Backport PR #12208: support for unicode identifiers

IPython/core/completerlib.py

@@ -1,354 +1,354 b''
 # encoding: utf-8
 """Implementations for various useful completers.
 
 These are all loaded by default by IPython.
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011 The IPython Development Team.
 #
 #  Distributed under the terms of the BSD License.
 #
 #  The full license is in the file COPYING.txt, distributed with this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 # Stdlib imports
 import glob
 import inspect
 import os
 import re
 import sys
 from importlib import import_module
 from importlib.machinery import all_suffixes
 
 
 # Third-party imports
 from time import time
 from zipimport import zipimporter
 
 # Our own imports
 from .completer import expand_user, compress_user
 from .error import TryNext
 from ..utils._process_common import arg_split
 
 # FIXME: this should be pulled in with the right call via the component system
 from IPython import get_ipython
 
 from typing import List
 
 #-----------------------------------------------------------------------------
 # Globals and constants
 #-----------------------------------------------------------------------------
 _suffixes = all_suffixes()
 
 # Time in seconds after which the rootmodules will be stored permanently in the
 # ipython ip.db database (kept in the user's .ipython dir).
 TIMEOUT_STORAGE = 2
 
 # Time in seconds after which we give up
 TIMEOUT_GIVEUP = 20
 
 # Regular expression for the python import statement
-import_re = re.compile(r'(?P<name>[a-zA-Z_][a-zA-Z0-9_]*?)'
+import_re = re.compile(r'(?P<name>[^\W\d]\w*?)'
                        r'(?P<package>[/\\]__init__)?'
                        r'(?P<suffix>%s)$' %
                        r'|'.join(re.escape(s) for s in _suffixes))
 
 # RE for the ipython %run command (python + ipython scripts)
 magic_run_re = re.compile(r'.*(\.ipy|\.ipynb|\.py[w]?)$')
 
 #-----------------------------------------------------------------------------
 # Local utilities
 #-----------------------------------------------------------------------------
 
 def module_list(path):
     """
     Return the list containing the names of the modules available in the given
     folder.
     """
     # sys.path has the cwd as an empty string, but isdir/listdir need it as '.'
     if path == '':
         path = '.'
 
     # A few local constants to be used in loops below
     pjoin = os.path.join
 
     if os.path.isdir(path):
         # Build a list of all files in the directory and all files
         # in its subdirectories. For performance reasons, do not
         # recurse more than one level into subdirectories.
         files = []
         for root, dirs, nondirs in os.walk(path, followlinks=True):
             subdir = root[len(path)+1:]
             if subdir:
                 files.extend(pjoin(subdir, f) for f in nondirs)
                 dirs[:] = [] # Do not recurse into additional subdirectories.
             else:
                 files.extend(nondirs)
 
     else:
         try:
             files = list(zipimporter(path)._files.keys())
         except:
             files = []
 
     # Build a list of modules which match the import_re regex.
     modules = []
     for f in files:
         m = import_re.match(f)
         if m:
             modules.append(m.group('name'))
     return list(set(modules))
 
 
 def get_root_modules():
     """
     Returns a list containing the names of all the modules available in the
     folders of the pythonpath.
 
     ip.db['rootmodules_cache'] maps sys.path entries to list of modules.
     """
     ip = get_ipython()
     if ip is None:
         # No global shell instance to store cached list of modules.
         # Don't try to scan for modules every time.
         return list(sys.builtin_module_names)
 
     rootmodules_cache = ip.db.get('rootmodules_cache', {})
     rootmodules = list(sys.builtin_module_names)
     start_time = time()
     store = False
     for path in sys.path:
         try:
             modules = rootmodules_cache[path]
         except KeyError:
             modules = module_list(path)
             try:
                 modules.remove('__init__')
             except ValueError:
                 pass
             if path not in ('', '.'): # cwd modules should not be cached
                 rootmodules_cache[path] = modules
             if time() - start_time > TIMEOUT_STORAGE and not store:
                 store = True
                 print("\nCaching the list of root modules, please wait!")
                 print("(This will only be done once - type '%rehashx' to "
                       "reset cache!)\n")
                 sys.stdout.flush()
             if time() - start_time > TIMEOUT_GIVEUP:
                 print("This is taking too long, we give up.\n")
                 return []
         rootmodules.extend(modules)
     if store:
         ip.db['rootmodules_cache'] = rootmodules_cache
     rootmodules = list(set(rootmodules))
     return rootmodules
 
 
 def is_importable(module, attr, only_modules):
     if only_modules:
         return inspect.ismodule(getattr(module, attr))
     else:
         return not(attr[:2] == '__' and attr[-2:] == '__')
 
 
 def try_import(mod: str, only_modules=False) -> List[str]:
     """
     Try to import given module and return list of potential completions.
     """
     mod = mod.rstrip('.')
     try:
         m = import_module(mod)
     except:
         return []
 
     m_is_init = '__init__' in (getattr(m, '__file__', '') or '')
 
     completions = []
     if (not hasattr(m, '__file__')) or (not only_modules) or m_is_init:
         completions.extend( [attr for attr in dir(m) if
                              is_importable(m, attr, only_modules)])
 
     completions.extend(getattr(m, '__all__', []))
     if m_is_init:
         completions.extend(module_list(os.path.dirname(m.__file__)))
     completions_set = {c for c in completions if isinstance(c, str)}
     completions_set.discard('__init__')
     return list(completions_set)
 
 
 #-----------------------------------------------------------------------------
 # Completion-related functions.
 #-----------------------------------------------------------------------------
 
 def quick_completer(cmd, completions):
     r""" Easily create a trivial completer for a command.
 
     Takes either a list of completions, or all completions in string (that will
     be split on whitespace).
 
     Example::
 
         [d:\ipython]|1> import ipy_completers
         [d:\ipython]|2> ipy_completers.quick_completer('foo', ['bar','baz'])
         [d:\ipython]|3> foo b<TAB>
         bar baz
         [d:\ipython]|3> foo ba
     """
 
     if isinstance(completions, str):
         completions = completions.split()
 
     def do_complete(self, event):
         return completions
 
     get_ipython().set_hook('complete_command',do_complete, str_key = cmd)
 
 def module_completion(line):
     """
     Returns a list containing the completion possibilities for an import line.
 
     The line looks like this :
     'import xml.d'
     'from xml.dom import'
     """
 
     words = line.split(' ')
     nwords = len(words)
 
     # from whatever <tab> -> 'import '
     if nwords == 3 and words[0] == 'from':
         return ['import ']
 
     # 'from xy<tab>' or 'import xy<tab>'
     if nwords < 3 and (words[0] in {'%aimport', 'import', 'from'}) :
         if nwords == 1:
             return get_root_modules()
         mod = words[1].split('.')
         if len(mod) < 2:
             return get_root_modules()
         completion_list = try_import('.'.join(mod[:-1]), True)
         return ['.'.join(mod[:-1] + [el]) for el in completion_list]
 
     # 'from xyz import abc<tab>'
     if nwords >= 3 and words[0] == 'from':
         mod = words[1]
         return try_import(mod)
 
 #-----------------------------------------------------------------------------
 # Completers
 #-----------------------------------------------------------------------------
 # These all have the func(self, event) signature to be used as custom
 # completers
 
 def module_completer(self,event):
     """Give completions after user has typed 'import ...' or 'from ...'"""
 
     # This works in all versions of python.  While 2.5 has
     # pkgutil.walk_packages(), that particular routine is fairly dangerous,
     # since it imports *EVERYTHING* on sys.path.  That is: a) very slow b) full
     # of possibly problematic side effects.
     # This search the folders in the sys.path for available modules.
 
     return module_completion(event.line)
 
 # FIXME: there's a lot of logic common to the run, cd and builtin file
 # completers, that is currently reimplemented in each.
 
 def magic_run_completer(self, event):
     """Complete files that end in .py or .ipy or .ipynb for the %run command.
     """
     comps = arg_split(event.line, strict=False)
     # relpath should be the current token that we need to complete.
     if (len(comps) > 1) and (not event.line.endswith(' ')):
         relpath = comps[-1].strip("'\"")
     else:
         relpath = ''
 
     #print("\nev=", event)  # dbg
     #print("rp=", relpath)  # dbg
     #print('comps=', comps)  # dbg
 
     lglob = glob.glob
     isdir = os.path.isdir
     relpath, tilde_expand, tilde_val = expand_user(relpath)
 
     # Find if the user has already typed the first filename, after which we
     # should complete on all files, since after the first one other files may
     # be arguments to the input script.
 
     if any(magic_run_re.match(c) for c in comps):
         matches =  [f.replace('\\','/') + ('/' if isdir(f) else '')
                             for f in lglob(relpath+'*')]
     else:
         dirs = [f.replace('\\','/') + "/" for f in lglob(relpath+'*') if isdir(f)]
         pys =  [f.replace('\\','/')
                 for f in lglob(relpath+'*.py') + lglob(relpath+'*.ipy') +
                 lglob(relpath+'*.ipynb') + lglob(relpath + '*.pyw')]
 
         matches = dirs + pys
 
     #print('run comp:', dirs+pys) # dbg
     return [compress_user(p, tilde_expand, tilde_val) for p in matches]
 
 
 def cd_completer(self, event):
     """Completer function for cd, which only returns directories."""
     ip = get_ipython()
     relpath = event.symbol
 
     #print(event) # dbg
     if event.line.endswith('-b') or ' -b ' in event.line:
         # return only bookmark completions
         bkms = self.db.get('bookmarks', None)
         if bkms:
             return bkms.keys()
         else:
             return []
 
     if event.symbol == '-':
         width_dh = str(len(str(len(ip.user_ns['_dh']) + 1)))
         # jump in directory history by number
         fmt = '-%0' + width_dh +'d [%s]'
         ents = [ fmt % (i,s) for i,s in enumerate(ip.user_ns['_dh'])]
         if len(ents) > 1:
             return ents
         return []
 
     if event.symbol.startswith('--'):
         return ["--" + os.path.basename(d) for d in ip.user_ns['_dh']]
 
     # Expand ~ in path and normalize directory separators.
     relpath, tilde_expand, tilde_val = expand_user(relpath)
     relpath = relpath.replace('\\','/')
 
     found = []
     for d in [f.replace('\\','/') + '/' for f in glob.glob(relpath+'*')
               if os.path.isdir(f)]:
         if ' ' in d:
             # we don't want to deal with any of that, complex code
             # for this is elsewhere
             raise TryNext
 
         found.append(d)
 
     if not found:
         if os.path.isdir(relpath):
             return [compress_user(relpath, tilde_expand, tilde_val)]
 
         # if no completions so far, try bookmarks
         bks = self.db.get('bookmarks',{})
         bkmatches = [s for s in bks if s.startswith(event.symbol)]
         if bkmatches:
             return bkmatches
 
         raise TryNext
 
     return [compress_user(p, tilde_expand, tilde_val) for p in found]
 
 def reset_completer(self, event):
     "A completer for %reset magic"
     return '-f -s in out array dhist'.split()

IPython/core/inputtransformer.py

@@ -1,536 +1,536 b''
 """DEPRECATED: Input transformer classes to support IPython special syntax.
 
 This module was deprecated in IPython 7.0, in favour of inputtransformer2.
 
 This includes the machinery to recognise and transform ``%magic`` commands,
 ``!system`` commands, ``help?`` querying, prompt stripping, and so forth.
 """
 import abc
 import functools
 import re
 import tokenize
 from tokenize import generate_tokens, untokenize, TokenError
 from io import StringIO
 
 from IPython.core.splitinput import LineInfo
 
 #-----------------------------------------------------------------------------
 # Globals
 #-----------------------------------------------------------------------------
 
 # The escape sequences that define the syntax transformations IPython will
 # apply to user input.  These can NOT be just changed here: many regular
 # expressions and other parts of the code may use their hardcoded values, and
 # for all intents and purposes they constitute the 'IPython syntax', so they
 # should be considered fixed.
 
 ESC_SHELL  = '!'     # Send line to underlying system shell
 ESC_SH_CAP = '!!'    # Send line to system shell and capture output
 ESC_HELP   = '?'     # Find information about object
 ESC_HELP2  = '??'    # Find extra-detailed information about object
 ESC_MAGIC  = '%'     # Call magic function
 ESC_MAGIC2 = '%%'    # Call cell-magic function
 ESC_QUOTE  = ','     # Split args on whitespace, quote each as string and call
 ESC_QUOTE2 = ';'     # Quote all args as a single string, call
 ESC_PAREN  = '/'     # Call first argument with rest of line as arguments
 
 ESC_SEQUENCES = [ESC_SHELL, ESC_SH_CAP, ESC_HELP ,\
                  ESC_HELP2, ESC_MAGIC, ESC_MAGIC2,\
                  ESC_QUOTE, ESC_QUOTE2, ESC_PAREN ]
 
 
 class InputTransformer(metaclass=abc.ABCMeta):
     """Abstract base class for line-based input transformers."""
     
     @abc.abstractmethod
     def push(self, line):
         """Send a line of input to the transformer, returning the transformed
         input or None if the transformer is waiting for more input.
         
         Must be overridden by subclasses.
 
         Implementations may raise ``SyntaxError`` if the input is invalid. No
         other exceptions may be raised.
         """
         pass
     
     @abc.abstractmethod
     def reset(self):
         """Return, transformed any lines that the transformer has accumulated,
         and reset its internal state.
         
