23 #-----------------------------------------------------------------------------
23 #-----------------------------------------------------------------------------
24 # Globals and constants
24 # Globals and constants
25 #-----------------------------------------------------------------------------
25 #-----------------------------------------------------------------------------
26
26
27 # timestamp formats
27 # timestamp formats
28 ISO8601="%Y-%m-%dT%H:%M:%S.%f"
28 ISO8601="%Y-%m-%dT%H:%M:%S.%f"
29 ISO8601_PAT=re.compile(r"^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\.\d+$")
29 ISO8601_PAT=re.compile(r"^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\.\d+$")
30
30
31 #-----------------------------------------------------------------------------
31 #-----------------------------------------------------------------------------
32 # Classes and functions
32 # Classes and functions
33 #-----------------------------------------------------------------------------
33 #-----------------------------------------------------------------------------
34
34
35 def rekey(dikt):
35 def rekey(dikt):
36 """Rekey a dict that has been forced to use str keys where there should be
36 """Rekey a dict that has been forced to use str keys where there should be
37 ints by json."""
37 ints by json."""
38 for k in dikt.iterkeys():
38 for k in dikt.iterkeys():
39 if isinstance(k, basestring):
39 if isinstance(k, basestring):
40 ik=fk=None
40 ik=fk=None
41 try:
41 try:
42 ik = int(k)
42 ik = int(k)
43 except ValueError:
43 except ValueError:
44 try:
44 try:
45 fk = float(k)
45 fk = float(k)
46 except ValueError:
46 except ValueError:
47 continue
47 continue
48 if ik is not None:
48 if ik is not None:
49 nk = ik
49 nk = ik
50 else:
50 else:
51 nk = fk
51 nk = fk
52 if nk in dikt:
52 if nk in dikt:
53 raise KeyError("already have key %r"%nk)
53 raise KeyError("already have key %r"%nk)
54 dikt[nk] = dikt.pop(k)
54 dikt[nk] = dikt.pop(k)
55 return dikt
55 return dikt
56
56
57
57
58 def extract_dates(obj):
58 def extract_dates(obj):
59 """extract ISO8601 dates from unpacked JSON"""
59 """extract ISO8601 dates from unpacked JSON"""
60 if isinstance(obj, dict):
60 if isinstance(obj, dict):
61 obj = dict(obj) # don't clobber
61 obj = dict(obj) # don't clobber
62 for k,v in obj.iteritems():
62 for k,v in obj.iteritems():
63 obj[k] = extract_dates(v)
63 obj[k] = extract_dates(v)
64 elif isinstance(obj, (list, tuple)):
64 elif isinstance(obj, (list, tuple)):
65 obj = [ extract_dates(o) for o in obj ]
65 obj = [ extract_dates(o) for o in obj ]
66 elif isinstance(obj, basestring):
66 elif isinstance(obj, basestring):
67 if ISO8601_PAT.match(obj):
67 if ISO8601_PAT.match(obj):
68 obj = datetime.strptime(obj, ISO8601)
68 obj = datetime.strptime(obj, ISO8601)
69 return obj
69 return obj
70
70
71 def squash_dates(obj):
71 def squash_dates(obj):
72 """squash datetime objects into ISO8601 strings"""
72 """squash datetime objects into ISO8601 strings"""
73 if isinstance(obj, dict):
73 if isinstance(obj, dict):
74 obj = dict(obj) # don't clobber
74 obj = dict(obj) # don't clobber
75 for k,v in obj.iteritems():
75 for k,v in obj.iteritems():
76 obj[k] = squash_dates(v)
76 obj[k] = squash_dates(v)
77 elif isinstance(obj, (list, tuple)):
77 elif isinstance(obj, (list, tuple)):
78 obj = [ squash_dates(o) for o in obj ]
78 obj = [ squash_dates(o) for o in obj ]
79 elif isinstance(obj, datetime):
79 elif isinstance(obj, datetime):
80 obj = obj.strftime(ISO8601)
80 obj = obj.strftime(ISO8601)
81 return obj
81 return obj
82
82
83 def date_default(obj):
83 def date_default(obj):
84 """default function for packing datetime objects in JSON."""
84 """default function for packing datetime objects in JSON."""
85 if isinstance(obj, datetime):
85 if isinstance(obj, datetime):
86 return obj.strftime(ISO8601)
86 return obj.strftime(ISO8601)
87 else:
87 else:
88 raise TypeError("%r is not JSON serializable"%obj)
88 raise TypeError("%r is not JSON serializable"%obj)
89
89
90
90
91
91
92 def json_clean(obj):
92 def json_clean(obj):
93 """Clean an object to ensure it's safe to encode in JSON.
93 """Clean an object to ensure it's safe to encode in JSON.
94
94
95 Atomic, immutable objects are returned unmodified. Sets and tuples are
95 Atomic, immutable objects are returned unmodified. Sets and tuples are
96 converted to lists, lists are copied and dicts are also copied.
96 converted to lists, lists are copied and dicts are also copied.
97
97
98 Note: dicts whose keys could cause collisions upon encoding (such as a dict
98 Note: dicts whose keys could cause collisions upon encoding (such as a dict
99 with both the number 1 and the string '1' as keys) will cause a ValueError
99 with both the number 1 and the string '1' as keys) will cause a ValueError
100 to be raised.
100 to be raised.
101
101
102 Parameters
102 Parameters
103 ----------
103 ----------
104 obj : any python object
104 obj : any python object
105
105
106 Returns
106 Returns
107 -------
107 -------
108 out : object
108 out : object
109
109
110 A version of the input which will not cause an encoding error when
110 A version of the input which will not cause an encoding error when
111 encoded as JSON. Note that this function does not *encode* its inputs,
111 encoded as JSON. Note that this function does not *encode* its inputs,
112 it simply sanitizes it so that there will be no encoding errors later.
112 it simply sanitizes it so that there will be no encoding errors later.
113
113
114 Examples
114 Examples
115 --------
115 --------
116 >>> json_clean(4)
116 >>> json_clean(4)
117 4
117 4
118 >>> json_clean(range(10))
118 >>> json_clean(range(10))
119 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
119 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
120 >>> json_clean(dict(x=1, y=2))
120 >>> json_clean(dict(x=1, y=2))
121 {'y': 2, 'x': 1}
121 {'y': 2, 'x': 1}
122 >>> json_clean(dict(x=1, y=2, z=[1,2,3]))
122 >>> json_clean(dict(x=1, y=2, z=[1,2,3]))
123 {'y': 2, 'x': 1, 'z': [1, 2, 3]}
123 {'y': 2, 'x': 1, 'z': [1, 2, 3]}
124 >>> json_clean(True)
124 >>> json_clean(True)
125 True
125 True
126 """
126 """
127 # types that are 'atomic' and ok in json as-is. bool doesn't need to be
127 # types that are 'atomic' and ok in json as-is. bool doesn't need to be
128 # listed explicitly because bools pass as int instances
128 # listed explicitly because bools pass as int instances
129 atomic_ok = (unicode, int, float, types.NoneType)
129 atomic_ok = (unicode, int, float, types.NoneType)
130
130
131 # containers that we need to convert into lists
131 # containers that we need to convert into lists
132 container_to_list = (tuple, set, types.GeneratorType)
132 container_to_list = (tuple, set, types.GeneratorType)
133
133
134 if isinstance(obj, atomic_ok):
134 if isinstance(obj, atomic_ok):
135 return obj
135 return obj
136
136
137 if isinstance(obj, bytes):
137 if isinstance(obj, bytes):
138 return obj.decode(text.getdefaultencoding(), 'replace')
138 return obj.decode(py3compat.getdefaultencoding(), 'replace')
139
139
140 if isinstance(obj, container_to_list) or (
140 if isinstance(obj, container_to_list) or (
141 hasattr(obj, '__iter__') and hasattr(obj, next_attr_name)):
141 hasattr(obj, '__iter__') and hasattr(obj, next_attr_name)):
142 obj = list(obj)
142 obj = list(obj)
143
143
144 if isinstance(obj, list):
144 if isinstance(obj, list):
145 return [json_clean(x) for x in obj]
145 return [json_clean(x) for x in obj]
146
146
147 if isinstance(obj, dict):
147 if isinstance(obj, dict):
148 # First, validate that the dict won't lose data in conversion due to
148 # First, validate that the dict won't lose data in conversion due to
149 # key collisions after stringification. This can happen with keys like
149 # key collisions after stringification. This can happen with keys like
150 # True and 'true' or 1 and '1', which collide in JSON.
150 # True and 'true' or 1 and '1', which collide in JSON.