         Must be overridden by subclasses.
         """
         pass
     
     @classmethod
     def wrap(cls, func):
         """Can be used by subclasses as a decorator, to return a factory that
         will allow instantiation with the decorated object.
         """
         @functools.wraps(func)
         def transformer_factory(**kwargs):
             return cls(func, **kwargs)
         
         return transformer_factory
 
 class StatelessInputTransformer(InputTransformer):
     """Wrapper for a stateless input transformer implemented as a function."""
     def __init__(self, func):
         self.func = func
     
     def __repr__(self):
         return "StatelessInputTransformer(func={0!r})".format(self.func)
     
     def push(self, line):
         """Send a line of input to the transformer, returning the
         transformed input."""
         return self.func(line)
     
     def reset(self):
         """No-op - exists for compatibility."""
         pass
 
 class CoroutineInputTransformer(InputTransformer):
     """Wrapper for an input transformer implemented as a coroutine."""
     def __init__(self, coro, **kwargs):
         # Prime it
         self.coro = coro(**kwargs)
         next(self.coro)
     
     def __repr__(self):
         return "CoroutineInputTransformer(coro={0!r})".format(self.coro)
     
     def push(self, line):
         """Send a line of input to the transformer, returning the
         transformed input or None if the transformer is waiting for more
         input.
         """
         return self.coro.send(line)
     
     def reset(self):
         """Return, transformed any lines that the transformer has
         accumulated, and reset its internal state.
         """
         return self.coro.send(None)
 
 class TokenInputTransformer(InputTransformer):
     """Wrapper for a token-based input transformer.
     
     func should accept a list of tokens (5-tuples, see tokenize docs), and
     return an iterable which can be passed to tokenize.untokenize().
     """
     def __init__(self, func):
         self.func = func
         self.buf = []
         self.reset_tokenizer()
 
     def reset_tokenizer(self):
         it = iter(self.buf)
         self.tokenizer = generate_tokens(it.__next__)
 
     def push(self, line):
         self.buf.append(line + '\n')
         if all(l.isspace() for l in self.buf):
             return self.reset()
 
         tokens = []
         stop_at_NL = False
         try:
             for intok in self.tokenizer:
                 tokens.append(intok)
                 t = intok[0]
                 if t == tokenize.NEWLINE or (stop_at_NL and t == tokenize.NL):
                     # Stop before we try to pull a line we don't have yet
                     break
                 elif t == tokenize.ERRORTOKEN:
                     stop_at_NL = True
         except TokenError:
             # Multi-line statement - stop and try again with the next line
             self.reset_tokenizer()
             return None
         
         return self.output(tokens)
     
     def output(self, tokens):
         self.buf.clear()
         self.reset_tokenizer()
         return untokenize(self.func(tokens)).rstrip('\n')
     
     def reset(self):
         l = ''.join(self.buf)
         self.buf.clear()
         self.reset_tokenizer()
         if l:
             return l.rstrip('\n')
 
 class assemble_python_lines(TokenInputTransformer):
     def __init__(self):
         super(assemble_python_lines, self).__init__(None)
     
     def output(self, tokens):
         return self.reset()
 
 @CoroutineInputTransformer.wrap
 def assemble_logical_lines():
     r"""Join lines following explicit line continuations (\)"""
     line = ''
     while True:
         line = (yield line)
         if not line or line.isspace():
             continue
         
         parts = []
         while line is not None:
             if line.endswith('\\') and (not has_comment(line)):
                 parts.append(line[:-1])
                 line = (yield None) # Get another line
             else:
                 parts.append(line)
                 break
         
         # Output
         line = ''.join(parts)
 
 # Utilities
 def _make_help_call(target, esc, lspace, next_input=None):
     """Prepares a pinfo(2)/psearch call from a target name and the escape
     (i.e. ? or ??)"""
     method  = 'pinfo2' if esc == '??' \
                 else 'psearch' if '*' in target \
                 else 'pinfo'
     arg = " ".join([method, target])
     #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args)
     t_magic_name, _, t_magic_arg_s = arg.partition(' ')
     t_magic_name = t_magic_name.lstrip(ESC_MAGIC)
     if next_input is None:
         return '%sget_ipython().run_line_magic(%r, %r)' % (lspace, t_magic_name, t_magic_arg_s)
     else:
         return '%sget_ipython().set_next_input(%r);get_ipython().run_line_magic(%r, %r)' % \
            (lspace, next_input, t_magic_name, t_magic_arg_s)
     
 # These define the transformations for the different escape characters.
 def _tr_system(line_info):
     "Translate lines escaped with: !"
     cmd = line_info.line.lstrip().lstrip(ESC_SHELL)
     return '%sget_ipython().system(%r)' % (line_info.pre, cmd)
 
 def _tr_system2(line_info):
     "Translate lines escaped with: !!"
     cmd = line_info.line.lstrip()[2:]
     return '%sget_ipython().getoutput(%r)' % (line_info.pre, cmd)
 
 def _tr_help(line_info):
     "Translate lines escaped with: ?/??"
     # A naked help line should just fire the intro help screen
     if not line_info.line[1:]:
         return 'get_ipython().show_usage()'
 
     return _make_help_call(line_info.ifun, line_info.esc, line_info.pre)
 
 def _tr_magic(line_info):
     "Translate lines escaped with: %"
     tpl = '%sget_ipython().run_line_magic(%r, %r)'
     if line_info.line.startswith(ESC_MAGIC2):
         return line_info.line
     cmd = ' '.join([line_info.ifun, line_info.the_rest]).strip()
     #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args)
     t_magic_name, _, t_magic_arg_s = cmd.partition(' ')
     t_magic_name = t_magic_name.lstrip(ESC_MAGIC)
     return tpl % (line_info.pre, t_magic_name, t_magic_arg_s)
 
 def _tr_quote(line_info):
     "Translate lines escaped with: ,"
     return '%s%s("%s")' % (line_info.pre, line_info.ifun,
                          '", "'.join(line_info.the_rest.split()) )
 
 def _tr_quote2(line_info):
     "Translate lines escaped with: ;"
     return '%s%s("%s")' % (line_info.pre, line_info.ifun,
                            line_info.the_rest)
 
 def _tr_paren(line_info):
     "Translate lines escaped with: /"
     return '%s%s(%s)' % (line_info.pre, line_info.ifun,
                          ", ".join(line_info.the_rest.split()))
 
 tr = { ESC_SHELL  : _tr_system,
        ESC_SH_CAP : _tr_system2,
        ESC_HELP   : _tr_help,
        ESC_HELP2  : _tr_help,
        ESC_MAGIC  : _tr_magic,
        ESC_QUOTE  : _tr_quote,
        ESC_QUOTE2 : _tr_quote2,
        ESC_PAREN  : _tr_paren }
 
 @StatelessInputTransformer.wrap
 def escaped_commands(line):
     """Transform escaped commands - %magic, !system, ?help + various autocalls.
     """
     if not line or line.isspace():
         return line
     lineinf = LineInfo(line)
     if lineinf.esc not in tr:
         return line
     
     return tr[lineinf.esc](lineinf)
 
 _initial_space_re = re.compile(r'\s*')
 
 _help_end_re = re.compile(r"""(%{0,2}
-                              [a-zA-Z_*][\w*]*        # Variable name
-                              (\.[a-zA-Z_*][\w*]*)*   # .etc.etc
+                              (?!\d)[\w*]+            # Variable name
+                              (\.(?!\d)[\w*]+)*       # .etc.etc
                               )
                               (\?\??)$                # ? or ??
                               """,
                               re.VERBOSE)
 
 # Extra pseudotokens for multiline strings and data structures
 _MULTILINE_STRING = object()
 _MULTILINE_STRUCTURE = object()
 
 def _line_tokens(line):
     """Helper for has_comment and ends_in_comment_or_string."""
     readline = StringIO(line).readline
     toktypes = set()
     try:
         for t in generate_tokens(readline):
             toktypes.add(t[0])
     except TokenError as e:
         # There are only two cases where a TokenError is raised.
         if 'multi-line string' in e.args[0]:
             toktypes.add(_MULTILINE_STRING)
         else:
             toktypes.add(_MULTILINE_STRUCTURE)
     return toktypes
 
 def has_comment(src):
     """Indicate whether an input line has (i.e. ends in, or is) a comment.
 
     This uses tokenize, so it can distinguish comments from # inside strings.
 
     Parameters
     ----------
     src : string
       A single line input string.
 
     Returns
     -------
     comment : bool
         True if source has a comment.
     """
     return (tokenize.COMMENT in _line_tokens(src))
 
 def ends_in_comment_or_string(src):
     """Indicates whether or not an input line ends in a comment or within
     a multiline string.
     
     Parameters
     ----------
     src : string
       A single line input string.
 
     Returns
     -------
     comment : bool
         True if source ends in a comment or multiline string.
     """
     toktypes = _line_tokens(src)
     return (tokenize.COMMENT in toktypes) or (_MULTILINE_STRING in toktypes)
         
 
 @StatelessInputTransformer.wrap
 def help_end(line):
     """Translate lines with ?/?? at the end"""
     m = _help_end_re.search(line)
     if m is None or ends_in_comment_or_string(line):
         return line
     target = m.group(1)
     esc = m.group(3)
     lspace = _initial_space_re.match(line).group(0)
 
     # If we're mid-command, put it back on the next prompt for the user.
     next_input = line.rstrip('?') if line.strip() != m.group(0) else None
 
     return _make_help_call(target, esc, lspace, next_input)
 
 
 @CoroutineInputTransformer.wrap
 def cellmagic(end_on_blank_line=False):
     """Captures & transforms cell magics.
     
     After a cell magic is started, this stores up any lines it gets until it is
     reset (sent None).
     """
     tpl = 'get_ipython().run_cell_magic(%r, %r, %r)'
     cellmagic_help_re = re.compile(r'%%\w+\?')
     line = ''
     while True:
         line = (yield line)
         # consume leading empty lines
         while not line:
             line = (yield line)
         
         if not line.startswith(ESC_MAGIC2):
             # This isn't a cell magic, idle waiting for reset then start over
             while line is not None:
                 line = (yield line)
             continue
         
         if cellmagic_help_re.match(line):
             # This case will be handled by help_end
             continue
         
         first = line
         body = []
         line = (yield None)
         while (line is not None) and \
                                 ((line.strip() != '') or not end_on_blank_line):
             body.append(line)
             line = (yield None)
         
         # Output
         magic_name, _, first = first.partition(' ')
         magic_name = magic_name.lstrip(ESC_MAGIC2)
         line = tpl % (magic_name, first, u'\n'.join(body))
 
 
 def _strip_prompts(prompt_re, initial_re=None, turnoff_re=None):
     """Remove matching input prompts from a block of input.
     
     Parameters
     ----------
     prompt_re : regular expression
         A regular expression matching any input prompt (including continuation)
     initial_re : regular expression, optional
         A regular expression matching only the initial prompt, but not continuation.
         If no initial expression is given, prompt_re will be used everywhere.
         Used mainly for plain Python prompts, where the continuation prompt
         ``...`` is a valid Python expression in Python 3, so shouldn't be stripped.
     
     If initial_re and prompt_re differ,
     only initial_re will be tested against the first line.
     If any prompt is found on the first two lines,
     prompts will be stripped from the rest of the block.
     """
     if initial_re is None:
         initial_re = prompt_re
     line = ''
     while True:
         line = (yield line)
         
         # First line of cell
         if line is None:
             continue
         out, n1 = initial_re.subn('', line, count=1)
         if turnoff_re and not n1:
             if turnoff_re.match(line):
                 # We're in e.g. a cell magic; disable this transformer for
                 # the rest of the cell.
                 while line is not None:
                     line = (yield line)
                 continue
 
         line = (yield out)
         
         if line is None:
             continue
         # check for any prompt on the second line of the cell,
         # because people often copy from just after the first prompt,
         # so we might not see it in the first line.
         out, n2 = prompt_re.subn('', line, count=1)
         line = (yield out)
         
         if n1 or n2:
             # Found a prompt in the first two lines - check for it in
             # the rest of the cell as well.
             while line is not None:
                 line = (yield prompt_re.sub('', line, count=1))
         
         else:
             # Prompts not in input - wait for reset
             while line is not None:
                 line = (yield line)
 
 @CoroutineInputTransformer.wrap
 def classic_prompt():
     """Strip the >>>/... prompts of the Python interactive shell."""
     # FIXME: non-capturing version (?:...) usable?
     prompt_re = re.compile(r'^(>>>|\.\.\.)( |$)')
     initial_re = re.compile(r'^>>>( |$)')
     # Any %magic/!system is IPython syntax, so we needn't look for >>> prompts
     turnoff_re = re.compile(r'^[%!]')
     return _strip_prompts(prompt_re, initial_re, turnoff_re)
 
 @CoroutineInputTransformer.wrap
 def ipy_prompt():
     """Strip IPython's In [1]:/...: prompts."""
     # FIXME: non-capturing version (?:...) usable?
     prompt_re = re.compile(r'^(In \[\d+\]: |\s*\.{3,}: ?)')
     # Disable prompt stripping inside cell magics
     turnoff_re = re.compile(r'^%%')
     return _strip_prompts(prompt_re, turnoff_re=turnoff_re)
 
 
 @CoroutineInputTransformer.wrap
 def leading_indent():
     """Remove leading indentation.
     
     If the first line starts with a spaces or tabs, the same whitespace will be
     removed from each following line until it is reset.
     """
     space_re = re.compile(r'^[ \t]+')
     line = ''
     while True:
         line = (yield line)
         
         if line is None:
             continue
         
         m = space_re.match(line)
         if m:
             space = m.group(0)
             while line is not None:
                 if line.startswith(space):
                     line = line[len(space):]
                 line = (yield line)
         else:
             # No leading spaces - wait for reset
             while line is not None:
                 line = (yield line)
 
 
 _assign_pat = \
 r'''(?P<lhs>(\s*)
     ([\w\.]+)                # Initial identifier
     (\s*,\s*
         \*?[\w\.]+)*         # Further identifiers for unpacking
     \s*?,?                   # Trailing comma
     )
     \s*=\s*
 '''
 
 assign_system_re = re.compile(r'{}!\s*(?P<cmd>.*)'.format(_assign_pat), re.VERBOSE)
 assign_system_template = '%s = get_ipython().getoutput(%r)'
 @StatelessInputTransformer.wrap
 def assign_from_system(line):
     """Transform assignment from system commands (e.g. files = !ls)"""
     m = assign_system_re.match(line)
     if m is None:
         return line
     
     return assign_system_template % m.group('lhs', 'cmd')
 
 assign_magic_re = re.compile(r'{}%\s*(?P<cmd>.*)'.format(_assign_pat), re.VERBOSE)
 assign_magic_template = '%s = get_ipython().run_line_magic(%r, %r)'
 @StatelessInputTransformer.wrap
 def assign_from_magic(line):
     """Transform assignment from magic commands (e.g. a = %who_ls)"""
     m = assign_magic_re.match(line)
     if m is None:
         return line
     #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args)
     m_lhs, m_cmd = m.group('lhs', 'cmd')
     t_magic_name, _, t_magic_arg_s = m_cmd.partition(' ')
     t_magic_name = t_magic_name.lstrip(ESC_MAGIC)
     return assign_magic_template % (m_lhs, t_magic_name, t_magic_arg_s)

IPython/core/inputtransformer2.py

@@ -1,721 +1,721 b''
 """Input transformer machinery to support IPython special syntax.
 
 This includes the machinery to recognise and transform ``%magic`` commands,
 ``!system`` commands, ``help?`` querying, prompt stripping, and so forth.
 
 Added: IPython 7.0. Replaces inputsplitter and inputtransformer which were
 deprecated in 7.0.
 """
 
 # Copyright (c) IPython Development Team.
 # Distributed under the terms of the Modified BSD License.
 
 from codeop import compile_command
 import re
 import tokenize
 from typing import List, Tuple, Union
 import warnings
 
 _indent_re = re.compile(r'^[ \t]+')
 
 def leading_empty_lines(lines):
     """Remove leading empty lines
 
     If the leading lines are empty or contain only whitespace, they will be
     removed.
     """
     if not lines:
         return lines
     for i, line in enumerate(lines):
         if line and not line.isspace():
             return lines[i:]
     return lines
 
 def leading_indent(lines):
     """Remove leading indentation.
 
     If the first line starts with a spaces or tabs, the same whitespace will be
     removed from each following line in the cell.
     """
     if not lines:
         return lines
     m = _indent_re.match(lines[0])
     if not m:
         return lines
     space = m.group(0)
     n = len(space)
     return [l[n:] if l.startswith(space) else l
             for l in lines]
 
 class PromptStripper:
     """Remove matching input prompts from a block of input.
 
     Parameters
     ----------
     prompt_re : regular expression
         A regular expression matching any input prompt (including continuation,
         e.g. ``...``)
     initial_re : regular expression, optional
         A regular expression matching only the initial prompt, but not continuation.
         If no initial expression is given, prompt_re will be used everywhere.
         Used mainly for plain Python prompts (``>>>``), where the continuation prompt
         ``...`` is a valid Python expression in Python 3, so shouldn't be stripped.
 
     If initial_re and prompt_re differ,
     only initial_re will be tested against the first line.
     If any prompt is found on the first two lines,
     prompts will be stripped from the rest of the block.
     """
     def __init__(self, prompt_re, initial_re=None):
         self.prompt_re = prompt_re
         self.initial_re = initial_re or prompt_re
 
     def _strip(self, lines):
         return [self.prompt_re.sub('', l, count=1) for l in lines]
 
     def __call__(self, lines):
         if not lines:
             return lines
         if self.initial_re.match(lines[0]) or \
                 (len(lines) > 1 and self.prompt_re.match(lines[1])):
             return self._strip(lines)
         return lines
 
 classic_prompt = PromptStripper(
     prompt_re=re.compile(r'^(>>>|\.\.\.)( |$)'),
     initial_re=re.compile(r'^>>>( |$)')
 )
 
 ipython_prompt = PromptStripper(re.compile(r'^(In \[\d+\]: |\s*\.{3,}: ?)'))
 
 def cell_magic(lines):
     if not lines or not lines[0].startswith('%%'):
         return lines
     if re.match(r'%%\w+\?', lines[0]):
         # This case will be handled by help_end
         return lines
     magic_name, _, first_line = lines[0][2:-1].partition(' ')
     body = ''.join(lines[1:])
     return ['get_ipython().run_cell_magic(%r, %r, %r)\n'
             % (magic_name, first_line, body)]
 
 
 def _find_assign_op(token_line) -> Union[int, None]:
     """Get the index of the first assignment in the line ('=' not inside brackets)
 
     Note: We don't try to support multiple special assignment (a = b = %foo)
     """
     paren_level = 0
     for i, ti in enumerate(token_line):
         s = ti.string
         if s == '=' and paren_level == 0:
             return i
         if s in {'(','[','{'}:
             paren_level += 1
         elif s in {')', ']', '}'}:
             if paren_level > 0:
                 paren_level -= 1
 
 def find_end_of_continued_line(lines, start_line: int):
     """Find the last line of a line explicitly extended using backslashes.
 
     Uses 0-indexed line numbers.
     """
     end_line = start_line
     while lines[end_line].endswith('\\\n'):
         end_line += 1
         if end_line >= len(lines):
             break
     return end_line
 
 def assemble_continued_line(lines, start: Tuple[int, int], end_line: int):
     r"""Assemble a single line from multiple continued line pieces
 
     Continued lines are lines ending in ``\``, and the line following the last
     ``\`` in the block.
 
     For example, this code continues over multiple lines::
 
         if (assign_ix is not None) \
              and (len(line) >= assign_ix + 2) \
              and (line[assign_ix+1].string == '%') \
              and (line[assign_ix+2].type == tokenize.NAME):
 
     This statement contains four continued line pieces.
     Assembling these pieces into a single line would give::
 
         if (assign_ix is not None) and (len(line) >= assign_ix + 2) and (line[...
 
     This uses 0-indexed line numbers. *start* is (lineno, colno).
 
     Used to allow ``%magic`` and ``!system`` commands to be continued over
     multiple lines.
     """
     parts = [lines[start[0]][start[1]:]] + lines[start[0]+1:end_line+1]
     return ' '.join([p[:-2] for p in parts[:-1]]  # Strip backslash+newline
                     + [parts[-1][:-1]])         # Strip newline from last line
 
 class TokenTransformBase:
     """Base class for transformations which examine tokens.
 
     Special syntax should not be transformed when it occurs inside strings or
     comments. This is hard to reliably avoid with regexes. The solution is to
     tokenise the code as Python, and recognise the special syntax in the tokens.
 
     IPython's special syntax is not valid Python syntax, so tokenising may go
     wrong after the special syntax starts. These classes therefore find and
     transform *one* instance of special syntax at a time into regular Python
     syntax. After each transformation, tokens are regenerated to find the next
     piece of special syntax.
 
     Subclasses need to implement one class method (find)
     and one regular method (transform).
 
     The priority attribute can select which transformation to apply if multiple
     transformers match in the same place. Lower numbers have higher priority.
     This allows "%magic?" to be turned into a help call rather than a magic call.
     """
     # Lower numbers -> higher priority (for matches in the same location)
     priority = 10
 
     def sortby(self):
         return self.start_line, self.start_col, self.priority
 
     def __init__(self, start):
         self.start_line = start[0] - 1   # Shift from 1-index to 0-index
         self.start_col = start[1]
 
     @classmethod
     def find(cls, tokens_by_line):
         """Find one instance of special syntax in the provided tokens.
 
         Tokens are grouped into logical lines for convenience,
         so it is easy to e.g. look at the first token of each line.
         *tokens_by_line* is a list of lists of tokenize.TokenInfo objects.
 
         This should return an instance of its class, pointing to the start
         position it has found, or None if it found no match.
         """
         raise NotImplementedError
 
     def transform(self, lines: List[str]):
         """Transform one instance of special syntax found by ``find()``
 
         Takes a list of strings representing physical lines,
         returns a similar list of transformed lines.
         """
         raise NotImplementedError
 
 class MagicAssign(TokenTransformBase):
     """Transformer for assignments from magics (a = %foo)"""
     @classmethod
     def find(cls, tokens_by_line):
         """Find the first magic assignment (a = %foo) in the cell.
         """
         for line in tokens_by_line:
             assign_ix = _find_assign_op(line)
             if (assign_ix is not None) \
                     and (len(line) >= assign_ix + 2) \
                     and (line[assign_ix+1].string == '%') \
                     and (line[assign_ix+2].type == tokenize.NAME):
                 return cls(line[assign_ix+1].start)
 
     def transform(self, lines: List[str]):
         """Transform a magic assignment found by the ``find()`` classmethod.
         """
         start_line, start_col = self.start_line, self.start_col
         lhs = lines[start_line][:start_col]
         end_line = find_end_of_continued_line(lines, start_line)
         rhs = assemble_continued_line(lines, (start_line, start_col), end_line)
         assert rhs.startswith('%'), rhs
         magic_name, _, args = rhs[1:].partition(' ')
 
         lines_before = lines[:start_line]
         call = "get_ipython().run_line_magic({!r}, {!r})".format(magic_name, args)
         new_line = lhs + call + '\n'
         lines_after = lines[end_line+1:]
 
         return lines_before + [new_line] + lines_after
 
 
 class SystemAssign(TokenTransformBase):
     """Transformer for assignments from system commands (a = !foo)"""
     @classmethod
     def find(cls, tokens_by_line):
         """Find the first system assignment (a = !foo) in the cell.
         """
         for line in tokens_by_line:
             assign_ix = _find_assign_op(line)
             if (assign_ix is not None) \
                     and not line[assign_ix].line.strip().startswith('=') \
                     and (len(line) >= assign_ix + 2) \
                     and (line[assign_ix + 1].type == tokenize.ERRORTOKEN):
                 ix = assign_ix + 1
 
                 while ix < len(line) and line[ix].type == tokenize.ERRORTOKEN:
                     if line[ix].string == '!':
                         return cls(line[ix].start)
                     elif not line[ix].string.isspace():
                         break
                     ix += 1
 
     def transform(self, lines: List[str]):
         """Transform a system assignment found by the ``find()`` classmethod.
         """
         start_line, start_col = self.start_line, self.start_col
 
         lhs = lines[start_line][:start_col]
         end_line = find_end_of_continued_line(lines, start_line)
         rhs = assemble_continued_line(lines, (start_line, start_col), end_line)
         assert rhs.startswith('!'), rhs
         cmd = rhs[1:]
 
         lines_before = lines[:start_line]
         call = "get_ipython().getoutput({!r})".format(cmd)
         new_line = lhs + call + '\n'
         lines_after = lines[end_line + 1:]
 
         return lines_before + [new_line] + lines_after
 
 # The escape sequences that define the syntax transformations IPython will
 # apply to user input.  These can NOT be just changed here: many regular
 # expressions and other parts of the code may use their hardcoded values, and
 # for all intents and purposes they constitute the 'IPython syntax', so they
 # should be considered fixed.
 
 ESC_SHELL  = '!'     # Send line to underlying system shell
 ESC_SH_CAP = '!!'    # Send line to system shell and capture output
 ESC_HELP   = '?'     # Find information about object
 ESC_HELP2  = '??'    # Find extra-detailed information about object
 ESC_MAGIC  = '%'     # Call magic function
 ESC_MAGIC2 = '%%'    # Call cell-magic function
 ESC_QUOTE  = ','     # Split args on whitespace, quote each as string and call
 ESC_QUOTE2 = ';'     # Quote all args as a single string, call
 ESC_PAREN  = '/'     # Call first argument with rest of line as arguments
 
 ESCAPE_SINGLES = {'!', '?', '%', ',', ';', '/'}
 ESCAPE_DOUBLES = {'!!', '??'}  # %% (cell magic) is handled separately
 
 def _make_help_call(target, esc, next_input=None):
     """Prepares a pinfo(2)/psearch call from a target name and the escape
     (i.e. ? or ??)"""
     method  = 'pinfo2' if esc == '??' \
                 else 'psearch' if '*' in target \
                 else 'pinfo'
     arg = " ".join([method, target])
     #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args)
     t_magic_name, _, t_magic_arg_s = arg.partition(' ')
     t_magic_name = t_magic_name.lstrip(ESC_MAGIC)
     if next_input is None:
         return 'get_ipython().run_line_magic(%r, %r)' % (t_magic_name, t_magic_arg_s)
     else:
         return 'get_ipython().set_next_input(%r);get_ipython().run_line_magic(%r, %r)' % \
            (next_input, t_magic_name, t_magic_arg_s)
 
 def _tr_help(content):
     """Translate lines escaped with: ?
 
     A naked help line should fire the intro help screen (shell.show_usage())
     """
     if not content:
         return 'get_ipython().show_usage()'
 
     return _make_help_call(content, '?')
 
 def _tr_help2(content):
     """Translate lines escaped with: ??
 
     A naked help line should fire the intro help screen (shell.show_usage())
     """
     if not content:
         return 'get_ipython().show_usage()'
 
     return _make_help_call(content, '??')
 
 def _tr_magic(content):
     "Translate lines escaped with a percent sign: %"
     name, _, args = content.partition(' ')
     return 'get_ipython().run_line_magic(%r, %r)' % (name, args)
 
 def _tr_quote(content):
     "Translate lines escaped with a comma: ,"
     name, _, args = content.partition(' ')
     return '%s("%s")' % (name, '", "'.join(args.split()) )
 
 def _tr_quote2(content):
     "Translate lines escaped with a semicolon: ;"
     name, _, args = content.partition(' ')
     return '%s("%s")' % (name, args)
 
 def _tr_paren(content):
     "Translate lines escaped with a slash: /"
     name, _, args = content.partition(' ')
     return '%s(%s)' % (name, ", ".join(args.split()))
 
 tr = { ESC_SHELL  : 'get_ipython().system({!r})'.format,
        ESC_SH_CAP : 'get_ipython().getoutput({!r})'.format,
        ESC_HELP   : _tr_help,
        ESC_HELP2  : _tr_help2,
        ESC_MAGIC  : _tr_magic,
        ESC_QUOTE  : _tr_quote,
        ESC_QUOTE2 : _tr_quote2,
        ESC_PAREN  : _tr_paren }
 
 class EscapedCommand(TokenTransformBase):
     """Transformer for escaped commands like %foo, !foo, or /foo"""
     @classmethod
     def find(cls, tokens_by_line):
         """Find the first escaped command (%foo, !foo, etc.) in the cell.
         """
         for line in tokens_by_line:
             if not line:
                 continue
             ix = 0
             ll = len(line)
             while ll > ix and line[ix].type in {tokenize.INDENT, tokenize.DEDENT}:
                 ix += 1
             if ix >= ll:
                 continue
             if line[ix].string in ESCAPE_SINGLES:
                 return cls(line[ix].start)
 
     def transform(self, lines):
         """Transform an escaped line found by the ``find()`` classmethod.
         """
         start_line, start_col = self.start_line, self.start_col
 
         indent = lines[start_line][:start_col]
         end_line = find_end_of_continued_line(lines, start_line)
         line = assemble_continued_line(lines, (start_line, start_col), end_line)
 
         if len(line) > 1 and line[:2] in ESCAPE_DOUBLES:
             escape, content = line[:2], line[2:]
         else:
             escape, content = line[:1], line[1:]
 
         if escape in tr:
             call = tr[escape](content)
         else:
             call = ''
 
         lines_before = lines[:start_line]
         new_line = indent + call + '\n'
         lines_after = lines[end_line + 1:]
 
         return lines_before + [new_line] + lines_after
 
 _help_end_re = re.compile(r"""(%{0,2}
-                              [a-zA-Z_*][\w*]*        # Variable name
-                              (\.[a-zA-Z_*][\w*]*)*   # .etc.etc
+                              (?!\d)[\w*]+            # Variable name
+                              (\.(?!\d)[\w*]+)*       # .etc.etc
                               )
                               (\?\??)$                # ? or ??
                               """,
                               re.VERBOSE)
 
 class HelpEnd(TokenTransformBase):
     """Transformer for help syntax: obj? and obj??"""
     # This needs to be higher priority (lower number) than EscapedCommand so
     # that inspecting magics (%foo?) works.
     priority = 5
 
     def __init__(self, start, q_locn):
         super().__init__(start)
         self.q_line = q_locn[0] - 1  # Shift from 1-indexed to 0-indexed
         self.q_col = q_locn[1]
 
     @classmethod
     def find(cls, tokens_by_line):
         """Find the first help command (foo?) in the cell.
         """
         for line in tokens_by_line:
             # Last token is NEWLINE; look at last but one
             if len(line) > 2 and line[-2].string == '?':
                 # Find the first token that's not INDENT/DEDENT
                 ix = 0
                 while line[ix].type in {tokenize.INDENT, tokenize.DEDENT}:
                     ix += 1
                 return cls(line[ix].start, line[-2].start)
 
     def transform(self, lines):
         """Transform a help command found by the ``find()`` classmethod.
         """
         piece = ''.join(lines[self.start_line:self.q_line+1])
         indent, content = piece[:self.start_col], piece[self.start_col:]
         lines_before = lines[:self.start_line]
         lines_after = lines[self.q_line + 1:]
 
         m = _help_end_re.search(content)
         if not m:
             raise SyntaxError(content)
         assert m is not None, content
         target = m.group(1)
         esc = m.group(3)
 
         # If we're mid-command, put it back on the next prompt for the user.
         next_input = None
         if (not lines_before) and (not lines_after) \
                 and content.strip() != m.group(0):
             next_input = content.rstrip('?\n')
 
         call = _make_help_call(target, esc, next_input=next_input)
         new_line = indent + call + '\n'
 
         return lines_before + [new_line] + lines_after
 
 def make_tokens_by_line(lines:List[str]):
     """Tokenize a series of lines and group tokens by line.
 