151 nkeys = len(obj)
151 nkeys = len(obj)
152 nkeys_collapsed = len(set(map(str, obj)))
152 nkeys_collapsed = len(set(map(str, obj)))
153 if nkeys != nkeys_collapsed:
153 if nkeys != nkeys_collapsed:
154 raise ValueError('dict can not be safely converted to JSON: '
154 raise ValueError('dict can not be safely converted to JSON: '
155 'key collision would lead to dropped values')
155 'key collision would lead to dropped values')
156 # If all OK, proceed by making the new dict that will be json-safe
156 # If all OK, proceed by making the new dict that will be json-safe
157 out = {}
157 out = {}
158 for k,v in obj.iteritems():
158 for k,v in obj.iteritems():
159 out[str(k)] = json_clean(v)
159 out[str(k)] = json_clean(v)
160 return out
160 return out
161
161
162 # If we get here, we don't know how to handle the object, so we just get
162 # If we get here, we don't know how to handle the object, so we just get
163 # its repr and return that. This will catch lambdas, open sockets, class
163 # its repr and return that. This will catch lambdas, open sockets, class
164 # objects, and any other complicated contraption that json can't encode
164 # objects, and any other complicated contraption that json can't encode
165 return repr(obj)
165 return repr(obj)
@@ -1,183 +1,207 b''
1 # coding: utf-8
1 # coding: utf-8
2 """Compatibility tricks for Python 3. Mainly to do with unicode."""
2 """Compatibility tricks for Python 3. Mainly to do with unicode."""
3 import __builtin__
3 import __builtin__
4 import functools
4 import functools
5 import sys
5 import sys
6 import re
6 import re
7 import types
7 import types
8 import locale
8
9
9 orig_open = open
10 orig_open = open
10
11
11 def no_code(x, encoding=None):
12 def no_code(x, encoding=None):
12 return x
13 return x
13
14
14 # to deal with the possibility of sys.std* not being a stream at all
15 # to deal with the possibility of sys.std* not being a stream at all
15 def get_stream_enc(stream, default=None):
16 def get_stream_enc(stream, default=None):
16 if not hasattr(stream, 'encoding') or not stream.encoding:
17 if not hasattr(stream, 'encoding') or not stream.encoding:
17 return default
18 return default
18 else:
19 else:
19 return stream.encoding
20 return stream.encoding
20
21
22 # Less conservative replacement for sys.getdefaultencoding, that will try
23 # to match the environment.
24 # Defined here as central function, so if we find better choices, we
25 # won't need to make changes all over IPython.
26 def getdefaultencoding():
27 """Return IPython's guess for the default encoding for bytes as text.
28
29 Asks for stdin.encoding first, to match the calling Terminal, but that
30 is often None for subprocesses. Fall back on locale.getpreferredencoding()
31 which should be a sensible platform default (that respects LANG environment),
32 and finally to sys.getdefaultencoding() which is the most conservative option,
33 and usually ASCII.
34 """
35 enc = get_stream_enc(sys.stdin)
36 if not enc or enc=='ascii':
37 try:
38 # There are reports of getpreferredencoding raising errors
39 # in some cases, which may well be fixed, but let's be conservative here.
40 enc = locale.getpreferredencoding()
41 except Exception:
42 pass
43 return enc or sys.getdefaultencoding()
44
21 def decode(s, encoding=None):
45 def decode(s, encoding=None):
22 encoding = get_stream_enc(sys.stdin, encoding) or sys.getdefaultencoding()
46 encoding = get_stream_enc(sys.stdin, encoding) or getdefaultencoding()
23 return s.decode(encoding, "replace")
47 return s.decode(encoding, "replace")
24
48
25 def encode(u, encoding=None):
49 def encode(u, encoding=None):
26 encoding = get_stream_enc(sys.stdin, encoding) or sys.getdefaultencoding()
50 encoding = get_stream_enc(sys.stdin, encoding) or getdefaultencoding()
27 return u.encode(encoding, "replace")
51 return u.encode(encoding, "replace")
28
52
29
53
30 def cast_unicode(s, encoding=None):
54 def cast_unicode(s, encoding=None):
31 if isinstance(s, bytes):
55 if isinstance(s, bytes):
32 return decode(s, encoding)
56 return decode(s, encoding)
33 return s
57 return s
34
58
35 def cast_bytes(s, encoding=None):
59 def cast_bytes(s, encoding=None):
36 if not isinstance(s, bytes):
60 if not isinstance(s, bytes):
37 return encode(s, encoding)
61 return encode(s, encoding)
38 return s
62 return s
39
63
40 def _modify_str_or_docstring(str_change_func):
64 def _modify_str_or_docstring(str_change_func):
41 @functools.wraps(str_change_func)
65 @functools.wraps(str_change_func)
42 def wrapper(func_or_str):
66 def wrapper(func_or_str):
43 if isinstance(func_or_str, basestring):
67 if isinstance(func_or_str, basestring):
44 func = None
68 func = None
45 doc = func_or_str
69 doc = func_or_str
46 else:
70 else:
47 func = func_or_str
71 func = func_or_str
48 doc = func.__doc__
72 doc = func.__doc__
49
73
50 doc = str_change_func(doc)
74 doc = str_change_func(doc)
51
75
52 if func:
76 if func:
53 func.__doc__ = doc
77 func.__doc__ = doc
54 return func
78 return func
55 return doc
79 return doc
56 return wrapper
80 return wrapper
57
81
58 if sys.version_info[0] >= 3:
82 if sys.version_info[0] >= 3:
59 PY3 = True
83 PY3 = True
60
84
61 input = input
85 input = input
62 builtin_mod_name = "builtins"
86 builtin_mod_name = "builtins"
63
87
64 str_to_unicode = no_code
88 str_to_unicode = no_code
65 unicode_to_str = no_code
89 unicode_to_str = no_code
66 str_to_bytes = encode
90 str_to_bytes = encode
67 bytes_to_str = decode
91 bytes_to_str = decode
68 cast_bytes_py2 = no_code
92 cast_bytes_py2 = no_code
69
93
70 def isidentifier(s, dotted=False):
94 def isidentifier(s, dotted=False):
71 if dotted:
95 if dotted:
72 return all(isidentifier(a) for a in s.split("."))
96 return all(isidentifier(a) for a in s.split("."))
73 return s.isidentifier()
97 return s.isidentifier()
74
98
75 open = orig_open
99 open = orig_open
76
100
77 MethodType = types.MethodType
101 MethodType = types.MethodType
78
102
79 def execfile(fname, glob, loc=None):
103 def execfile(fname, glob, loc=None):
80 loc = loc if (loc is not None) else glob
104 loc = loc if (loc is not None) else glob
81 exec compile(open(fname, 'rb').read(), fname, 'exec') in glob, loc
105 exec compile(open(fname, 'rb').read(), fname, 'exec') in glob, loc
82
106
83 # Refactor print statements in doctests.
107 # Refactor print statements in doctests.
84 _print_statement_re = re.compile(r"\bprint (?P<expr>.*)$", re.MULTILINE)
108 _print_statement_re = re.compile(r"\bprint (?P<expr>.*)$", re.MULTILINE)
85 def _print_statement_sub(match):
109 def _print_statement_sub(match):
86 expr = match.groups('expr')
110 expr = match.groups('expr')
87 return "print(%s)" % expr
111 return "print(%s)" % expr
88
112
89 @_modify_str_or_docstring
113 @_modify_str_or_docstring
90 def doctest_refactor_print(doc):
114 def doctest_refactor_print(doc):
91 """Refactor 'print x' statements in a doctest to print(x) style. 2to3
115 """Refactor 'print x' statements in a doctest to print(x) style. 2to3
92 unfortunately doesn't pick up on our doctests.
116 unfortunately doesn't pick up on our doctests.
93
117
94 Can accept a string or a function, so it can be used as a decorator."""
118 Can accept a string or a function, so it can be used as a decorator."""
95 return _print_statement_re.sub(_print_statement_sub, doc)
119 return _print_statement_re.sub(_print_statement_sub, doc)
96
120
97 # Abstract u'abc' syntax:
121 # Abstract u'abc' syntax:
98 @_modify_str_or_docstring
122 @_modify_str_or_docstring
99 def u_format(s):
123 def u_format(s):
100 """"{u}'abc'" --> "'abc'" (Python 3)
124 """"{u}'abc'" --> "'abc'" (Python 3)
101
125
102 Accepts a string or a function, so it can be used as a decorator."""
126 Accepts a string or a function, so it can be used as a decorator."""
103 return s.format(u='')
127 return s.format(u='')
104
128
105 else:
129 else:
106 PY3 = False
130 PY3 = False
107
131
108 input = raw_input
132 input = raw_input
109 builtin_mod_name = "__builtin__"
133 builtin_mod_name = "__builtin__"
110
134
111 str_to_unicode = decode
135 str_to_unicode = decode
112 unicode_to_str = encode
136 unicode_to_str = encode
113 str_to_bytes = no_code
137 str_to_bytes = no_code
114 bytes_to_str = no_code
138 bytes_to_str = no_code
115 cast_bytes_py2 = cast_bytes
139 cast_bytes_py2 = cast_bytes
116
140
117 import re
141 import re
118 _name_re = re.compile(r"[a-zA-Z_][a-zA-Z0-9_]*$")
142 _name_re = re.compile(r"[a-zA-Z_][a-zA-Z0-9_]*$")
119 def isidentifier(s, dotted=False):
143 def isidentifier(s, dotted=False):
120 if dotted:
144 if dotted:
121 return all(isidentifier(a) for a in s.split("."))
145 return all(isidentifier(a) for a in s.split("."))
122 return bool(_name_re.match(s))
146 return bool(_name_re.match(s))
123
147
124 class open(object):
148 class open(object):
125 """Wrapper providing key part of Python 3 open() interface."""
149 """Wrapper providing key part of Python 3 open() interface."""
126 def __init__(self, fname, mode="r", encoding="utf-8"):
150 def __init__(self, fname, mode="r", encoding="utf-8"):
127 self.f = orig_open(fname, mode)
151 self.f = orig_open(fname, mode)
128 self.enc = encoding
152 self.enc = encoding
129
153
130 def write(self, s):
154 def write(self, s):
131 return self.f.write(s.encode(self.enc))
155 return self.f.write(s.encode(self.enc))
132
156
133 def read(self, size=-1):
157 def read(self, size=-1):
134 return self.f.read(size).decode(self.enc)
158 return self.f.read(size).decode(self.enc)
135
159
136 def close(self):
160 def close(self):
137 return self.f.close()
161 return self.f.close()
138
162
139 def __enter__(self):
163 def __enter__(self):
140 return self
164 return self
141
165
142 def __exit__(self, etype, value, traceback):
166 def __exit__(self, etype, value, traceback):
143 self.f.close()
167 self.f.close()
144
168
145 def MethodType(func, instance):
169 def MethodType(func, instance):
146 return types.MethodType(func, instance, type(instance))
170 return types.MethodType(func, instance, type(instance))
147
171
148 # don't override system execfile on 2.x:
172 # don't override system execfile on 2.x:
149 execfile = execfile
173 execfile = execfile
150
174
151 def doctest_refactor_print(func_or_str):
175 def doctest_refactor_print(func_or_str):
152 return func_or_str
176 return func_or_str
153
177
154
178
155 # Abstract u'abc' syntax:
179 # Abstract u'abc' syntax:
156 @_modify_str_or_docstring
180 @_modify_str_or_docstring
157 def u_format(s):
181 def u_format(s):
158 """"{u}'abc'" --> "u'abc'" (Python 2)
182 """"{u}'abc'" --> "u'abc'" (Python 2)
159
183
160 Accepts a string or a function, so it can be used as a decorator."""
184 Accepts a string or a function, so it can be used as a decorator."""
161 return s.format(u='u')
185 return s.format(u='u')
162
186
163 if sys.platform == 'win32':
187 if sys.platform == 'win32':
164 def execfile(fname, glob=None, loc=None):
188 def execfile(fname, glob=None, loc=None):
165 loc = loc if (loc is not None) else glob
189 loc = loc if (loc is not None) else glob
166 # The rstrip() is necessary b/c trailing whitespace in files will
190 # The rstrip() is necessary b/c trailing whitespace in files will
167 # cause an IndentationError in Python 2.6 (this was fixed in 2.7,
191 # cause an IndentationError in Python 2.6 (this was fixed in 2.7,
168 # but we still support 2.6). See issue 1027.
192 # but we still support 2.6). See issue 1027.
169 scripttext = __builtin__.open(fname).read().rstrip() + '\n'
193 scripttext = __builtin__.open(fname).read().rstrip() + '\n'
170 # compile converts unicode filename to str assuming
194 # compile converts unicode filename to str assuming
171 # ascii. Let's do the conversion before calling compile
195 # ascii. Let's do the conversion before calling compile
172 if isinstance(fname, unicode):
196 if isinstance(fname, unicode):
173 filename = unicode_to_str(fname)
197 filename = unicode_to_str(fname)
174 else:
198 else:
175 filename = fname
199 filename = fname
176 exec compile(scripttext, filename, 'exec') in glob, loc
200 exec compile(scripttext, filename, 'exec') in glob, loc
177 else:
201 else:
178 def execfile(fname, *where):
202 def execfile(fname, *where):
179 if isinstance(fname, unicode):
203 if isinstance(fname, unicode):
180 filename = fname.encode(sys.getfilesystemencoding())
204 filename = fname.encode(sys.getfilesystemencoding())
181 else:
205 else:
182 filename = fname
206 filename = fname
183 __builtin__.execfile(filename, *where)
207 __builtin__.execfile(filename, *where)
@@ -1,760 +1,736 b''
1 # encoding: utf-8
1 # encoding: utf-8
2 """
2 """
3 Utilities for working with strings and text.
3 Utilities for working with strings and text.
4 """
4 """
5
5
6 #-----------------------------------------------------------------------------
6 #-----------------------------------------------------------------------------
7 # Copyright (C) 2008-2011 The IPython Development Team
7 # Copyright (C) 2008-2011 The IPython Development Team
8 #
8 #
9 # Distributed under the terms of the BSD License. The full license is in
9 # Distributed under the terms of the BSD License. The full license is in
10 # the file COPYING, distributed as part of this software.
10 # the file COPYING, distributed as part of this software.
11 #-----------------------------------------------------------------------------
11 #-----------------------------------------------------------------------------
12
12
13 #-----------------------------------------------------------------------------
13 #-----------------------------------------------------------------------------
14 # Imports
14 # Imports
15 #-----------------------------------------------------------------------------
15 #-----------------------------------------------------------------------------
16
16
17 import __main__
17 import __main__
18
18
19 import locale
20 import os
19 import os
21 import re
20 import re
22 import shutil
21 import shutil
23 import sys
22 import sys
24 import textwrap
23 import textwrap
25 from string import Formatter
24 from string import Formatter
26
25
27 from IPython.external.path import path
26 from IPython.external.path import path
28 from IPython.testing.skipdoctest import skip_doctest_py3
27 from IPython.testing.skipdoctest import skip_doctest_py3
29 from IPython.utils import py3compat
28 from IPython.utils import py3compat
30 from IPython.utils.io import nlprint
29 from IPython.utils.io import nlprint
31 from IPython.utils.data import flatten
30 from IPython.utils.data import flatten
32
31
33 #-----------------------------------------------------------------------------
32 #-----------------------------------------------------------------------------
34 # Code
33 # Code
35 #-----------------------------------------------------------------------------
34 #-----------------------------------------------------------------------------
36
35
37 # Less conservative replacement for sys.getdefaultencoding, that will try
38 # to match the environment.
39 # Defined here as central function, so if we find better choices, we
40 # won't need to make changes all over IPython.
41 def getdefaultencoding():
42 """Return IPython's guess for the default encoding for bytes as text.
43
44 Asks for stdin.encoding first, to match the calling Terminal, but that
45 is often None for subprocesses. Fall back on locale.getpreferredencoding()
46 which should be a sensible platform default (that respects LANG environment),
47 and finally to sys.getdefaultencoding() which is the most conservative option,
48 and usually ASCII.
49 """
50 enc = py3compat.get_stream_enc(sys.stdin)
51 if not enc or enc=='ascii':
52 try:
53 # There are reports of getpreferredencoding raising errors
54 # in some cases, which may well be fixed, but let's be conservative here.
55 enc = locale.getpreferredencoding()
56 except Exception:
57 pass
58 return enc or sys.getdefaultencoding()
59
60 def unquote_ends(istr):
36 def unquote_ends(istr):
61 """Remove a single pair of quotes from the endpoints of a string."""
37 """Remove a single pair of quotes from the endpoints of a string."""
62
38
63 if not istr:
39 if not istr:
64 return istr
40 return istr
65 if (istr[0]=="'" and istr[-1]=="'") or \
41 if (istr[0]=="'" and istr[-1]=="'") or \
66 (istr[0]=='"' and istr[-1]=='"'):
42 (istr[0]=='"' and istr[-1]=='"'):
67 return istr[1:-1]
43 return istr[1:-1]
68 else:
44 else:
69 return istr
45 return istr
70
46
71
47
72 class LSString(str):
48 class LSString(str):
73 """String derivative with a special access attributes.
49 """String derivative with a special access attributes.
74
50
75 These are normal strings, but with the special attributes:
51 These are normal strings, but with the special attributes:
76
52
77 .l (or .list) : value as list (split on newlines).
53 .l (or .list) : value as list (split on newlines).
78 .n (or .nlstr): original value (the string itself).
54 .n (or .nlstr): original value (the string itself).
79 .s (or .spstr): value as whitespace-separated string.
55 .s (or .spstr): value as whitespace-separated string.
80 .p (or .paths): list of path objects
56 .p (or .paths): list of path objects
81
57
82 Any values which require transformations are computed only once and
58 Any values which require transformations are computed only once and
83 cached.
59 cached.
84
60
85 Such strings are very useful to efficiently interact with the shell, which
61 Such strings are very useful to efficiently interact with the shell, which
86 typically only understands whitespace-separated options for commands."""
62 typically only understands whitespace-separated options for commands."""