     The tokens for a multiline Python string or expression are grouped as one
     line. All lines except the last lines should keep their line ending ('\\n',
     '\\r\\n') for this to properly work. Use `.splitlines(keeplineending=True)`
     for example when passing block of text to this function.
 
     """
     # NL tokens are used inside multiline expressions, but also after blank
     # lines or comments. This is intentional - see https://bugs.python.org/issue17061
     # We want to group the former case together but split the latter, so we
     # track parentheses level, similar to the internals of tokenize.
     NEWLINE, NL = tokenize.NEWLINE, tokenize.NL
     tokens_by_line = [[]]
     if len(lines) > 1 and not lines[0].endswith(('\n', '\r', '\r\n', '\x0b', '\x0c')):
         warnings.warn("`make_tokens_by_line` received a list of lines which do not have lineending markers ('\\n', '\\r', '\\r\\n', '\\x0b', '\\x0c'), behavior will be unspecified")
     parenlev = 0
     try:
         for token in tokenize.generate_tokens(iter(lines).__next__):
             tokens_by_line[-1].append(token)
             if (token.type == NEWLINE) \
                     or ((token.type == NL) and (parenlev <= 0)):
                 tokens_by_line.append([])
             elif token.string in {'(', '[', '{'}:
                 parenlev += 1
             elif token.string in {')', ']', '}'}:
                 if parenlev > 0:
                     parenlev -= 1
     except tokenize.TokenError:
         # Input ended in a multiline string or expression. That's OK for us.
         pass
 
 
     if not tokens_by_line[-1]:
         tokens_by_line.pop()
 
 
     return tokens_by_line
 
 def show_linewise_tokens(s: str):
     """For investigation and debugging"""
     if not s.endswith('\n'):
         s += '\n'
     lines = s.splitlines(keepends=True)
     for line in make_tokens_by_line(lines):
         print("Line -------")
         for tokinfo in line:
             print(" ", tokinfo)
 
 # Arbitrary limit to prevent getting stuck in infinite loops
 TRANSFORM_LOOP_LIMIT = 500
 
 class TransformerManager:
     """Applies various transformations to a cell or code block.
 
     The key methods for external use are ``transform_cell()``
     and ``check_complete()``.
     """
     def __init__(self):
         self.cleanup_transforms = [
             leading_empty_lines,
             leading_indent,
             classic_prompt,
             ipython_prompt,
         ]
         self.line_transforms = [
             cell_magic,
         ]
         self.token_transformers = [
             MagicAssign,
             SystemAssign,
             EscapedCommand,
             HelpEnd,
         ]
 
     def do_one_token_transform(self, lines):
         """Find and run the transform earliest in the code.
 
         Returns (changed, lines).
 
         This method is called repeatedly until changed is False, indicating
         that all available transformations are complete.
 
         The tokens following IPython special syntax might not be valid, so
         the transformed code is retokenised every time to identify the next
         piece of special syntax. Hopefully long code cells are mostly valid
         Python, not using lots of IPython special syntax, so this shouldn't be
         a performance issue.
         """
         tokens_by_line = make_tokens_by_line(lines)
         candidates = []
         for transformer_cls in self.token_transformers:
             transformer = transformer_cls.find(tokens_by_line)
             if transformer:
                 candidates.append(transformer)
 
         if not candidates:
             # Nothing to transform
             return False, lines
         ordered_transformers = sorted(candidates, key=TokenTransformBase.sortby)
         for transformer in ordered_transformers:
             try:
                 return True, transformer.transform(lines)
             except SyntaxError:
                 pass
         return False, lines
 
     def do_token_transforms(self, lines):
         for _ in range(TRANSFORM_LOOP_LIMIT):
             changed, lines = self.do_one_token_transform(lines)
             if not changed:
                 return lines
 
         raise RuntimeError("Input transformation still changing after "
                            "%d iterations. Aborting." % TRANSFORM_LOOP_LIMIT)
 
     def transform_cell(self, cell: str) -> str:
         """Transforms a cell of input code"""
         if not cell.endswith('\n'):
             cell += '\n'  # Ensure the cell has a trailing newline
         lines = cell.splitlines(keepends=True)
         for transform in self.cleanup_transforms + self.line_transforms:
             lines = transform(lines)
 
         lines = self.do_token_transforms(lines)
         return ''.join(lines)
 
     def check_complete(self, cell: str):
         """Return whether a block of code is ready to execute, or should be continued
 
         Parameters
         ----------
         source : string
           Python input code, which can be multiline.
 
         Returns
         -------
         status : str
           One of 'complete', 'incomplete', or 'invalid' if source is not a
           prefix of valid code.
         indent_spaces : int or None
           The number of spaces by which to indent the next line of code. If
           status is not 'incomplete', this is None.
         """
         # Remember if the lines ends in a new line.
         ends_with_newline = False
         for character in reversed(cell):
             if character == '\n':
                 ends_with_newline = True
                 break
             elif character.strip():
                 break
             else:
                 continue
 
         if not ends_with_newline:
             # Append an newline for consistent tokenization
             # See https://bugs.python.org/issue33899
             cell += '\n'
 
         lines = cell.splitlines(keepends=True)
 
         if not lines:
             return 'complete', None
 
         if lines[-1].endswith('\\'):
             # Explicit backslash continuation
             return 'incomplete', find_last_indent(lines)
 
         try:
             for transform in self.cleanup_transforms:
                 lines = transform(lines)
         except SyntaxError:
             return 'invalid', None
 
         if lines[0].startswith('%%'):
             # Special case for cell magics - completion marked by blank line
             if lines[-1].strip():
                 return 'incomplete', find_last_indent(lines)
             else:
                 return 'complete', None
 
         try:
             for transform in self.line_transforms:
                 lines = transform(lines)
             lines = self.do_token_transforms(lines)
         except SyntaxError:
             return 'invalid', None
 
         tokens_by_line = make_tokens_by_line(lines)
 
         if not tokens_by_line:
             return 'incomplete', find_last_indent(lines)
 
         if tokens_by_line[-1][-1].type != tokenize.ENDMARKER:
             # We're in a multiline string or expression
             return 'incomplete', find_last_indent(lines)
 
         newline_types = {tokenize.NEWLINE, tokenize.COMMENT, tokenize.ENDMARKER}
 
         # Pop the last line which only contains DEDENTs and ENDMARKER
         last_token_line = None
         if {t.type for t in tokens_by_line[-1]} in [
             {tokenize.DEDENT, tokenize.ENDMARKER},
             {tokenize.ENDMARKER}
         ] and len(tokens_by_line) > 1:
             last_token_line = tokens_by_line.pop()
 
         while tokens_by_line[-1] and tokens_by_line[-1][-1].type in newline_types:
             tokens_by_line[-1].pop()
 
         if not tokens_by_line[-1]:
             return 'incomplete', find_last_indent(lines)
 
         if tokens_by_line[-1][-1].string == ':':
             # The last line starts a block (e.g. 'if foo:')
             ix = 0
             while tokens_by_line[-1][ix].type in {tokenize.INDENT, tokenize.DEDENT}:
                 ix += 1
 
             indent = tokens_by_line[-1][ix].start[1]
             return 'incomplete', indent + 4
 
         if tokens_by_line[-1][0].line.endswith('\\'):
             return 'incomplete', None
 
         # At this point, our checks think the code is complete (or invalid).
         # We'll use codeop.compile_command to check this with the real parser
         try:
             with warnings.catch_warnings():
                 warnings.simplefilter('error', SyntaxWarning)
                 res = compile_command(''.join(lines), symbol='exec')
         except (SyntaxError, OverflowError, ValueError, TypeError,
                 MemoryError, SyntaxWarning):
             return 'invalid', None
         else:
             if res is None:
                 return 'incomplete', find_last_indent(lines)
 
         if last_token_line and last_token_line[0].type == tokenize.DEDENT:
             if ends_with_newline:
                 return 'complete', None
             return 'incomplete', find_last_indent(lines)
 
         # If there's a blank line at the end, assume we're ready to execute
         if not lines[-1].strip():
             return 'complete', None
 
         return 'complete', None
 
 
 def find_last_indent(lines):
     m = _indent_re.match(lines[-1])
     if not m:
         return 0
     return len(m.group(0).replace('\t', ' '*4))

IPython/core/magics/namespace.py

@@ -1,714 +1,708 b''
 """Implementation of namespace-related magic functions.
 """
 #-----------------------------------------------------------------------------
 #  Copyright (c) 2012 The IPython Development Team.
 #
 #  Distributed under the terms of the Modified BSD License.
 #
 #  The full license is in the file COPYING.txt, distributed with this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 # Stdlib
 import gc
 import re
 import sys
 
 # Our own packages
 from IPython.core import page
 from IPython.core.error import StdinNotImplementedError, UsageError
 from IPython.core.magic import Magics, magics_class, line_magic
 from IPython.testing.skipdoctest import skip_doctest
 from IPython.utils.encoding import DEFAULT_ENCODING
 from IPython.utils.openpy import read_py_file
 from IPython.utils.path import get_py_filename
 
 #-----------------------------------------------------------------------------
 # Magic implementation classes
 #-----------------------------------------------------------------------------
 
 @magics_class
 class NamespaceMagics(Magics):
     """Magics to manage various aspects of the user's namespace.
 
     These include listing variables, introspecting into them, etc.
     """
 
     @line_magic
     def pinfo(self, parameter_s='', namespaces=None):
         """Provide detailed information about an object.
 
         '%pinfo object' is just a synonym for object? or ?object."""
 
         #print 'pinfo par: <%s>' % parameter_s  # dbg
         # detail_level: 0 -> obj? , 1 -> obj??
         detail_level = 0
         # We need to detect if we got called as 'pinfo pinfo foo', which can
         # happen if the user types 'pinfo foo?' at the cmd line.
         pinfo,qmark1,oname,qmark2 = \
                re.match(r'(pinfo )?(\?*)(.*?)(\??$)',parameter_s).groups()
         if pinfo or qmark1 or qmark2:
             detail_level = 1
         if "*" in oname:
             self.psearch(oname)
         else:
             self.shell._inspect('pinfo', oname, detail_level=detail_level,
                                 namespaces=namespaces)
 
     @line_magic
     def pinfo2(self, parameter_s='', namespaces=None):
         """Provide extra detailed information about an object.
 
         '%pinfo2 object' is just a synonym for object?? or ??object."""
         self.shell._inspect('pinfo', parameter_s, detail_level=1,
                             namespaces=namespaces)
 
     @skip_doctest
     @line_magic
     def pdef(self, parameter_s='', namespaces=None):
         """Print the call signature for any callable object.
 
         If the object is a class, print the constructor information.
 
         Examples
         --------
         ::
 
           In [3]: %pdef urllib.urlopen
           urllib.urlopen(url, data=None, proxies=None)
         """
         self.shell._inspect('pdef',parameter_s, namespaces)
 
     @line_magic
     def pdoc(self, parameter_s='', namespaces=None):
         """Print the docstring for an object.
 
         If the given object is a class, it will print both the class and the
         constructor docstrings."""
         self.shell._inspect('pdoc',parameter_s, namespaces)
 
     @line_magic
     def psource(self, parameter_s='', namespaces=None):
         """Print (or run through pager) the source code for an object."""
         if not parameter_s:
             raise UsageError('Missing object name.')
         self.shell._inspect('psource',parameter_s, namespaces)
 
     @line_magic
     def pfile(self, parameter_s='', namespaces=None):
         """Print (or run through pager) the file where an object is defined.
 
         The file opens at the line where the object definition begins. IPython
         will honor the environment variable PAGER if set, and otherwise will
         do its best to print the file in a convenient form.
 
         If the given argument is not an object currently defined, IPython will
         try to interpret it as a filename (automatically adding a .py extension
         if needed). You can thus use %pfile as a syntax highlighting code
         viewer."""
 
         # first interpret argument as an object name
         out = self.shell._inspect('pfile',parameter_s, namespaces)
         # if not, try the input as a filename
         if out == 'not found':
             try:
                 filename = get_py_filename(parameter_s)
             except IOError as msg:
                 print(msg)
                 return
             page.page(self.shell.pycolorize(read_py_file(filename, skip_encoding_cookie=False)))
 
     @line_magic
     def psearch(self, parameter_s=''):
         """Search for object in namespaces by wildcard.
 
         %psearch [options] PATTERN [OBJECT TYPE]
 
         Note: ? can be used as a synonym for %psearch, at the beginning or at
         the end: both a*? and ?a* are equivalent to '%psearch a*'.  Still, the
         rest of the command line must be unchanged (options come first), so
         for example the following forms are equivalent
 
         %psearch -i a* function
         -i a* function?
         ?-i a* function
 
         Arguments:
 
           PATTERN
 
           where PATTERN is a string containing * as a wildcard similar to its
           use in a shell.  The pattern is matched in all namespaces on the
           search path. By default objects starting with a single _ are not
           matched, many IPython generated objects have a single
           underscore. The default is case insensitive matching. Matching is
           also done on the attributes of objects and not only on the objects
           in a module.
 
           [OBJECT TYPE]
 
           Is the name of a python type from the types module. The name is
           given in lowercase without the ending type, ex. StringType is
           written string. By adding a type here only objects matching the
           given type are matched. Using all here makes the pattern match all
           types (this is the default).
 
         Options:
 
           -a: makes the pattern match even objects whose names start with a
           single underscore.  These names are normally omitted from the
           search.
 
           -i/-c: make the pattern case insensitive/sensitive.  If neither of
           these options are given, the default is read from your configuration
           file, with the option ``InteractiveShell.wildcards_case_sensitive``.
           If this option is not specified in your configuration file, IPython's
           internal default is to do a case sensitive search.
 
           -e/-s NAMESPACE: exclude/search a given namespace.  The pattern you
           specify can be searched in any of the following namespaces:
           'builtin', 'user', 'user_global','internal', 'alias', where
           'builtin' and 'user' are the search defaults.  Note that you should
           not use quotes when specifying namespaces.
           
           -l: List all available object types for object matching. This function
           can be used without arguments.
 
           'Builtin' contains the python module builtin, 'user' contains all
           user data, 'alias' only contain the shell aliases and no python
           objects, 'internal' contains objects used by IPython.  The
           'user_global' namespace is only used by embedded IPython instances,
           and it contains module-level globals.  You can add namespaces to the
           search with -s or exclude them with -e (these options can be given
           more than once).
 
         Examples
         --------
         ::
 
           %psearch a*            -> objects beginning with an a
           %psearch -e builtin a* -> objects NOT in the builtin space starting in a
           %psearch a* function   -> all functions beginning with an a
           %psearch re.e*         -> objects beginning with an e in module re
           %psearch r*.e*         -> objects that start with e in modules starting in r
           %psearch r*.* string   -> all strings in modules beginning with r
 
         Case sensitive search::
 
           %psearch -c a*         list all object beginning with lower case a
 
         Show objects beginning with a single _::
 
           %psearch -a _*         list objects beginning with a single underscore
           
         List available objects::
         
           %psearch -l            list all available object types
         """
-        try:
-            parameter_s.encode('ascii')
-        except UnicodeEncodeError:
-            print('Python identifiers can only contain ascii characters.')
-            return
-
         # default namespaces to be searched
         def_search = ['user_local', 'user_global', 'builtin']
 
         # Process options/args
         opts,args = self.parse_options(parameter_s,'cias:e:l',list_all=True)
         opt = opts.get
         shell = self.shell
         psearch = shell.inspector.psearch
         
         # select list object types
         list_types = False
         if 'l' in opts:
             list_types = True
 
         # select case options
         if 'i' in opts:
             ignore_case = True
         elif 'c' in opts:
             ignore_case = False
         else:
             ignore_case = not shell.wildcards_case_sensitive
 
         # Build list of namespaces to search from user options
         def_search.extend(opt('s',[]))
         ns_exclude = ns_exclude=opt('e',[])
         ns_search = [nm for nm in def_search if nm not in ns_exclude]
 
         # Call the actual search
         try:
             psearch(args,shell.ns_table,ns_search,
                     show_all=opt('a'),ignore_case=ignore_case, list_types=list_types)
         except:
             shell.showtraceback()
 
     @skip_doctest
     @line_magic
     def who_ls(self, parameter_s=''):
         """Return a sorted list of all interactive variables.
 
         If arguments are given, only variables of types matching these
         arguments are returned.
 
         Examples
         --------
 
         Define two variables and list them with who_ls::
 
           In [1]: alpha = 123
 
           In [2]: beta = 'test'
 
           In [3]: %who_ls
           Out[3]: ['alpha', 'beta']
 
           In [4]: %who_ls int
           Out[4]: ['alpha']
 
           In [5]: %who_ls str
           Out[5]: ['beta']
         """
 
         user_ns = self.shell.user_ns
         user_ns_hidden = self.shell.user_ns_hidden
         nonmatching = object()  # This can never be in user_ns
         out = [ i for i in user_ns
                 if not i.startswith('_') \
                 and (user_ns[i] is not user_ns_hidden.get(i, nonmatching)) ]
 
         typelist = parameter_s.split()
         if typelist:
             typeset = set(typelist)
             out = [i for i in out if type(user_ns[i]).__name__ in typeset]
 
         out.sort()
         return out
 
     @skip_doctest
     @line_magic
     def who(self, parameter_s=''):
         """Print all interactive variables, with some minimal formatting.
 
         If any arguments are given, only variables whose type matches one of
         these are printed.  For example::
 
           %who function str
 
         will only list functions and strings, excluding all other types of
         variables.  To find the proper type names, simply use type(var) at a
         command line to see how python prints type names.  For example:
 
         ::
 
           In [1]: type('hello')\\
           Out[1]: <type 'str'>
 
         indicates that the type name for strings is 'str'.
 
         ``%who`` always excludes executed names loaded through your configuration
         file and things which are internal to IPython.
 
         This is deliberate, as typically you may load many modules and the
         purpose of %who is to show you only what you've manually defined.
 
         Examples
         --------
 
         Define two variables and list them with who::
 
           In [1]: alpha = 123
 
           In [2]: beta = 'test'
 
           In [3]: %who
           alpha   beta
 
           In [4]: %who int
           alpha
 
           In [5]: %who str
           beta
         """
 
         varlist = self.who_ls(parameter_s)
         if not varlist:
             if parameter_s:
                 print('No variables match your requested type.')
             else:
                 print('Interactive namespace is empty.')
             return
 
         # if we have variables, move on...
         count = 0
         for i in varlist:
             print(i+'\t', end=' ')
             count += 1
             if count > 8:
                 count = 0
                 print()
         print()
 
     @skip_doctest
     @line_magic
     def whos(self, parameter_s=''):
         """Like %who, but gives some extra information about each variable.
 
         The same type filtering of %who can be applied here.
 
         For all variables, the type is printed. Additionally it prints:
 
           - For {},[],(): their length.
 
           - For numpy arrays, a summary with shape, number of
             elements, typecode and size in memory.
 
           - Everything else: a string representation, snipping their middle if
             too long.
 
         Examples
         --------
 
         Define two variables and list them with whos::
 
           In [1]: alpha = 123
 
           In [2]: beta = 'test'
 
           In [3]: %whos
           Variable   Type        Data/Info
           --------------------------------
           alpha      int         123
           beta       str         test
         """
 
         varnames = self.who_ls(parameter_s)
         if not varnames:
             if parameter_s:
                 print('No variables match your requested type.')
             else:
                 print('Interactive namespace is empty.')
             return
 
         # if we have variables, move on...
 
         # for these types, show len() instead of data:
         seq_types = ['dict', 'list', 'tuple']
 
         # for numpy arrays, display summary info
         ndarray_type = None
         if 'numpy' in sys.modules:
             try:
                 from numpy import ndarray
             except ImportError:
                 pass
             else:
                 ndarray_type = ndarray.__name__
 
         # Find all variable names and types so we can figure out column sizes
 
         # some types are well known and can be shorter
         abbrevs = {'IPython.core.macro.Macro' : 'Macro'}
         def type_name(v):
             tn = type(v).__name__
             return abbrevs.get(tn,tn)
 
         varlist = [self.shell.user_ns[n] for n in varnames]
 
         typelist = []
         for vv in varlist:
             tt = type_name(vv)
 
             if tt=='instance':
                 typelist.append( abbrevs.get(str(vv.__class__),
                                              str(vv.__class__)))
             else:
                 typelist.append(tt)
 
         # column labels and # of spaces as separator
         varlabel = 'Variable'
         typelabel = 'Type'
         datalabel = 'Data/Info'
         colsep = 3
         # variable format strings
         vformat    = "{0:<{varwidth}}{1:<{typewidth}}"
         aformat    = "%s: %s elems, type `%s`, %s bytes"
         # find the size of the columns to format the output nicely
         varwidth = max(max(map(len,varnames)), len(varlabel)) + colsep
         typewidth = max(max(map(len,typelist)), len(typelabel)) + colsep
         # table header
         print(varlabel.ljust(varwidth) + typelabel.ljust(typewidth) + \
               ' '+datalabel+'\n' + '-'*(varwidth+typewidth+len(datalabel)+1))
         # and the table itself
         kb = 1024
         Mb = 1048576  # kb**2
         for vname,var,vtype in zip(varnames,varlist,typelist):
             print(vformat.format(vname, vtype, varwidth=varwidth, typewidth=typewidth), end=' ')
             if vtype in seq_types:
                 print("n="+str(len(var)))
             elif vtype == ndarray_type:
                 vshape = str(var.shape).replace(',','').replace(' ','x')[1:-1]
                 if vtype==ndarray_type:
                     # numpy
                     vsize  = var.size
                     vbytes = vsize*var.itemsize
                     vdtype = var.dtype
 
                 if vbytes < 100000:
                     print(aformat % (vshape, vsize, vdtype, vbytes))
                 else:
                     print(aformat % (vshape, vsize, vdtype, vbytes), end=' ')
                     if vbytes < Mb:
                         print('(%s kb)' % (vbytes/kb,))
                     else:
                         print('(%s Mb)' % (vbytes/Mb,))
             else:
                 try:
                     vstr = str(var)
                 except UnicodeEncodeError:
                     vstr = var.encode(DEFAULT_ENCODING,
                                       'backslashreplace')
                 except:
                     vstr = "<object with id %d (str() failed)>" % id(var)
                 vstr = vstr.replace('\n', '\\n')
                 if len(vstr) < 50:
                     print(vstr)
                 else:
                     print(vstr[:25] + "<...>" + vstr[-25:])
 
     @line_magic
     def reset(self, parameter_s=''):
         """Resets the namespace by removing all names defined by the user, if
         called without arguments, or by removing some types of objects, such
         as everything currently in IPython's In[] and Out[] containers (see
         the parameters for details).
 
         Parameters
         ----------
         -f : force reset without asking for confirmation.
 
         -s : 'Soft' reset: Only clears your namespace, leaving history intact.
             References to objects may be kept. By default (without this option),
             we do a 'hard' reset, giving you a new session and removing all
             references to objects from the current session.
 
         in : reset input history
 
         out : reset output history
 
         dhist : reset directory history
 
         array : reset only variables that are NumPy arrays
 
         See Also
         --------
         reset_selective : invoked as ``%reset_selective``
 
         Examples
         --------
         ::
 
           In [6]: a = 1
 
           In [7]: a
           Out[7]: 1
 
           In [8]: 'a' in get_ipython().user_ns
           Out[8]: True
 
           In [9]: %reset -f
 
           In [1]: 'a' in get_ipython().user_ns
           Out[1]: False
 
           In [2]: %reset -f in
           Flushing input history
 
           In [3]: %reset -f dhist in
           Flushing directory history
           Flushing input history
 
         Notes
         -----
         Calling this magic from clients that do not implement standard input,
         such as the ipython notebook interface, will reset the namespace
         without confirmation.
         """
         opts, args = self.parse_options(parameter_s,'sf', mode='list')
         if 'f' in opts:
             ans = True
         else:
             try:
                 ans = self.shell.ask_yes_no(
                 "Once deleted, variables cannot be recovered. Proceed (y/[n])?",
                 default='n')
             except StdinNotImplementedError:
                 ans = True
         if not ans:
             print('Nothing done.')
             return
 
         if 's' in opts:                     # Soft reset
             user_ns = self.shell.user_ns
             for i in self.who_ls():
                 del(user_ns[i])
         elif len(args) == 0:                # Hard reset
             self.shell.reset(new_session = False)
 
         # reset in/out/dhist/array: previously extensinions/clearcmd.py
         ip = self.shell
         user_ns = self.shell.user_ns  # local lookup, heavily used
 
         for target in args:
             target = target.lower() # make matches case insensitive
             if target == 'out':
                 print("Flushing output cache (%d entries)" % len(user_ns['_oh']))
                 self.shell.displayhook.flush()
 
             elif target == 'in':
                 print("Flushing input history")
                 pc = self.shell.displayhook.prompt_count + 1
                 for n in range(1, pc):
                     key = '_i'+repr(n)
                     user_ns.pop(key,None)
                 user_ns.update(dict(_i=u'',_ii=u'',_iii=u''))
                 hm = ip.history_manager
                 # don't delete these, as %save and %macro depending on the
                 # length of these lists to be preserved
                 hm.input_hist_parsed[:] = [''] * pc
                 hm.input_hist_raw[:] = [''] * pc
                 # hm has internal machinery for _i,_ii,_iii, clear it out
                 hm._i = hm._ii = hm._iii = hm._i00 =  u''
 
             elif target == 'array':
                 # Support cleaning up numpy arrays
                 try:
                     from numpy import ndarray
                     # This must be done with items and not iteritems because
                     # we're going to modify the dict in-place.
                     for x,val in list(user_ns.items()):
                         if isinstance(val,ndarray):
                             del user_ns[x]
                 except ImportError:
                     print("reset array only works if Numpy is available.")
 
             elif target == 'dhist':
                 print("Flushing directory history")
                 del user_ns['_dh'][:]
 
             else:
                 print("Don't know how to reset ", end=' ')
                 print(target + ", please run `%reset?` for details")
 
         gc.collect()
 
     @line_magic
     def reset_selective(self, parameter_s=''):
         """Resets the namespace by removing names defined by the user.
 
         Input/Output history are left around in case you need them.
 
         %reset_selective [-f] regex
 
         No action is taken if regex is not included
 
         Options
           -f : force reset without asking for confirmation.
 
         See Also
         --------
         reset : invoked as ``%reset``
 
         Examples
         --------
 
         We first fully reset the namespace so your output looks identical to
         this example for pedagogical reasons; in practice you do not need a
         full reset::
 
           In [1]: %reset -f
 
         Now, with a clean namespace we can make a few variables and use
         ``%reset_selective`` to only delete names that match our regexp::
 
           In [2]: a=1; b=2; c=3; b1m=4; b2m=5; b3m=6; b4m=7; b2s=8
 
           In [3]: who_ls
           Out[3]: ['a', 'b', 'b1m', 'b2m', 'b2s', 'b3m', 'b4m', 'c']
 
           In [4]: %reset_selective -f b[2-3]m
 
           In [5]: who_ls
           Out[5]: ['a', 'b', 'b1m', 'b2s', 'b4m', 'c']
 
           In [6]: %reset_selective -f d
 
           In [7]: who_ls
           Out[7]: ['a', 'b', 'b1m', 'b2s', 'b4m', 'c']
 
           In [8]: %reset_selective -f c
 
           In [9]: who_ls
           Out[9]: ['a', 'b', 'b1m', 'b2s', 'b4m']
 
           In [10]: %reset_selective -f b
 
           In [11]: who_ls
           Out[11]: ['a']
 
         Notes
         -----
         Calling this magic from clients that do not implement standard input,
         such as the ipython notebook interface, will reset the namespace
         without confirmation.
         """
 
         opts, regex = self.parse_options(parameter_s,'f')
 
         if 'f' in opts:
             ans = True
         else:
             try:
                 ans = self.shell.ask_yes_no(
                 "Once deleted, variables cannot be recovered. Proceed (y/[n])? ",
                 default='n')
             except StdinNotImplementedError:
                 ans = True
         if not ans:
             print('Nothing done.')
             return
         user_ns = self.shell.user_ns
         if not regex:
             print('No regex pattern specified. Nothing done.')
             return
         else:
             try:
                 m = re.compile(regex)
             except TypeError:
                 raise TypeError('regex must be a string or compiled pattern')
             for i in self.who_ls():
                 if m.search(i):
                     del(user_ns[i])
 
     @line_magic
     def xdel(self, parameter_s=''):
         """Delete a variable, trying to clear it from anywhere that
         IPython's machinery has references to it. By default, this uses
         the identity of the named object in the user namespace to remove
         references held under other names. The object is also removed
         from the output history.
 