87
63
88 def get_list(self):
64 def get_list(self):
89 try:
65 try:
90 return self.__list
66 return self.__list
91 except AttributeError:
67 except AttributeError:
92 self.__list = self.split('\n')
68 self.__list = self.split('\n')
93 return self.__list
69 return self.__list
94
70
95 l = list = property(get_list)
71 l = list = property(get_list)
96
72
97 def get_spstr(self):
73 def get_spstr(self):
98 try:
74 try:
99 return self.__spstr
75 return self.__spstr
100 except AttributeError:
76 except AttributeError:
101 self.__spstr = self.replace('\n',' ')
77 self.__spstr = self.replace('\n',' ')
102 return self.__spstr
78 return self.__spstr
103
79
104 s = spstr = property(get_spstr)
80 s = spstr = property(get_spstr)
105
81
106 def get_nlstr(self):
82 def get_nlstr(self):
107 return self
83 return self
108
84
109 n = nlstr = property(get_nlstr)
85 n = nlstr = property(get_nlstr)
110
86
111 def get_paths(self):
87 def get_paths(self):
112 try:
88 try:
113 return self.__paths
89 return self.__paths
114 except AttributeError:
90 except AttributeError:
115 self.__paths = [path(p) for p in self.split('\n') if os.path.exists(p)]
91 self.__paths = [path(p) for p in self.split('\n') if os.path.exists(p)]
116 return self.__paths
92 return self.__paths
117
93
118 p = paths = property(get_paths)
94 p = paths = property(get_paths)
119
95
120 # FIXME: We need to reimplement type specific displayhook and then add this
96 # FIXME: We need to reimplement type specific displayhook and then add this
121 # back as a custom printer. This should also be moved outside utils into the
97 # back as a custom printer. This should also be moved outside utils into the
122 # core.
98 # core.
123
99
124 # def print_lsstring(arg):
100 # def print_lsstring(arg):
125 # """ Prettier (non-repr-like) and more informative printer for LSString """
101 # """ Prettier (non-repr-like) and more informative printer for LSString """
126 # print "LSString (.p, .n, .l, .s available). Value:"
102 # print "LSString (.p, .n, .l, .s available). Value:"
127 # print arg
103 # print arg
128 #
104 #
129 #
105 #
130 # print_lsstring = result_display.when_type(LSString)(print_lsstring)
106 # print_lsstring = result_display.when_type(LSString)(print_lsstring)
131
107
132
108
133 class SList(list):
109 class SList(list):
134 """List derivative with a special access attributes.
110 """List derivative with a special access attributes.
135
111
136 These are normal lists, but with the special attributes:
112 These are normal lists, but with the special attributes:
137
113
138 .l (or .list) : value as list (the list itself).
114 .l (or .list) : value as list (the list itself).
139 .n (or .nlstr): value as a string, joined on newlines.
115 .n (or .nlstr): value as a string, joined on newlines.
140 .s (or .spstr): value as a string, joined on spaces.
116 .s (or .spstr): value as a string, joined on spaces.
141 .p (or .paths): list of path objects
117 .p (or .paths): list of path objects
142
118
143 Any values which require transformations are computed only once and
119 Any values which require transformations are computed only once and
144 cached."""
120 cached."""
145
121
146 def get_list(self):
122 def get_list(self):
147 return self
123 return self
148
124
149 l = list = property(get_list)
125 l = list = property(get_list)
150
126
151 def get_spstr(self):
127 def get_spstr(self):
152 try:
128 try:
153 return self.__spstr
129 return self.__spstr
154 except AttributeError:
130 except AttributeError:
155 self.__spstr = ' '.join(self)
131 self.__spstr = ' '.join(self)
156 return self.__spstr
132 return self.__spstr
157
133
158 s = spstr = property(get_spstr)
134 s = spstr = property(get_spstr)
159
135
160 def get_nlstr(self):
136 def get_nlstr(self):
161 try:
137 try:
162 return self.__nlstr
138 return self.__nlstr
163 except AttributeError:
139 except AttributeError:
164 self.__nlstr = '\n'.join(self)
140 self.__nlstr = '\n'.join(self)
165 return self.__nlstr
141 return self.__nlstr
166
142
167 n = nlstr = property(get_nlstr)
143 n = nlstr = property(get_nlstr)
168
144
169 def get_paths(self):
145 def get_paths(self):
170 try:
146 try:
171 return self.__paths
147 return self.__paths
172 except AttributeError:
148 except AttributeError:
173 self.__paths = [path(p) for p in self if os.path.exists(p)]
149 self.__paths = [path(p) for p in self if os.path.exists(p)]
174 return self.__paths
150 return self.__paths
175
151
176 p = paths = property(get_paths)
152 p = paths = property(get_paths)
177
153
178 def grep(self, pattern, prune = False, field = None):
154 def grep(self, pattern, prune = False, field = None):
179 """ Return all strings matching 'pattern' (a regex or callable)
155 """ Return all strings matching 'pattern' (a regex or callable)
180
156
181 This is case-insensitive. If prune is true, return all items
157 This is case-insensitive. If prune is true, return all items
182 NOT matching the pattern.
158 NOT matching the pattern.
183
159
184 If field is specified, the match must occur in the specified
160 If field is specified, the match must occur in the specified
185 whitespace-separated field.
161 whitespace-separated field.
186
162
187 Examples::
163 Examples::
188
164
189 a.grep( lambda x: x.startswith('C') )
165 a.grep( lambda x: x.startswith('C') )
190 a.grep('Cha.*log', prune=1)
166 a.grep('Cha.*log', prune=1)
191 a.grep('chm', field=-1)
167 a.grep('chm', field=-1)
192 """
168 """
193
169
194 def match_target(s):
170 def match_target(s):
195 if field is None:
171 if field is None:
196 return s
172 return s
197 parts = s.split()
173 parts = s.split()
198 try:
174 try:
199 tgt = parts[field]
175 tgt = parts[field]
200 return tgt
176 return tgt
201 except IndexError:
177 except IndexError:
202 return ""
178 return ""
203
179
204 if isinstance(pattern, basestring):
180 if isinstance(pattern, basestring):
205 pred = lambda x : re.search(pattern, x, re.IGNORECASE)
181 pred = lambda x : re.search(pattern, x, re.IGNORECASE)
206 else:
182 else:
207 pred = pattern
183 pred = pattern
208 if not prune:
184 if not prune:
209 return SList([el for el in self if pred(match_target(el))])
185 return SList([el for el in self if pred(match_target(el))])
210 else:
186 else:
211 return SList([el for el in self if not pred(match_target(el))])
187 return SList([el for el in self if not pred(match_target(el))])
212
188
213 def fields(self, *fields):
189 def fields(self, *fields):
214 """ Collect whitespace-separated fields from string list
190 """ Collect whitespace-separated fields from string list
215
191
216 Allows quick awk-like usage of string lists.
192 Allows quick awk-like usage of string lists.
217
193
218 Example data (in var a, created by 'a = !ls -l')::
194 Example data (in var a, created by 'a = !ls -l')::
219 -rwxrwxrwx 1 ville None 18 Dec 14 2006 ChangeLog
195 -rwxrwxrwx 1 ville None 18 Dec 14 2006 ChangeLog
220 drwxrwxrwx+ 6 ville None 0 Oct 24 18:05 IPython
196 drwxrwxrwx+ 6 ville None 0 Oct 24 18:05 IPython
221
197
222 a.fields(0) is ['-rwxrwxrwx', 'drwxrwxrwx+']
198 a.fields(0) is ['-rwxrwxrwx', 'drwxrwxrwx+']
223 a.fields(1,0) is ['1 -rwxrwxrwx', '6 drwxrwxrwx+']
199 a.fields(1,0) is ['1 -rwxrwxrwx', '6 drwxrwxrwx+']
224 (note the joining by space).
200 (note the joining by space).
225 a.fields(-1) is ['ChangeLog', 'IPython']
201 a.fields(-1) is ['ChangeLog', 'IPython']
226
202
227 IndexErrors are ignored.
203 IndexErrors are ignored.
228
204
229 Without args, fields() just split()'s the strings.
205 Without args, fields() just split()'s the strings.
230 """
206 """
231 if len(fields) == 0:
207 if len(fields) == 0:
232 return [el.split() for el in self]
208 return [el.split() for el in self]
233
209
234 res = SList()
210 res = SList()
235 for el in [f.split() for f in self]:
211 for el in [f.split() for f in self]:
236 lineparts = []
212 lineparts = []
237
213
238 for fd in fields:
214 for fd in fields:
239 try:
215 try:
240 lineparts.append(el[fd])
216 lineparts.append(el[fd])
241 except IndexError:
217 except IndexError:
242 pass
218 pass
243 if lineparts:
219 if lineparts:
244 res.append(" ".join(lineparts))
220 res.append(" ".join(lineparts))
245
221
246 return res
222 return res
247
223
248 def sort(self,field= None, nums = False):
224 def sort(self,field= None, nums = False):
249 """ sort by specified fields (see fields())
225 """ sort by specified fields (see fields())
250
226
251 Example::
227 Example::
252 a.sort(1, nums = True)
228 a.sort(1, nums = True)
253
229
254 Sorts a by second field, in numerical order (so that 21 > 3)
230 Sorts a by second field, in numerical order (so that 21 > 3)
255
231
256 """
232 """
257
233
258 #decorate, sort, undecorate
234 #decorate, sort, undecorate
259 if field is not None:
235 if field is not None:
260 dsu = [[SList([line]).fields(field), line] for line in self]
236 dsu = [[SList([line]).fields(field), line] for line in self]
261 else:
237 else:
262 dsu = [[line, line] for line in self]
238 dsu = [[line, line] for line in self]
263 if nums:
239 if nums:
264 for i in range(len(dsu)):
240 for i in range(len(dsu)):
265 numstr = "".join([ch for ch in dsu[i][0] if ch.isdigit()])
241 numstr = "".join([ch for ch in dsu[i][0] if ch.isdigit()])
266 try:
242 try:
267 n = int(numstr)
243 n = int(numstr)
268 except ValueError:
244 except ValueError:
269 n = 0;
245 n = 0;
270 dsu[i][0] = n
246 dsu[i][0] = n
271
247
272
248
273 dsu.sort()
249 dsu.sort()
274 return SList([t[1] for t in dsu])
250 return SList([t[1] for t in dsu])
275
251
276
252
277 # FIXME: We need to reimplement type specific displayhook and then add this
253 # FIXME: We need to reimplement type specific displayhook and then add this
278 # back as a custom printer. This should also be moved outside utils into the
254 # back as a custom printer. This should also be moved outside utils into the
279 # core.