         Options
           -n : Delete the specified name from all namespaces, without
           checking their identity.
         """
         opts, varname = self.parse_options(parameter_s,'n')
         try:
             self.shell.del_var(varname, ('n' in opts))
         except (NameError, ValueError) as e:
             print(type(e).__name__ +": "+ str(e))

IPython/core/prefilter.py

@@ -1,709 +1,709 b''
 # encoding: utf-8
 """
 Prefiltering components.
 
 Prefilters transform user input before it is exec'd by Python.  These
 transforms are used to implement additional syntax such as !ls and %magic.
 """
 
 # Copyright (c) IPython Development Team.
 # Distributed under the terms of the Modified BSD License.
 
 from keyword import iskeyword
 import re
 
 from .autocall import IPyAutocall
 from traitlets.config.configurable import Configurable
 from .inputtransformer2 import (
     ESC_MAGIC,
     ESC_QUOTE,
     ESC_QUOTE2,
     ESC_PAREN,
 )
 from .macro import Macro
 from .splitinput import LineInfo
 
 from traitlets import (
     List, Integer, Unicode, Bool, Instance, CRegExp
 )
 
 #-----------------------------------------------------------------------------
 # Global utilities, errors and constants
 #-----------------------------------------------------------------------------
 
 
 class PrefilterError(Exception):
     pass
 
 
 # RegExp to identify potential function names
-re_fun_name = re.compile(r'[a-zA-Z_]([a-zA-Z0-9_.]*) *$')
+re_fun_name = re.compile(r'[^\W\d]([\w.]*) *$')
 
 # RegExp to exclude strings with this start from autocalling.  In
 # particular, all binary operators should be excluded, so that if foo is
 # callable, foo OP bar doesn't become foo(OP bar), which is invalid.  The
 # characters '!=()' don't need to be checked for, as the checkPythonChars
 # routine explicitly does so, to catch direct calls and rebindings of
 # existing names.
 
 # Warning: the '-' HAS TO BE AT THE END of the first group, otherwise
 # it affects the rest of the group in square brackets.
 re_exclude_auto = re.compile(r'^[,&^\|\*/\+-]'
                              r'|^is |^not |^in |^and |^or ')
 
 # try to catch also methods for stuff in lists/tuples/dicts: off
 # (experimental). For this to work, the line_split regexp would need
 # to be modified so it wouldn't break things at '['. That line is
 # nasty enough that I shouldn't change it until I can test it _well_.
 #self.re_fun_name = re.compile (r'[a-zA-Z_]([a-zA-Z0-9_.\[\]]*) ?$')
 
 
 # Handler Check Utilities
 def is_shadowed(identifier, ip):
     """Is the given identifier defined in one of the namespaces which shadow
     the alias and magic namespaces?  Note that an identifier is different
     than ifun, because it can not contain a '.' character."""
     # This is much safer than calling ofind, which can change state
     return (identifier in ip.user_ns \
             or identifier in ip.user_global_ns \
             or identifier in ip.ns_table['builtin']\
             or iskeyword(identifier))
 
 
 #-----------------------------------------------------------------------------
 # Main Prefilter manager
 #-----------------------------------------------------------------------------
 
 
 class PrefilterManager(Configurable):
     """Main prefilter component.
 
     The IPython prefilter is run on all user input before it is run.  The
     prefilter consumes lines of input and produces transformed lines of
     input.
 
     The implementation consists of two phases:
 
     1. Transformers
     2. Checkers and handlers
 
     Over time, we plan on deprecating the checkers and handlers and doing
     everything in the transformers.
 
     The transformers are instances of :class:`PrefilterTransformer` and have
     a single method :meth:`transform` that takes a line and returns a
     transformed line.  The transformation can be accomplished using any
     tool, but our current ones use regular expressions for speed.
 
     After all the transformers have been run, the line is fed to the checkers,
     which are instances of :class:`PrefilterChecker`.  The line is passed to
     the :meth:`check` method, which either returns `None` or a
     :class:`PrefilterHandler` instance.  If `None` is returned, the other
     checkers are tried.  If an :class:`PrefilterHandler` instance is returned,
     the line is passed to the :meth:`handle` method of the returned
     handler and no further checkers are tried.
 
     Both transformers and checkers have a `priority` attribute, that determines
     the order in which they are called.  Smaller priorities are tried first.
 
     Both transformers and checkers also have `enabled` attribute, which is
     a boolean that determines if the instance is used.
 
     Users or developers can change the priority or enabled attribute of
     transformers or checkers, but they must call the :meth:`sort_checkers`
     or :meth:`sort_transformers` method after changing the priority.
     """
 
     multi_line_specials = Bool(True).tag(config=True)
     shell = Instance('IPython.core.interactiveshell.InteractiveShellABC', allow_none=True)
 
     def __init__(self, shell=None, **kwargs):
         super(PrefilterManager, self).__init__(shell=shell, **kwargs)
         self.shell = shell
         self.init_transformers()
         self.init_handlers()
         self.init_checkers()
 
     #-------------------------------------------------------------------------
     # API for managing transformers
     #-------------------------------------------------------------------------
 
     def init_transformers(self):
         """Create the default transformers."""
         self._transformers = []
         for transformer_cls in _default_transformers:
             transformer_cls(
                 shell=self.shell, prefilter_manager=self, parent=self
             )
 
     def sort_transformers(self):
         """Sort the transformers by priority.
 
         This must be called after the priority of a transformer is changed.
         The :meth:`register_transformer` method calls this automatically.
         """
         self._transformers.sort(key=lambda x: x.priority)
 
     @property
     def transformers(self):
         """Return a list of checkers, sorted by priority."""
         return self._transformers
 
     def register_transformer(self, transformer):
         """Register a transformer instance."""
         if transformer not in self._transformers:
             self._transformers.append(transformer)
             self.sort_transformers()
 
     def unregister_transformer(self, transformer):
         """Unregister a transformer instance."""
         if transformer in self._transformers:
             self._transformers.remove(transformer)
 
     #-------------------------------------------------------------------------
     # API for managing checkers
     #-------------------------------------------------------------------------
 
     def init_checkers(self):
         """Create the default checkers."""
         self._checkers = []
         for checker in _default_checkers:
             checker(
                 shell=self.shell, prefilter_manager=self, parent=self
             )
 
     def sort_checkers(self):
         """Sort the checkers by priority.
 
         This must be called after the priority of a checker is changed.
         The :meth:`register_checker` method calls this automatically.
         """
         self._checkers.sort(key=lambda x: x.priority)
 
     @property
     def checkers(self):
         """Return a list of checkers, sorted by priority."""
         return self._checkers
 
     def register_checker(self, checker):
         """Register a checker instance."""
         if checker not in self._checkers:
             self._checkers.append(checker)
             self.sort_checkers()
 
     def unregister_checker(self, checker):
         """Unregister a checker instance."""
         if checker in self._checkers:
             self._checkers.remove(checker)
 
     #-------------------------------------------------------------------------
     # API for managing handlers
     #-------------------------------------------------------------------------
 
     def init_handlers(self):
         """Create the default handlers."""
         self._handlers = {}
         self._esc_handlers = {}
         for handler in _default_handlers:
             handler(
                 shell=self.shell, prefilter_manager=self, parent=self
             )
 
     @property
     def handlers(self):
         """Return a dict of all the handlers."""
         return self._handlers
 
     def register_handler(self, name, handler, esc_strings):
         """Register a handler instance by name with esc_strings."""
         self._handlers[name] = handler
         for esc_str in esc_strings:
             self._esc_handlers[esc_str] = handler
 
     def unregister_handler(self, name, handler, esc_strings):
         """Unregister a handler instance by name with esc_strings."""
         try:
             del self._handlers[name]
         except KeyError:
             pass
         for esc_str in esc_strings:
             h = self._esc_handlers.get(esc_str)
             if h is handler:
                 del self._esc_handlers[esc_str]
 
     def get_handler_by_name(self, name):
         """Get a handler by its name."""
         return self._handlers.get(name)
 
     def get_handler_by_esc(self, esc_str):
         """Get a handler by its escape string."""
         return self._esc_handlers.get(esc_str)
 
     #-------------------------------------------------------------------------
     # Main prefiltering API
     #-------------------------------------------------------------------------
 
     def prefilter_line_info(self, line_info):
         """Prefilter a line that has been converted to a LineInfo object.
 
         This implements the checker/handler part of the prefilter pipe.
         """
         # print "prefilter_line_info: ", line_info
         handler = self.find_handler(line_info)
         return handler.handle(line_info)
 
     def find_handler(self, line_info):
         """Find a handler for the line_info by trying checkers."""
         for checker in self.checkers:
             if checker.enabled:
                 handler = checker.check(line_info)
                 if handler:
                     return handler
         return self.get_handler_by_name('normal')
 
     def transform_line(self, line, continue_prompt):
         """Calls the enabled transformers in order of increasing priority."""
         for transformer in self.transformers:
             if transformer.enabled:
                 line = transformer.transform(line, continue_prompt)
         return line
 
     def prefilter_line(self, line, continue_prompt=False):
         """Prefilter a single input line as text.
 
         This method prefilters a single line of text by calling the
         transformers and then the checkers/handlers.
         """
 
         # print "prefilter_line: ", line, continue_prompt
         # All handlers *must* return a value, even if it's blank ('').
 
         # save the line away in case we crash, so the post-mortem handler can
         # record it
         self.shell._last_input_line = line
 
         if not line:
             # Return immediately on purely empty lines, so that if the user
             # previously typed some whitespace that started a continuation
             # prompt, he can break out of that loop with just an empty line.
             # This is how the default python prompt works.
             return ''
 
         # At this point, we invoke our transformers.
         if not continue_prompt or (continue_prompt and self.multi_line_specials):
             line = self.transform_line(line, continue_prompt)
 
         # Now we compute line_info for the checkers and handlers
         line_info = LineInfo(line, continue_prompt)
 
         # the input history needs to track even empty lines
         stripped = line.strip()
 
         normal_handler = self.get_handler_by_name('normal')
         if not stripped:
             return normal_handler.handle(line_info)
 
         # special handlers are only allowed for single line statements
         if continue_prompt and not self.multi_line_specials:
             return normal_handler.handle(line_info)
 
         prefiltered = self.prefilter_line_info(line_info)
         # print "prefiltered line: %r" % prefiltered
         return prefiltered
 
     def prefilter_lines(self, lines, continue_prompt=False):
         """Prefilter multiple input lines of text.
 
         This is the main entry point for prefiltering multiple lines of
         input.  This simply calls :meth:`prefilter_line` for each line of
         input.
 
         This covers cases where there are multiple lines in the user entry,
         which is the case when the user goes back to a multiline history
         entry and presses enter.
         """
         llines = lines.rstrip('\n').split('\n')
         # We can get multiple lines in one shot, where multiline input 'blends'
         # into one line, in cases like recalling from the readline history
         # buffer.  We need to make sure that in such cases, we correctly
         # communicate downstream which line is first and which are continuation
         # ones.
         if len(llines) > 1:
             out = '\n'.join([self.prefilter_line(line, lnum>0)
                              for lnum, line in enumerate(llines) ])
         else:
             out = self.prefilter_line(llines[0], continue_prompt)
 
         return out
 
 #-----------------------------------------------------------------------------
 # Prefilter transformers
 #-----------------------------------------------------------------------------
 
 
 class PrefilterTransformer(Configurable):
     """Transform a line of user input."""
 
     priority = Integer(100).tag(config=True)
     # Transformers don't currently use shell or prefilter_manager, but as we
     # move away from checkers and handlers, they will need them.
     shell = Instance('IPython.core.interactiveshell.InteractiveShellABC', allow_none=True)
     prefilter_manager = Instance('IPython.core.prefilter.PrefilterManager', allow_none=True)
     enabled = Bool(True).tag(config=True)
 
     def __init__(self, shell=None, prefilter_manager=None, **kwargs):
         super(PrefilterTransformer, self).__init__(
             shell=shell, prefilter_manager=prefilter_manager, **kwargs
         )
         self.prefilter_manager.register_transformer(self)
 
     def transform(self, line, continue_prompt):
         """Transform a line, returning the new one."""
         return None
 
     def __repr__(self):
         return "<%s(priority=%r, enabled=%r)>" % (
             self.__class__.__name__, self.priority, self.enabled)
 
 
 #-----------------------------------------------------------------------------
 # Prefilter checkers
 #-----------------------------------------------------------------------------
 
 
 class PrefilterChecker(Configurable):
     """Inspect an input line and return a handler for that line."""
 
     priority = Integer(100).tag(config=True)
     shell = Instance('IPython.core.interactiveshell.InteractiveShellABC', allow_none=True)
     prefilter_manager = Instance('IPython.core.prefilter.PrefilterManager', allow_none=True)
     enabled = Bool(True).tag(config=True)
 
     def __init__(self, shell=None, prefilter_manager=None, **kwargs):
         super(PrefilterChecker, self).__init__(
             shell=shell, prefilter_manager=prefilter_manager, **kwargs
         )
         self.prefilter_manager.register_checker(self)
 
     def check(self, line_info):
         """Inspect line_info and return a handler instance or None."""
         return None
 
     def __repr__(self):
         return "<%s(priority=%r, enabled=%r)>" % (
             self.__class__.__name__, self.priority, self.enabled)
 
 
 class EmacsChecker(PrefilterChecker):
 
     priority = Integer(100).tag(config=True)
     enabled = Bool(False).tag(config=True)
 
     def check(self, line_info):
         "Emacs ipython-mode tags certain input lines."
         if line_info.line.endswith('# PYTHON-MODE'):
             return self.prefilter_manager.get_handler_by_name('emacs')
         else:
             return None
 
 
 class MacroChecker(PrefilterChecker):
 
     priority = Integer(250).tag(config=True)
 
     def check(self, line_info):
         obj = self.shell.user_ns.get(line_info.ifun)
         if isinstance(obj, Macro):
             return self.prefilter_manager.get_handler_by_name('macro')
         else:
             return None
 
 
 class IPyAutocallChecker(PrefilterChecker):
 
     priority = Integer(300).tag(config=True)
 
     def check(self, line_info):
         "Instances of IPyAutocall in user_ns get autocalled immediately"
         obj = self.shell.user_ns.get(line_info.ifun, None)
         if isinstance(obj, IPyAutocall):
             obj.set_ip(self.shell)
             return self.prefilter_manager.get_handler_by_name('auto')
         else:
             return None
 
 
 class AssignmentChecker(PrefilterChecker):
 
     priority = Integer(600).tag(config=True)
 
     def check(self, line_info):
         """Check to see if user is assigning to a var for the first time, in
         which case we want to avoid any sort of automagic / autocall games.
 