255 # core.
280
256
281 # def print_slist(arg):
257 # def print_slist(arg):
282 # """ Prettier (non-repr-like) and more informative printer for SList """
258 # """ Prettier (non-repr-like) and more informative printer for SList """
283 # print "SList (.p, .n, .l, .s, .grep(), .fields(), sort() available):"
259 # print "SList (.p, .n, .l, .s, .grep(), .fields(), sort() available):"
284 # if hasattr(arg, 'hideonce') and arg.hideonce:
260 # if hasattr(arg, 'hideonce') and arg.hideonce:
285 # arg.hideonce = False
261 # arg.hideonce = False
286 # return
262 # return
287 #
263 #
288 # nlprint(arg)
264 # nlprint(arg)
289 #
265 #
290 # print_slist = result_display.when_type(SList)(print_slist)
266 # print_slist = result_display.when_type(SList)(print_slist)
291
267
292
268
293 def esc_quotes(strng):
269 def esc_quotes(strng):
294 """Return the input string with single and double quotes escaped out"""
270 """Return the input string with single and double quotes escaped out"""
295
271
296 return strng.replace('"','\\"').replace("'","\\'")
272 return strng.replace('"','\\"').replace("'","\\'")
297
273
298
274
299 def qw(words,flat=0,sep=None,maxsplit=-1):
275 def qw(words,flat=0,sep=None,maxsplit=-1):
300 """Similar to Perl's qw() operator, but with some more options.
276 """Similar to Perl's qw() operator, but with some more options.
301
277
302 qw(words,flat=0,sep=' ',maxsplit=-1) -> words.split(sep,maxsplit)
278 qw(words,flat=0,sep=' ',maxsplit=-1) -> words.split(sep,maxsplit)
303
279
304 words can also be a list itself, and with flat=1, the output will be
280 words can also be a list itself, and with flat=1, the output will be
305 recursively flattened.
281 recursively flattened.
306
282
307 Examples:
283 Examples:
308
284
309 >>> qw('1 2')
285 >>> qw('1 2')
310 ['1', '2']
286 ['1', '2']
311
287
312 >>> qw(['a b','1 2',['m n','p q']])
288 >>> qw(['a b','1 2',['m n','p q']])
313 [['a', 'b'], ['1', '2'], [['m', 'n'], ['p', 'q']]]
289 [['a', 'b'], ['1', '2'], [['m', 'n'], ['p', 'q']]]
314
290
315 >>> qw(['a b','1 2',['m n','p q']],flat=1)
291 >>> qw(['a b','1 2',['m n','p q']],flat=1)
316 ['a', 'b', '1', '2', 'm', 'n', 'p', 'q']
292 ['a', 'b', '1', '2', 'm', 'n', 'p', 'q']
317 """
293 """
318
294
319 if isinstance(words, basestring):
295 if isinstance(words, basestring):
320 return [word.strip() for word in words.split(sep,maxsplit)
296 return [word.strip() for word in words.split(sep,maxsplit)
321 if word and not word.isspace() ]
297 if word and not word.isspace() ]
322 if flat:
298 if flat:
323 return flatten(map(qw,words,[1]*len(words)))
299 return flatten(map(qw,words,[1]*len(words)))
324 return map(qw,words)
300 return map(qw,words)
325
301
326
302
327 def qwflat(words,sep=None,maxsplit=-1):
303 def qwflat(words,sep=None,maxsplit=-1):
328 """Calls qw(words) in flat mode. It's just a convenient shorthand."""
304 """Calls qw(words) in flat mode. It's just a convenient shorthand."""
329 return qw(words,1,sep,maxsplit)
305 return qw(words,1,sep,maxsplit)
330
306
331
307
332 def qw_lol(indata):
308 def qw_lol(indata):
333 """qw_lol('a b') -> [['a','b']],
309 """qw_lol('a b') -> [['a','b']],
334 otherwise it's just a call to qw().
310 otherwise it's just a call to qw().
335
311
336 We need this to make sure the modules_some keys *always* end up as a
312 We need this to make sure the modules_some keys *always* end up as a
337 list of lists."""
313 list of lists."""
338
314
339 if isinstance(indata, basestring):
315 if isinstance(indata, basestring):
340 return [qw(indata)]
316 return [qw(indata)]
341 else:
317 else:
342 return qw(indata)
318 return qw(indata)
343
319
344
320
345 def grep(pat,list,case=1):
321 def grep(pat,list,case=1):
346 """Simple minded grep-like function.
322 """Simple minded grep-like function.
347 grep(pat,list) returns occurrences of pat in list, None on failure.
323 grep(pat,list) returns occurrences of pat in list, None on failure.
348
324
349 It only does simple string matching, with no support for regexps. Use the
325 It only does simple string matching, with no support for regexps. Use the
350 option case=0 for case-insensitive matching."""
326 option case=0 for case-insensitive matching."""
351
327
352 # This is pretty crude. At least it should implement copying only references
328 # This is pretty crude. At least it should implement copying only references
353 # to the original data in case it's big. Now it copies the data for output.
329 # to the original data in case it's big. Now it copies the data for output.
354 out=[]
330 out=[]
355 if case:
331 if case:
356 for term in list:
332 for term in list:
357 if term.find(pat)>-1: out.append(term)
333 if term.find(pat)>-1: out.append(term)
358 else:
334 else:
359 lpat=pat.lower()
335 lpat=pat.lower()
360 for term in list:
336 for term in list:
361 if term.lower().find(lpat)>-1: out.append(term)
337 if term.lower().find(lpat)>-1: out.append(term)
362
338
363 if len(out): return out
339 if len(out): return out
364 else: return None
340 else: return None
365
341
366
342
367 def dgrep(pat,*opts):
343 def dgrep(pat,*opts):
368 """Return grep() on dir()+dir(__builtins__).
344 """Return grep() on dir()+dir(__builtins__).
369
345
370 A very common use of grep() when working interactively."""
346 A very common use of grep() when working interactively."""
371
347
372 return grep(pat,dir(__main__)+dir(__main__.__builtins__),*opts)
348 return grep(pat,dir(__main__)+dir(__main__.__builtins__),*opts)
373
349
374
350
375 def idgrep(pat):
351 def idgrep(pat):
376 """Case-insensitive dgrep()"""
352 """Case-insensitive dgrep()"""
377
353
378 return dgrep(pat,0)
354 return dgrep(pat,0)
379
355
380
356
381 def igrep(pat,list):
357 def igrep(pat,list):
382 """Synonym for case-insensitive grep."""
358 """Synonym for case-insensitive grep."""
383
359
384 return grep(pat,list,case=0)
360 return grep(pat,list,case=0)
385
361
386
362
387 def indent(instr,nspaces=4, ntabs=0, flatten=False):
363 def indent(instr,nspaces=4, ntabs=0, flatten=False):
388 """Indent a string a given number of spaces or tabstops.
364 """Indent a string a given number of spaces or tabstops.
389
365
390 indent(str,nspaces=4,ntabs=0) -> indent str by ntabs+nspaces.
366 indent(str,nspaces=4,ntabs=0) -> indent str by ntabs+nspaces.
391
367
392 Parameters
368 Parameters
393 ----------
369 ----------
394
370
395 instr : basestring
371 instr : basestring
396 The string to be indented.
372 The string to be indented.
397 nspaces : int (default: 4)
373 nspaces : int (default: 4)
398 The number of spaces to be indented.
374 The number of spaces to be indented.
399 ntabs : int (default: 0)
375 ntabs : int (default: 0)
400 The number of tabs to be indented.
376 The number of tabs to be indented.
401 flatten : bool (default: False)
377 flatten : bool (default: False)
402 Whether to scrub existing indentation. If True, all lines will be
378 Whether to scrub existing indentation. If True, all lines will be
403 aligned to the same indentation. If False, existing indentation will
379 aligned to the same indentation. If False, existing indentation will
404 be strictly increased.
380 be strictly increased.
405
381
406 Returns
382 Returns
407 -------
383 -------
408
384
409 str|unicode : string indented by ntabs and nspaces.
385 str|unicode : string indented by ntabs and nspaces.
410
386
411 """
387 """
412 if instr is None:
388 if instr is None:
413 return
389 return
414 ind = '\t'*ntabs+' '*nspaces
390 ind = '\t'*ntabs+' '*nspaces
415 if flatten:
391 if flatten:
416 pat = re.compile(r'^\s*', re.MULTILINE)
392 pat = re.compile(r'^\s*', re.MULTILINE)
417 else:
393 else:
418 pat = re.compile(r'^', re.MULTILINE)
394 pat = re.compile(r'^', re.MULTILINE)
419 outstr = re.sub(pat, ind, instr)
395 outstr = re.sub(pat, ind, instr)
420 if outstr.endswith(os.linesep+ind):
396 if outstr.endswith(os.linesep+ind):
421 return outstr[:-len(ind)]
397 return outstr[:-len(ind)]
422 else:
398 else:
423 return outstr
399 return outstr
424
400
425 def native_line_ends(filename,backup=1):
401 def native_line_ends(filename,backup=1):
426 """Convert (in-place) a file to line-ends native to the current OS.
402 """Convert (in-place) a file to line-ends native to the current OS.
427
403
428 If the optional backup argument is given as false, no backup of the
404 If the optional backup argument is given as false, no backup of the
429 original file is left. """
405 original file is left. """
430
406
431 backup_suffixes = {'posix':'~','dos':'.bak','nt':'.bak','mac':'.bak'}
407 backup_suffixes = {'posix':'~','dos':'.bak','nt':'.bak','mac':'.bak'}
432
408
433 bak_filename = filename + backup_suffixes[os.name]
409 bak_filename = filename + backup_suffixes[os.name]
434
410
435 original = open(filename).read()
411 original = open(filename).read()
436 shutil.copy2(filename,bak_filename)
412 shutil.copy2(filename,bak_filename)
437 try:
413 try:
438 new = open(filename,'wb')
414 new = open(filename,'wb')
439 new.write(os.linesep.join(original.splitlines()))
415 new.write(os.linesep.join(original.splitlines()))
440 new.write(os.linesep) # ALWAYS put an eol at the end of the file
416 new.write(os.linesep) # ALWAYS put an eol at the end of the file
441 new.close()
417 new.close()
442 except:
418 except:
443 os.rename(bak_filename,filename)
419 os.rename(bak_filename,filename)
444 if not backup:
420 if not backup:
445 try:
421 try:
446 os.remove(bak_filename)
422 os.remove(bak_filename)
447 except:
423 except:
448 pass
424 pass
449
425
450
426
451 def list_strings(arg):
427 def list_strings(arg):
452 """Always return a list of strings, given a string or list of strings
428 """Always return a list of strings, given a string or list of strings
453 as input.
429 as input.
454
430
455 :Examples:
431 :Examples:
456
432
457 In [7]: list_strings('A single string')
433 In [7]: list_strings('A single string')
458 Out[7]: ['A single string']
434 Out[7]: ['A single string']
459
435
460 In [8]: list_strings(['A single string in a list'])
436 In [8]: list_strings(['A single string in a list'])
461 Out[8]: ['A single string in a list']
437 Out[8]: ['A single string in a list']
462
438
463 In [9]: list_strings(['A','list','of','strings'])
439 In [9]: list_strings(['A','list','of','strings'])
464 Out[9]: ['A', 'list', 'of', 'strings']
440 Out[9]: ['A', 'list', 'of', 'strings']
465 """
441 """
466
442
467 if isinstance(arg,basestring): return [arg]
443 if isinstance(arg,basestring): return [arg]
468 else: return arg
444 else: return arg
469
445
470
446
471 def marquee(txt='',width=78,mark='*'):
447 def marquee(txt='',width=78,mark='*'):
472 """Return the input string centered in a 'marquee'.
448 """Return the input string centered in a 'marquee'.
473
449
474 :Examples:
450 :Examples:
475
451
476 In [16]: marquee('A test',40)
452 In [16]: marquee('A test',40)
477 Out[16]: '**************** A test ****************'
453 Out[16]: '**************** A test ****************'
478
454
479 In [17]: marquee('A test',40,'-')
455 In [17]: marquee('A test',40,'-')
480 Out[17]: '---------------- A test ----------------'
456 Out[17]: '---------------- A test ----------------'
481
457
482 In [18]: marquee('A test',40,' ')
458 In [18]: marquee('A test',40,' ')
483 Out[18]: ' A test '
459 Out[18]: ' A test '
484
460
485 """
461 """
486 if not txt:
462 if not txt:
487 return (mark*width)[:width]
463 return (mark*width)[:width]
488 nmark = (width-len(txt)-2)//len(mark)//2
464 nmark = (width-len(txt)-2)//len(mark)//2
489 if nmark < 0: nmark =0
465 if nmark < 0: nmark =0
490 marks = mark*nmark
466 marks = mark*nmark
491 return '%s %s %s' % (marks,txt,marks)
467 return '%s %s %s' % (marks,txt,marks)
492
468
493
469
494 ini_spaces_re = re.compile(r'^(\s+)')
470 ini_spaces_re = re.compile(r'^(\s+)')
495
471
496 def num_ini_spaces(strng):
472 def num_ini_spaces(strng):
497 """Return the number of initial spaces in a string"""
473 """Return the number of initial spaces in a string"""
498
474
499 ini_spaces = ini_spaces_re.match(strng)
475 ini_spaces = ini_spaces_re.match(strng)
500 if ini_spaces:
476 if ini_spaces:
501 return ini_spaces.end()
477 return ini_spaces.end()
502 else:
478 else:
503 return 0
479 return 0
504
480
505
481
506 def format_screen(strng):
482 def format_screen(strng):
507 """Format a string for screen printing.
483 """Format a string for screen printing.
508
484
509 This removes some latex-type format codes."""
485 This removes some latex-type format codes."""
510 # Paragraph continue
486 # Paragraph continue
511 par_re = re.compile(r'\\$',re.MULTILINE)
487 par_re = re.compile(r'\\$',re.MULTILINE)
512 strng = par_re.sub('',strng)
488 strng = par_re.sub('',strng)
513 return strng
489 return strng
514
490
515 def dedent(text):
491 def dedent(text):
516 """Equivalent of textwrap.dedent that ignores unindented first line.
492 """Equivalent of textwrap.dedent that ignores unindented first line.
517
493
518 This means it will still dedent strings like:
494 This means it will still dedent strings like:
519 '''foo
495 '''foo
520 is a bar
496 is a bar
521 '''
497 '''
522
498
523 For use in wrap_paragraphs.
499 For use in wrap_paragraphs.
524 """
500 """
525
501
526 if text.startswith('\n'):
502 if text.startswith('\n'):
527 # text starts with blank line, don't ignore the first line
503 # text starts with blank line, don't ignore the first line
528 return textwrap.dedent(text)
504 return textwrap.dedent(text)
529
505
530 # split first line
506 # split first line
531 splits = text.split('\n',1)
507 splits = text.split('\n',1)
532 if len(splits) == 1:
508 if len(splits) == 1:
533 # only one line
509 # only one line
534 return textwrap.dedent(text)
510 return textwrap.dedent(text)
535
511
536 first, rest = splits
512 first, rest = splits
537 # dedent everything but the first line
513 # dedent everything but the first line
538 rest = textwrap.dedent(rest)
514 rest = textwrap.dedent(rest)
539 return '\n'.join([first, rest])
515 return '\n'.join([first, rest])
540
516
541 def wrap_paragraphs(text, ncols=80):
517 def wrap_paragraphs(text, ncols=80):
542 """Wrap multiple paragraphs to fit a specified width.
518 """Wrap multiple paragraphs to fit a specified width.
543
519
544 This is equivalent to textwrap.wrap, but with support for multiple
520 This is equivalent to textwrap.wrap, but with support for multiple
545 paragraphs, as separated by empty lines.
521 paragraphs, as separated by empty lines.
546
522
547 Returns
523 Returns
548 -------
524 -------
549
525
550 list of complete paragraphs, wrapped to fill `ncols` columns.
526 list of complete paragraphs, wrapped to fill `ncols` columns.
551 """
527 """
552 paragraph_re = re.compile(r'\n(\s*\n)+', re.MULTILINE)
528 paragraph_re = re.compile(r'\n(\s*\n)+', re.MULTILINE)
553 text = dedent(text).strip()
529 text = dedent(text).strip()
554 paragraphs = paragraph_re.split(text)[::2] # every other entry is space
530 paragraphs = paragraph_re.split(text)[::2] # every other entry is space
555 out_ps = []
531 out_ps = []
556 indent_re = re.compile(r'\n\s+', re.MULTILINE)
532 indent_re = re.compile(r'\n\s+', re.MULTILINE)
557 for p in paragraphs:
533 for p in paragraphs:
558 # presume indentation that survives dedent is meaningful formatting,
534 # presume indentation that survives dedent is meaningful formatting,
559 # so don't fill unless text is flush.
535 # so don't fill unless text is flush.