         This allows users to assign to either alias or magic names true python
         variables (the magic/alias systems always take second seat to true
         python code).  E.g. ls='hi', or ls,that=1,2"""
         if line_info.the_rest:
             if line_info.the_rest[0] in '=,':
                 return self.prefilter_manager.get_handler_by_name('normal')
         else:
             return None
 
 
 class AutoMagicChecker(PrefilterChecker):
 
     priority = Integer(700).tag(config=True)
 
     def check(self, line_info):
         """If the ifun is magic, and automagic is on, run it.  Note: normal,
         non-auto magic would already have been triggered via '%' in
         check_esc_chars. This just checks for automagic.  Also, before
         triggering the magic handler, make sure that there is nothing in the
         user namespace which could shadow it."""
         if not self.shell.automagic or not self.shell.find_magic(line_info.ifun):
             return None
 
         # We have a likely magic method.  Make sure we should actually call it.
         if line_info.continue_prompt and not self.prefilter_manager.multi_line_specials:
             return None
 
         head = line_info.ifun.split('.',1)[0]
         if is_shadowed(head, self.shell):
             return None
 
         return self.prefilter_manager.get_handler_by_name('magic')
 
 
 class PythonOpsChecker(PrefilterChecker):
 
     priority = Integer(900).tag(config=True)
 
     def check(self, line_info):
         """If the 'rest' of the line begins with a function call or pretty much
         any python operator, we should simply execute the line (regardless of
         whether or not there's a possible autocall expansion).  This avoids
         spurious (and very confusing) geattr() accesses."""
         if line_info.the_rest and line_info.the_rest[0] in '!=()<>,+*/%^&|':
             return self.prefilter_manager.get_handler_by_name('normal')
         else:
             return None
 
 
 class AutocallChecker(PrefilterChecker):
 
     priority = Integer(1000).tag(config=True)
 
     function_name_regexp = CRegExp(re_fun_name,
         help="RegExp to identify potential function names."
         ).tag(config=True)
     exclude_regexp = CRegExp(re_exclude_auto,
         help="RegExp to exclude strings with this start from autocalling."
         ).tag(config=True)
 
     def check(self, line_info):
         "Check if the initial word/function is callable and autocall is on."
         if not self.shell.autocall:
             return None
 
         oinfo = line_info.ofind(self.shell) # This can mutate state via getattr
         if not oinfo['found']:
             return None
 
         ignored_funs = ['b', 'f', 'r', 'u', 'br', 'rb', 'fr', 'rf']
         ifun = line_info.ifun
         line = line_info.line
         if ifun.lower() in ignored_funs and (line.startswith(ifun + "'") or line.startswith(ifun + '"')):
             return None
 
         if callable(oinfo['obj']) \
                and (not self.exclude_regexp.match(line_info.the_rest)) \
                and self.function_name_regexp.match(line_info.ifun):
             return self.prefilter_manager.get_handler_by_name('auto')
         else:
             return None
 
 
 #-----------------------------------------------------------------------------
 # Prefilter handlers
 #-----------------------------------------------------------------------------
 
 
 class PrefilterHandler(Configurable):
 
     handler_name = Unicode('normal')
     esc_strings = List([])
     shell = Instance('IPython.core.interactiveshell.InteractiveShellABC', allow_none=True)
     prefilter_manager = Instance('IPython.core.prefilter.PrefilterManager', allow_none=True)
 
     def __init__(self, shell=None, prefilter_manager=None, **kwargs):
         super(PrefilterHandler, self).__init__(
             shell=shell, prefilter_manager=prefilter_manager, **kwargs
         )
         self.prefilter_manager.register_handler(
             self.handler_name,
             self,
             self.esc_strings
         )
 
     def handle(self, line_info):
         # print "normal: ", line_info
         """Handle normal input lines. Use as a template for handlers."""
 
         # With autoindent on, we need some way to exit the input loop, and I
         # don't want to force the user to have to backspace all the way to
         # clear the line.  The rule will be in this case, that either two
         # lines of pure whitespace in a row, or a line of pure whitespace but
         # of a size different to the indent level, will exit the input loop.
         line = line_info.line
         continue_prompt = line_info.continue_prompt
 
         if (continue_prompt and
             self.shell.autoindent and
             line.isspace() and
             0 < abs(len(line) - self.shell.indent_current_nsp) <= 2):
             line = ''
 
         return line
 
     def __str__(self):
         return "<%s(name=%s)>" % (self.__class__.__name__, self.handler_name)
 
 
 class MacroHandler(PrefilterHandler):
     handler_name = Unicode("macro")
 
     def handle(self, line_info):
         obj = self.shell.user_ns.get(line_info.ifun)
         pre_space = line_info.pre_whitespace
         line_sep = "\n" + pre_space
         return pre_space + line_sep.join(obj.value.splitlines())
 
 
 class MagicHandler(PrefilterHandler):
 
     handler_name = Unicode('magic')
     esc_strings = List([ESC_MAGIC])
 
     def handle(self, line_info):
         """Execute magic functions."""
         ifun    = line_info.ifun
         the_rest = line_info.the_rest
         #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args)
         t_arg_s = ifun + " " + the_rest
         t_magic_name, _, t_magic_arg_s = t_arg_s.partition(' ')
         t_magic_name = t_magic_name.lstrip(ESC_MAGIC)
         cmd = '%sget_ipython().run_line_magic(%r, %r)' % (line_info.pre_whitespace, t_magic_name, t_magic_arg_s)
         return cmd
 
 
 class AutoHandler(PrefilterHandler):
 
     handler_name = Unicode('auto')
     esc_strings = List([ESC_PAREN, ESC_QUOTE, ESC_QUOTE2])
 
     def handle(self, line_info):
         """Handle lines which can be auto-executed, quoting if requested."""
         line    = line_info.line
         ifun    = line_info.ifun
         the_rest = line_info.the_rest
         esc     = line_info.esc
         continue_prompt = line_info.continue_prompt
         obj = line_info.ofind(self.shell)['obj']
 
         # This should only be active for single-line input!
         if continue_prompt:
             return line
 
         force_auto = isinstance(obj, IPyAutocall)
 
         # User objects sometimes raise exceptions on attribute access other
         # than AttributeError (we've seen it in the past), so it's safest to be
         # ultra-conservative here and catch all.
         try:
             auto_rewrite = obj.rewrite
         except Exception:
             auto_rewrite = True
 
         if esc == ESC_QUOTE:
             # Auto-quote splitting on whitespace
             newcmd = '%s("%s")' % (ifun,'", "'.join(the_rest.split()) )
         elif esc == ESC_QUOTE2:
             # Auto-quote whole string
             newcmd = '%s("%s")' % (ifun,the_rest)
         elif esc == ESC_PAREN:
             newcmd = '%s(%s)' % (ifun,",".join(the_rest.split()))
         else:
             # Auto-paren.
             if force_auto:
                 # Don't rewrite if it is already a call.
                 do_rewrite = not the_rest.startswith('(')
             else:
                 if not the_rest:
                     # We only apply it to argument-less calls if the autocall
                     # parameter is set to 2.
                     do_rewrite = (self.shell.autocall >= 2)
                 elif the_rest.startswith('[') and hasattr(obj, '__getitem__'):
                     # Don't autocall in this case: item access for an object
                     # which is BOTH callable and implements __getitem__.
                     do_rewrite = False
                 else:
                     do_rewrite = True
 
             # Figure out the rewritten command
             if do_rewrite:
                 if the_rest.endswith(';'):
                     newcmd = '%s(%s);' % (ifun.rstrip(),the_rest[:-1])
                 else:
                     newcmd = '%s(%s)' % (ifun.rstrip(), the_rest)
             else:
                 normal_handler = self.prefilter_manager.get_handler_by_name('normal')
                 return normal_handler.handle(line_info)
 
         # Display the rewritten call
         if auto_rewrite:
             self.shell.auto_rewrite_input(newcmd)
 
         return newcmd
 
 
 class EmacsHandler(PrefilterHandler):
 
     handler_name = Unicode('emacs')
     esc_strings = List([])
 
     def handle(self, line_info):
         """Handle input lines marked by python-mode."""
 
         # Currently, nothing is done.  Later more functionality can be added
         # here if needed.
 
         # The input cache shouldn't be updated
         return line_info.line
 
 
 #-----------------------------------------------------------------------------
 # Defaults
 #-----------------------------------------------------------------------------
 
 
 _default_transformers = [
 ]
 
 _default_checkers = [
     EmacsChecker,
     MacroChecker,
     IPyAutocallChecker,
     AssignmentChecker,
     AutoMagicChecker,
     PythonOpsChecker,
     AutocallChecker
 ]
 
 _default_handlers = [
     PrefilterHandler,
     MacroHandler,
     MagicHandler,
     AutoHandler,
     EmacsHandler
 ]

IPython/core/tests/test_inputtransformer.py

@@ -1,494 +1,495 b''
 import tokenize
 import nose.tools as nt
 
 from IPython.testing import tools as tt
 from IPython.utils import py3compat
 u_fmt = py3compat.u_format
 
 from IPython.core import inputtransformer as ipt
 
 def transform_and_reset(transformer):
     transformer = transformer()
     def transform(inp):
         try:
             return transformer.push(inp)
         finally:
             transformer.reset()
     
     return transform
 
 # Transformer tests
 def transform_checker(tests, transformer, **kwargs):
     """Utility to loop over test inputs"""
     transformer = transformer(**kwargs)
     try:
         for inp, tr in tests:
             if inp is None:
                 out = transformer.reset()
             else:
                 out = transformer.push(inp)
             nt.assert_equal(out, tr)
     finally:
         transformer.reset()
 
 # Data for all the syntax tests in the form of lists of pairs of
 # raw/transformed input.  We store it here as a global dict so that we can use
 # it both within single-function tests and also to validate the behavior of the
 # larger objects
 
 syntax = \
   dict(assign_system =
        [(i,py3compat.u_format(o)) for i,o in \
        [(u'a =! ls', "a = get_ipython().getoutput('ls')"),
         (u'b = !ls', "b = get_ipython().getoutput('ls')"),
         (u'c= !ls', "c = get_ipython().getoutput('ls')"),
         (u'd == !ls', u'd == !ls'), # Invalid syntax, but we leave == alone.
         ('x=1', 'x=1'), # normal input is unmodified
         ('    ','    '),  # blank lines are kept intact
         # Tuple unpacking
         (u"a, b = !echo 'a\\nb'", u"a, b = get_ipython().getoutput(\"echo 'a\\\\nb'\")"),
         (u"a,= !echo 'a'", u"a, = get_ipython().getoutput(\"echo 'a'\")"),
         (u"a, *bc = !echo 'a\\nb\\nc'", u"a, *bc = get_ipython().getoutput(\"echo 'a\\\\nb\\\\nc'\")"),
         # Tuple unpacking with regular Python expressions, not our syntax.
         (u"a, b = range(2)", u"a, b = range(2)"),
         (u"a, = range(1)", u"a, = range(1)"),
         (u"a, *bc = range(3)", u"a, *bc = range(3)"),
         ]],
 
        assign_magic =
        [(i,py3compat.u_format(o)) for i,o in \
        [(u'a =% who', "a = get_ipython().run_line_magic('who', '')"),
         (u'b = %who', "b = get_ipython().run_line_magic('who', '')"),
         (u'c= %ls', "c = get_ipython().run_line_magic('ls', '')"),
         (u'd == %ls', u'd == %ls'), # Invalid syntax, but we leave == alone.
         ('x=1', 'x=1'), # normal input is unmodified
         ('    ','    '),  # blank lines are kept intact
         (u"a, b = %foo", u"a, b = get_ipython().run_line_magic('foo', '')"),
         ]],
 
        classic_prompt =
        [('>>> x=1', 'x=1'),
         ('x=1', 'x=1'), # normal input is unmodified
         ('    ', '    '),  # blank lines are kept intact
         ],
 
        ipy_prompt =
        [('In [1]: x=1', 'x=1'),
         ('x=1', 'x=1'), # normal input is unmodified
         ('    ','    '),  # blank lines are kept intact
         ],
 
        # Tests for the escape transformer to leave normal code alone
        escaped_noesc =
        [ ('    ', '    '),
          ('x=1', 'x=1'),
          ],
 
        # System calls
        escaped_shell =
        [(i,py3compat.u_format(o)) for i,o in \
        [ (u'!ls', "get_ipython().system('ls')"),
          # Double-escape shell, this means to capture the output of the
          # subprocess and return it
          (u'!!ls', "get_ipython().getoutput('ls')"),
          ]],
 
        # Help/object info
        escaped_help =
        [(i,py3compat.u_format(o)) for i,o in \
        [ (u'?', 'get_ipython().show_usage()'),
          (u'?x1', "get_ipython().run_line_magic('pinfo', 'x1')"),
          (u'??x2', "get_ipython().run_line_magic('pinfo2', 'x2')"),
          (u'?a.*s', "get_ipython().run_line_magic('psearch', 'a.*s')"),
          (u'?%hist1', "get_ipython().run_line_magic('pinfo', '%hist1')"),
          (u'?%%hist2', "get_ipython().run_line_magic('pinfo', '%%hist2')"),
          (u'?abc = qwe', "get_ipython().run_line_magic('pinfo', 'abc')"),
          ]],
 
       end_help =
       [(i,py3compat.u_format(o)) for i,o in \
       [ (u'x3?', "get_ipython().run_line_magic('pinfo', 'x3')"),
         (u'x4??', "get_ipython().run_line_magic('pinfo2', 'x4')"),
         (u'%hist1?', "get_ipython().run_line_magic('pinfo', '%hist1')"),
         (u'%hist2??', "get_ipython().run_line_magic('pinfo2', '%hist2')"),
         (u'%%hist3?', "get_ipython().run_line_magic('pinfo', '%%hist3')"),
         (u'%%hist4??', "get_ipython().run_line_magic('pinfo2', '%%hist4')"),
+        (u'π.foo?', "get_ipython().run_line_magic('pinfo', 'π.foo')"),
         (u'f*?', "get_ipython().run_line_magic('psearch', 'f*')"),
         (u'ax.*aspe*?', "get_ipython().run_line_magic('psearch', 'ax.*aspe*')"),
         (u'a = abc?', "get_ipython().set_next_input('a = abc');"
                       "get_ipython().run_line_magic('pinfo', 'abc')"),
         (u'a = abc.qe??', "get_ipython().set_next_input('a = abc.qe');"
                           "get_ipython().run_line_magic('pinfo2', 'abc.qe')"),
         (u'a = *.items?', "get_ipython().set_next_input('a = *.items');"
                           "get_ipython().run_line_magic('psearch', '*.items')"),
         (u'plot(a?', "get_ipython().set_next_input('plot(a');"
                      "get_ipython().run_line_magic('pinfo', 'a')"),
         (u'a*2 #comment?', 'a*2 #comment?'),
         ]],
 