560 if indent_re.search(p) is None:
536 if indent_re.search(p) is None:
561 # wrap paragraph
537 # wrap paragraph
562 p = textwrap.fill(p, ncols)
538 p = textwrap.fill(p, ncols)
563 out_ps.append(p)
539 out_ps.append(p)
564 return out_ps
540 return out_ps
565
541
566
542
567 class EvalFormatter(Formatter):
543 class EvalFormatter(Formatter):
568 """A String Formatter that allows evaluation of simple expressions.
544 """A String Formatter that allows evaluation of simple expressions.
569
545
570 Note that this version interprets a : as specifying a format string (as per
546 Note that this version interprets a : as specifying a format string (as per
571 standard string formatting), so if slicing is required, you must explicitly
547 standard string formatting), so if slicing is required, you must explicitly
572 create a slice.
548 create a slice.
573
549
574 This is to be used in templating cases, such as the parallel batch
550 This is to be used in templating cases, such as the parallel batch
575 script templates, where simple arithmetic on arguments is useful.
551 script templates, where simple arithmetic on arguments is useful.
576
552
577 Examples
553 Examples
578 --------
554 --------
579
555
580 In [1]: f = EvalFormatter()
556 In [1]: f = EvalFormatter()
581 In [2]: f.format('{n//4}', n=8)
557 In [2]: f.format('{n//4}', n=8)
582 Out [2]: '2'
558 Out [2]: '2'
583
559
584 In [3]: f.format("{greeting[slice(2,4)]}", greeting="Hello")
560 In [3]: f.format("{greeting[slice(2,4)]}", greeting="Hello")
585 Out [3]: 'll'
561 Out [3]: 'll'
586 """
562 """
587 def get_field(self, name, args, kwargs):
563 def get_field(self, name, args, kwargs):
588 v = eval(name, kwargs)
564 v = eval(name, kwargs)
589 return v, name
565 return v, name
590
566
591 @skip_doctest_py3
567 @skip_doctest_py3
592 class FullEvalFormatter(Formatter):
568 class FullEvalFormatter(Formatter):
593 """A String Formatter that allows evaluation of simple expressions.
569 """A String Formatter that allows evaluation of simple expressions.
594
570
595 Any time a format key is not found in the kwargs,
571 Any time a format key is not found in the kwargs,
596 it will be tried as an expression in the kwargs namespace.
572 it will be tried as an expression in the kwargs namespace.
597
573
598 Note that this version allows slicing using [1:2], so you cannot specify
574 Note that this version allows slicing using [1:2], so you cannot specify
599 a format string. Use :class:`EvalFormatter` to permit format strings.
575 a format string. Use :class:`EvalFormatter` to permit format strings.
600
576
601 Examples
577 Examples
602 --------
578 --------
603
579
604 In [1]: f = FullEvalFormatter()
580 In [1]: f = FullEvalFormatter()
605 In [2]: f.format('{n//4}', n=8)
581 In [2]: f.format('{n//4}', n=8)
606 Out[2]: u'2'
582 Out[2]: u'2'
607
583
608 In [3]: f.format('{list(range(5))[2:4]}')
584 In [3]: f.format('{list(range(5))[2:4]}')
609 Out[3]: u'[2, 3]'
585 Out[3]: u'[2, 3]'
610
586
611 In [4]: f.format('{3*2}')
587 In [4]: f.format('{3*2}')
612 Out[4]: u'6'
588 Out[4]: u'6'
613 """
589 """
614 # copied from Formatter._vformat with minor changes to allow eval
590 # copied from Formatter._vformat with minor changes to allow eval
615 # and replace the format_spec code with slicing
591 # and replace the format_spec code with slicing
616 def _vformat(self, format_string, args, kwargs, used_args, recursion_depth):
592 def _vformat(self, format_string, args, kwargs, used_args, recursion_depth):
617 if recursion_depth < 0:
593 if recursion_depth < 0:
618 raise ValueError('Max string recursion exceeded')
594 raise ValueError('Max string recursion exceeded')
619 result = []
595 result = []
620 for literal_text, field_name, format_spec, conversion in \
596 for literal_text, field_name, format_spec, conversion in \
621 self.parse(format_string):
597 self.parse(format_string):
622
598
623 # output the literal text
599 # output the literal text
624 if literal_text:
600 if literal_text:
625 result.append(literal_text)
601 result.append(literal_text)
626
602
627 # if there's a field, output it
603 # if there's a field, output it
628 if field_name is not None:
604 if field_name is not None:
629 # this is some markup, find the object and do
605 # this is some markup, find the object and do
630 # the formatting
606 # the formatting
631
607
632 if format_spec:
608 if format_spec:
633 # override format spec, to allow slicing:
609 # override format spec, to allow slicing:
634 field_name = ':'.join([field_name, format_spec])
610 field_name = ':'.join([field_name, format_spec])
635
611
636 # eval the contents of the field for the object
612 # eval the contents of the field for the object
637 # to be formatted
613 # to be formatted
638 obj = eval(field_name, kwargs)
614 obj = eval(field_name, kwargs)
639
615
640 # do any conversion on the resulting object
616 # do any conversion on the resulting object
641 obj = self.convert_field(obj, conversion)
617 obj = self.convert_field(obj, conversion)
642
618
643 # format the object and append to the result
619 # format the object and append to the result
644 result.append(self.format_field(obj, ''))
620 result.append(self.format_field(obj, ''))
645
621
646 return u''.join(py3compat.cast_unicode(s) for s in result)
622 return u''.join(py3compat.cast_unicode(s) for s in result)
647
623
648 @skip_doctest_py3
624 @skip_doctest_py3
649 class DollarFormatter(FullEvalFormatter):
625 class DollarFormatter(FullEvalFormatter):
650 """Formatter allowing Itpl style $foo replacement, for names and attribute
626 """Formatter allowing Itpl style $foo replacement, for names and attribute
651 access only. Standard {foo} replacement also works, and allows full
627 access only. Standard {foo} replacement also works, and allows full
652 evaluation of its arguments.
628 evaluation of its arguments.
653
629
654 Examples
630 Examples
655 --------
631 --------
656 In [1]: f = DollarFormatter()
632 In [1]: f = DollarFormatter()
657 In [2]: f.format('{n//4}', n=8)
633 In [2]: f.format('{n//4}', n=8)
658 Out[2]: u'2'
634 Out[2]: u'2'
659
635
660 In [3]: f.format('23 * 76 is $result', result=23*76)
636 In [3]: f.format('23 * 76 is $result', result=23*76)
661 Out[3]: u'23 * 76 is 1748'
637 Out[3]: u'23 * 76 is 1748'
662
638
663 In [4]: f.format('$a or {b}', a=1, b=2)
639 In [4]: f.format('$a or {b}', a=1, b=2)
664 Out[4]: u'1 or 2'
640 Out[4]: u'1 or 2'
665 """
641 """
666 _dollar_pattern = re.compile("(.*?)\$(\$?[\w\.]+)")
642 _dollar_pattern = re.compile("(.*?)\$(\$?[\w\.]+)")
667 def parse(self, fmt_string):
643 def parse(self, fmt_string):
668 for literal_txt, field_name, format_spec, conversion \
644 for literal_txt, field_name, format_spec, conversion \
669 in Formatter.parse(self, fmt_string):
645 in Formatter.parse(self, fmt_string):
670
646
671 # Find $foo patterns in the literal text.
647 # Find $foo patterns in the literal text.
672 continue_from = 0
648 continue_from = 0
673 txt = ""
649 txt = ""
674 for m in self._dollar_pattern.finditer(literal_txt):
650 for m in self._dollar_pattern.finditer(literal_txt):
675 new_txt, new_field = m.group(1,2)
651 new_txt, new_field = m.group(1,2)
676 # $$foo --> $foo
652 # $$foo --> $foo
677 if new_field.startswith("$"):
653 if new_field.startswith("$"):
678 txt += new_txt + new_field
654 txt += new_txt + new_field
679 else:
655 else:
680 yield (txt + new_txt, new_field, "", None)
656 yield (txt + new_txt, new_field, "", None)
681 txt = ""
657 txt = ""
682 continue_from = m.end()
658 continue_from = m.end()
683
659
684 # Re-yield the {foo} style pattern
660 # Re-yield the {foo} style pattern
685 yield (txt + literal_txt[continue_from:], field_name, format_spec, conversion)
661 yield (txt + literal_txt[continue_from:], field_name, format_spec, conversion)
686
662
687
663
688 def columnize(items, separator=' ', displaywidth=80):
664 def columnize(items, separator=' ', displaywidth=80):
689 """ Transform a list of strings into a single string with columns.
665 """ Transform a list of strings into a single string with columns.
690
666
691 Parameters
667 Parameters
692 ----------
668 ----------
693 items : sequence of strings
669 items : sequence of strings
694 The strings to process.
670 The strings to process.
695
671
696 separator : str, optional [default is two spaces]
672 separator : str, optional [default is two spaces]
697 The string that separates columns.
673 The string that separates columns.
698
674
699 displaywidth : int, optional [default is 80]
675 displaywidth : int, optional [default is 80]
700 Width of the display in number of characters.
676 Width of the display in number of characters.
701
677
702 Returns
678 Returns
703 -------
679 -------
704 The formatted string.
680 The formatted string.
705 """
681 """
706 # Note: this code is adapted from columnize 0.3.2.
682 # Note: this code is adapted from columnize 0.3.2.
707 # See http://code.google.com/p/pycolumnize/
683 # See http://code.google.com/p/pycolumnize/
708
684
709 # Some degenerate cases.
685 # Some degenerate cases.
710 size = len(items)
686 size = len(items)
711 if size == 0:
687 if size == 0:
712 return '\n'
688 return '\n'
713 elif size == 1:
689 elif size == 1:
714 return '%s\n' % items[0]
690 return '%s\n' % items[0]
715
691
716 # Special case: if any item is longer than the maximum width, there's no
692 # Special case: if any item is longer than the maximum width, there's no
717 # point in triggering the logic below...
693 # point in triggering the logic below...
718 item_len = map(len, items) # save these, we can reuse them below
694 item_len = map(len, items) # save these, we can reuse them below
719 longest = max(item_len)
695 longest = max(item_len)
720 if longest >= displaywidth:
696 if longest >= displaywidth:
721 return '\n'.join(items+[''])
697 return '\n'.join(items+[''])
722
698
723 # Try every row count from 1 upwards
699 # Try every row count from 1 upwards
724 array_index = lambda nrows, row, col: nrows*col + row
700 array_index = lambda nrows, row, col: nrows*col + row
725 for nrows in range(1, size):
701 for nrows in range(1, size):
726 ncols = (size + nrows - 1) // nrows
702 ncols = (size + nrows - 1) // nrows
727 colwidths = []
703 colwidths = []
728 totwidth = -len(separator)
704 totwidth = -len(separator)
729 for col in range(ncols):
705 for col in range(ncols):
730 # Get max column width for this column
706 # Get max column width for this column
731 colwidth = 0
707 colwidth = 0
732 for row in range(nrows):
708 for row in range(nrows):
733 i = array_index(nrows, row, col)
709 i = array_index(nrows, row, col)
734 if i >= size: break
710 if i >= size: break
735 x, len_x = items[i], item_len[i]
711 x, len_x = items[i], item_len[i]
736 colwidth = max(colwidth, len_x)
712 colwidth = max(colwidth, len_x)
737 colwidths.append(colwidth)
713 colwidths.append(colwidth)
738 totwidth += colwidth + len(separator)
714 totwidth += colwidth + len(separator)
739 if totwidth > displaywidth:
715 if totwidth > displaywidth:
740 break
716 break
741 if totwidth <= displaywidth:
717 if totwidth <= displaywidth:
742 break
718 break
743
719
744 # The smallest number of rows computed and the max widths for each
720 # The smallest number of rows computed and the max widths for each
745 # column has been obtained. Now we just have to format each of the rows.
721 # column has been obtained. Now we just have to format each of the rows.
746 string = ''
722 string = ''
747 for row in range(nrows):
723 for row in range(nrows):
748 texts = []
724 texts = []
749 for col in range(ncols):
725 for col in range(ncols):
750 i = row + nrows*col
726 i = row + nrows*col
751 if i >= size:
727 if i >= size:
752 texts.append('')
728 texts.append('')
753 else:
729 else:
754 texts.append(items[i])
730 texts.append(items[i])
755 while texts and not texts[-1]:
731 while texts and not texts[-1]:
756 del texts[-1]
732 del texts[-1]
757 for col in range(len(texts)):
733 for col in range(len(texts)):
758 texts[col] = texts[col].ljust(colwidths[col])
734 texts[col] = texts[col].ljust(colwidths[col])
759 string += '%s\n' % separator.join(texts)
735 string += '%s\n' % separator.join(texts)
760 return string
736 return string
@@ -1,91 +1,91 b''
1 import sys
1 import sys
2 import time
2 import time
3 from io import StringIO
3 from io import StringIO
4
4
5 from session import extract_header, Message
5 from session import extract_header, Message
6
6
7 from IPython.utils import io, text
7 from IPython.utils import io, text, py3compat
8
8
9 #-----------------------------------------------------------------------------
9 #-----------------------------------------------------------------------------
10 # Globals
10 # Globals
11 #-----------------------------------------------------------------------------
11 #-----------------------------------------------------------------------------
12
12
13 #-----------------------------------------------------------------------------
13 #-----------------------------------------------------------------------------
14 # Stream classes
14 # Stream classes
15 #-----------------------------------------------------------------------------
15 #-----------------------------------------------------------------------------
16
16
17 class OutStream(object):
17 class OutStream(object):
18 """A file like object that publishes the stream to a 0MQ PUB socket."""
18 """A file like object that publishes the stream to a 0MQ PUB socket."""
19
19
20 # The time interval between automatic flushes, in seconds.
20 # The time interval between automatic flushes, in seconds.
21 flush_interval = 0.05
21 flush_interval = 0.05
22 topic=None
22 topic=None
23
23
24 def __init__(self, session, pub_socket, name):
24 def __init__(self, session, pub_socket, name):
25 self.session = session
25 self.session = session
26 self.pub_socket = pub_socket
26 self.pub_socket = pub_socket
27 self.name = name
27 self.name = name
28 self.parent_header = {}
28 self.parent_header = {}
29 self._new_buffer()
29 self._new_buffer()
30
30
31 def set_parent(self, parent):
31 def set_parent(self, parent):
32 self.parent_header = extract_header(parent)
32 self.parent_header = extract_header(parent)
33
33
34 def close(self):
34 def close(self):
35 self.pub_socket = None
35 self.pub_socket = None
36
36
37 def flush(self):
37 def flush(self):
38 #io.rprint('>>>flushing output buffer: %s<<<' % self.name) # dbg
38 #io.rprint('>>>flushing output buffer: %s<<<' % self.name) # dbg
39 if self.pub_socket is None:
39 if self.pub_socket is None:
40 raise ValueError(u'I/O operation on closed file')
40 raise ValueError(u'I/O operation on closed file')
41 else:
41 else:
42 data = self._buffer.getvalue()
42 data = self._buffer.getvalue()
43 if data:
43 if data:
44 content = {u'name':self.name, u'data':data}
44 content = {u'name':self.name, u'data':data}
45 msg = self.session.send(self.pub_socket, u'stream', content=content,
45 msg = self.session.send(self.pub_socket, u'stream', content=content,
46 parent=self.parent_header, ident=self.topic)
46 parent=self.parent_header, ident=self.topic)
47
47
48 if hasattr(self.pub_socket, 'flush'):
48 if hasattr(self.pub_socket, 'flush'):
49 # socket itself has flush (presumably ZMQStream)
49 # socket itself has flush (presumably ZMQStream)
50 self.pub_socket.flush()
50 self.pub_socket.flush()
51 self._buffer.close()
51 self._buffer.close()
52 self._new_buffer()
52 self._new_buffer()
53
53
54 def isatty(self):
54 def isatty(self):
55 return False
55 return False
56
56
57 def next(self):
57 def next(self):
58 raise IOError('Read not supported on a write only stream.')
58 raise IOError('Read not supported on a write only stream.')
59
59
60 def read(self, size=-1):
60 def read(self, size=-1):
61 raise IOError('Read not supported on a write only stream.')
61 raise IOError('Read not supported on a write only stream.')
62
62
63 def readline(self, size=-1):
63 def readline(self, size=-1):
64 raise IOError('Read not supported on a write only stream.')
64 raise IOError('Read not supported on a write only stream.')
65
65
66 def write(self, string):
66 def write(self, string):
67 if self.pub_socket is None:
67 if self.pub_socket is None:
68 raise ValueError('I/O operation on closed file')
68 raise ValueError('I/O operation on closed file')
69 else:
69 else:
70 # Make sure that we're handling unicode
70 # Make sure that we're handling unicode
71 if not isinstance(string, unicode):
71 if not isinstance(string, unicode):
72 enc = text.getdefaultencoding()
72 enc = py3compat.getdefaultencoding()
73 string = string.decode(enc, 'replace')
73 string = string.decode(enc, 'replace')
74
74
75 self._buffer.write(string)
75 self._buffer.write(string)
76 current_time = time.time()
76 current_time = time.time()
77 if self._start <= 0:
77 if self._start <= 0:
78 self._start = current_time
78 self._start = current_time
79 elif current_time - self._start > self.flush_interval:
79 elif current_time - self._start > self.flush_interval:
80 self.flush()
80 self.flush()
81
81
82 def writelines(self, sequence):
82 def writelines(self, sequence):
83 if self.pub_socket is None:
83 if self.pub_socket is None:
84 raise ValueError('I/O operation on closed file')
84 raise ValueError('I/O operation on closed file')
85 else:
85 else:
86 for string in sequence:
86 for string in sequence:
87 self.write(string)
87 self.write(string)
88
88
89 def _new_buffer(self):
89 def _new_buffer(self):
90 self._buffer = StringIO()
90 self._buffer = StringIO()
91 self._start = -1
91 self._start = -1
General Comments 0
You need to be logged in to leave comments. Login now