        # Explicit magic calls
        escaped_magic =
        [(i,py3compat.u_format(o)) for i,o in \
        [ (u'%cd', "get_ipython().run_line_magic('cd', '')"),
          (u'%cd /home', "get_ipython().run_line_magic('cd', '/home')"),
          # Backslashes need to be escaped.
          (u'%cd C:\\User', "get_ipython().run_line_magic('cd', 'C:\\\\User')"),
          (u'    %magic', "    get_ipython().run_line_magic('magic', '')"),
          ]],
 
        # Quoting with separate arguments
        escaped_quote =
        [ (',f', 'f("")'),
          (',f x', 'f("x")'),
          ('  ,f y', '  f("y")'),
          (',f a b', 'f("a", "b")'),
          ],
 
        # Quoting with single argument
        escaped_quote2 =
        [ (';f', 'f("")'),
          (';f x', 'f("x")'),
          ('  ;f y', '  f("y")'),
          (';f a b', 'f("a b")'),
          ],
 
        # Simply apply parens
        escaped_paren =
        [ ('/f', 'f()'),
          ('/f x', 'f(x)'),
          ('  /f y', '  f(y)'),
          ('/f a b', 'f(a, b)'),
          ],
 
        # Check that we transform prompts before other transforms
        mixed =
        [(i,py3compat.u_format(o)) for i,o in \
        [ (u'In [1]: %lsmagic', "get_ipython().run_line_magic('lsmagic', '')"),
          (u'>>> %lsmagic', "get_ipython().run_line_magic('lsmagic', '')"),
          (u'In [2]: !ls', "get_ipython().system('ls')"),
          (u'In [3]: abs?', "get_ipython().run_line_magic('pinfo', 'abs')"),
          (u'In [4]: b = %who', "b = get_ipython().run_line_magic('who', '')"),
          ]],
        )
 
 # multiline syntax examples.  Each of these should be a list of lists, with
 # each entry itself having pairs of raw/transformed input.  The union (with
 # '\n'.join() of the transformed inputs is what the splitter should produce
 # when fed the raw lines one at a time via push.
 syntax_ml = \
   dict(classic_prompt =
        [ [('>>> for i in range(10):','for i in range(10):'),
           ('...     print i','    print i'),
           ('... ', ''),
           ],
          [('>>> a="""','a="""'),
           ('... 123"""','123"""'),
           ],
          [('a="""','a="""'),
           ('... 123','123'),
           ('... 456"""','456"""'),
           ],
          [('a="""','a="""'),
           ('>>> 123','123'),
           ('... 456"""','456"""'),
           ],
          [('a="""','a="""'),
           ('123','123'),
           ('... 456"""','... 456"""'),
           ],
          [('....__class__','....__class__'),
          ],
          [('a=5', 'a=5'),
           ('...', ''),
          ],
          [('>>> def f(x):', 'def f(x):'),
           ('...', ''),
           ('...     return x', '    return x'),
           ],
          [('board = """....', 'board = """....'),
           ('....', '....'),
           ('...."""', '...."""'),
           ],
         ],
 
        ipy_prompt =
        [ [('In [24]: for i in range(10):','for i in range(10):'),
           ('   ....:     print i','    print i'),
           ('   ....: ', ''),
           ],
          [('In [24]: for i in range(10):','for i in range(10):'),
           # Qt console prompts expand with spaces, not dots
           ('    ...:     print i','    print i'),
           ('    ...: ', ''),
           ],
          [('In [24]: for i in range(10):','for i in range(10):'),
           # Sometimes whitespace preceding '...' has been removed
           ('...:     print i','    print i'),
           ('...: ', ''),
           ],
          [('In [24]: for i in range(10):','for i in range(10):'),
           # Space after last continuation prompt has been removed (issue #6674)
           ('...:     print i','    print i'),
           ('...:', ''),
           ],
          [('In [2]: a="""','a="""'),
           ('   ...: 123"""','123"""'),
           ],
          [('a="""','a="""'),
           ('   ...: 123','123'),
           ('   ...: 456"""','456"""'),
           ],
          [('a="""','a="""'),
           ('In [1]: 123','123'),
           ('   ...: 456"""','456"""'),
           ],
          [('a="""','a="""'),
           ('123','123'),
           ('   ...: 456"""','   ...: 456"""'),
           ],
          ],
 
        multiline_datastructure_prompt =
        [ [('>>> a = [1,','a = [1,'),
           ('... 2]','2]'),
          ],
        ],
         
        multiline_datastructure =
        [ [('b = ("%s"', None),
           ('# comment', None),
           ('%foo )', 'b = ("%s"\n# comment\n%foo )'),
          ],
        ],
        
        multiline_string =
        [ [("'''foo?", None),
           ("bar'''", "'''foo?\nbar'''"),
          ],
        ],
        
        leading_indent =
        [ [('    print "hi"','print "hi"'),
           ],
          [('  for a in range(5):','for a in range(5):'),
           ('    a*2','  a*2'),
           ],
          [('    a="""','a="""'),
           ('    123"""','123"""'),
            ],
          [('a="""','a="""'),
           ('    123"""','    123"""'),
           ],
        ],
        
        cellmagic =
        [ [(u'%%foo a', None),
           (None, u_fmt("get_ipython().run_cell_magic('foo', 'a', '')")),
           ],
          [(u'%%bar 123', None),
           (u'hello', None),
           (None , u_fmt("get_ipython().run_cell_magic('bar', '123', 'hello')")),
           ],
          [(u'a=5', 'a=5'),
           (u'%%cellmagic', '%%cellmagic'),
           ],
        ],
        
        escaped =
        [ [('%abc def \\', None),
           ('ghi', u_fmt("get_ipython().run_line_magic('abc', 'def ghi')")),
           ],
          [('%abc def \\', None),
           ('ghi\\', None),
           (None, u_fmt("get_ipython().run_line_magic('abc', 'def ghi')")),
           ],
        ],
        
        assign_magic =
        [ [(u'a = %bc de \\', None),
           (u'fg', u_fmt("a = get_ipython().run_line_magic('bc', 'de fg')")),
           ],
          [(u'a = %bc de \\', None),
           (u'fg\\', None),
           (None, u_fmt("a = get_ipython().run_line_magic('bc', 'de fg')")),
           ],
        ],
        
        assign_system =
        [ [(u'a = !bc de \\', None),
           (u'fg', u_fmt("a = get_ipython().getoutput('bc de fg')")),
           ],
          [(u'a = !bc de \\', None),
           (u'fg\\', None),
           (None, u_fmt("a = get_ipython().getoutput('bc de fg')")),
           ],
        ],
        )
 
 
 def test_assign_system():
     tt.check_pairs(transform_and_reset(ipt.assign_from_system), syntax['assign_system'])
 
 def test_assign_magic():
     tt.check_pairs(transform_and_reset(ipt.assign_from_magic), syntax['assign_magic'])
 
 def test_classic_prompt():
     tt.check_pairs(transform_and_reset(ipt.classic_prompt), syntax['classic_prompt'])
     for example in syntax_ml['classic_prompt']:
         transform_checker(example, ipt.classic_prompt)
     for example in syntax_ml['multiline_datastructure_prompt']:
         transform_checker(example, ipt.classic_prompt)
 
     # Check that we don't transform the second line if the first is obviously
     # IPython syntax
     transform_checker([
         (u'%foo', '%foo'),
         (u'>>> bar', '>>> bar'),
     ], ipt.classic_prompt)
 
 
 def test_ipy_prompt():
     tt.check_pairs(transform_and_reset(ipt.ipy_prompt), syntax['ipy_prompt'])
     for example in syntax_ml['ipy_prompt']:
         transform_checker(example, ipt.ipy_prompt)
 
     # Check that we don't transform the second line if we're inside a cell magic
     transform_checker([
         (u'%%foo', '%%foo'),
         (u'In [1]: bar', 'In [1]: bar'),
     ], ipt.ipy_prompt)
 
 def test_assemble_logical_lines():
     tests = \
     [ [(u"a = \\", None),
        (u"123", u"a = 123"),
       ],
       [(u"a = \\", None),  # Test resetting when within a multi-line string
        (u"12 *\\", None),
        (None, u"a = 12 *"),
       ],
       [(u"# foo\\", u"# foo\\"), # Comments can't be continued like this
       ],
     ]
     for example in tests:
         transform_checker(example, ipt.assemble_logical_lines)
 
 def test_assemble_python_lines():
     tests = \
     [ [(u"a = '''", None),
        (u"abc'''", u"a = '''\nabc'''"),
       ],
       [(u"a = '''", None),  # Test resetting when within a multi-line string
        (u"def", None),
        (None, u"a = '''\ndef"),
       ],
       [(u"a = [1,", None),
        (u"2]", u"a = [1,\n2]"),
       ],
       [(u"a = [1,", None),  # Test resetting when within a multi-line string
        (u"2,", None),
        (None, u"a = [1,\n2,"),
       ],
       [(u"a = '''", None),  # Test line continuation within a multi-line string
        (u"abc\\", None),
        (u"def", None),
        (u"'''", u"a = '''\nabc\\\ndef\n'''"),
       ],
     ] + syntax_ml['multiline_datastructure']
     for example in tests:
         transform_checker(example, ipt.assemble_python_lines)
 
 
 def test_help_end():
     tt.check_pairs(transform_and_reset(ipt.help_end), syntax['end_help'])
 
 def test_escaped_noesc():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_noesc'])
 
 
 def test_escaped_shell():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_shell'])
 
 
 def test_escaped_help():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_help'])
 
 
 def test_escaped_magic():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_magic'])
 
 
 def test_escaped_quote():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_quote'])
 
 
 def test_escaped_quote2():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_quote2'])
 
 
 def test_escaped_paren():
     tt.check_pairs(transform_and_reset(ipt.escaped_commands), syntax['escaped_paren'])
 
 
 def test_cellmagic():
     for example in syntax_ml['cellmagic']:
         transform_checker(example, ipt.cellmagic)
     
     line_example = [(u'%%bar 123', None),
                     (u'hello', None),
                     (u'' , u_fmt("get_ipython().run_cell_magic('bar', '123', 'hello')")),
                    ]
     transform_checker(line_example, ipt.cellmagic, end_on_blank_line=True)
 
 def test_has_comment():
     tests = [('text', False),
              ('text #comment', True),
              ('text #comment\n', True),
              ('#comment', True),
              ('#comment\n', True),
              ('a = "#string"', False),
              ('a = "#string" # comment', True),
              ('a #comment not "string"', True),
              ]
     tt.check_pairs(ipt.has_comment, tests)
 
 @ipt.TokenInputTransformer.wrap
 def decistmt(tokens):
     """Substitute Decimals for floats in a string of statements.
 
     Based on an example from the tokenize module docs.
     """
     result = []
     for toknum, tokval, _, _, _  in tokens:
         if toknum == tokenize.NUMBER and '.' in tokval:  # replace NUMBER tokens
             for newtok in [
                 (tokenize.NAME, 'Decimal'),
                 (tokenize.OP, '('),
                 (tokenize.STRING, repr(tokval)),
                 (tokenize.OP, ')')
             ]:
                 yield newtok
         else:
             yield (toknum, tokval)
 
 
 
 def test_token_input_transformer():
     tests = [(u'1.2', u_fmt(u"Decimal ('1.2')")),
              (u'"1.2"', u'"1.2"'),
              ]
     tt.check_pairs(transform_and_reset(decistmt), tests)
     ml_tests = \
     [ [(u"a = 1.2; b = '''x", None),
        (u"y'''", u_fmt(u"a =Decimal ('1.2');b ='''x\ny'''")),
       ],
       [(u"a = [1.2,", None),
        (u"3]", u_fmt(u"a =[Decimal ('1.2'),\n3 ]")),
       ],
       [(u"a = '''foo", None),  # Test resetting when within a multi-line string
        (u"bar", None),
        (None, u"a = '''foo\nbar"),
       ],
     ]
     for example in ml_tests:
         transform_checker(example, decistmt)

IPython/core/tests/test_inputtransformer2.py

@@ -1,284 +1,292 b''
 """Tests for the token-based transformers in IPython.core.inputtransformer2
 
 Line-based transformers are the simpler ones; token-based transformers are
 more complex. See test_inputtransformer2_line for tests for line-based
 transformations.
 """
 import nose.tools as nt
 import string
 
 from IPython.core import inputtransformer2 as ipt2
 from IPython.core.inputtransformer2 import make_tokens_by_line, _find_assign_op
 
 from textwrap import dedent
 
 MULTILINE_MAGIC = ("""\
 a = f()
 %foo \\
 bar
 g()
 """.splitlines(keepends=True), (2, 0), """\
 a = f()
 get_ipython().run_line_magic('foo', ' bar')
 g()
 """.splitlines(keepends=True))
 
 INDENTED_MAGIC = ("""\
 for a in range(5):
     %ls
 """.splitlines(keepends=True), (2, 4), """\
 for a in range(5):
     get_ipython().run_line_magic('ls', '')
 """.splitlines(keepends=True))
 
 MULTILINE_MAGIC_ASSIGN = ("""\
 a = f()
 b = %foo \\
   bar
 g()
 """.splitlines(keepends=True), (2, 4), """\
 a = f()
 b = get_ipython().run_line_magic('foo', '   bar')
 g()
 """.splitlines(keepends=True))
 
 MULTILINE_SYSTEM_ASSIGN = ("""\
 a = f()
 b = !foo \\
   bar
 g()
 """.splitlines(keepends=True), (2, 4), """\
 a = f()
 b = get_ipython().getoutput('foo    bar')
 g()
 """.splitlines(keepends=True))
 
 #####
 
 MULTILINE_SYSTEM_ASSIGN_AFTER_DEDENT = ("""\
 def test():
   for i in range(1):
     print(i)
   res =! ls
 """.splitlines(keepends=True), (4, 7), '''\
 def test():
   for i in range(1):
     print(i)
   res =get_ipython().getoutput(\' ls\')
 '''.splitlines(keepends=True))
 
 ######
 
 AUTOCALL_QUOTE = (
     [",f 1 2 3\n"], (1, 0),
     ['f("1", "2", "3")\n']
 )
 
 AUTOCALL_QUOTE2 = (
     [";f 1 2 3\n"], (1, 0),
     ['f("1 2 3")\n']
 )
 
 AUTOCALL_PAREN = (
     ["/f 1 2 3\n"], (1, 0),
     ['f(1, 2, 3)\n']
 )
 
 SIMPLE_HELP = (
     ["foo?\n"], (1, 0),
     ["get_ipython().run_line_magic('pinfo', 'foo')\n"]
 )
 
 DETAILED_HELP = (
     ["foo??\n"], (1, 0),
     ["get_ipython().run_line_magic('pinfo2', 'foo')\n"]
 )
 
 MAGIC_HELP = (
     ["%foo?\n"], (1, 0),
     ["get_ipython().run_line_magic('pinfo', '%foo')\n"]
 )
 
 HELP_IN_EXPR = (
     ["a = b + c?\n"], (1, 0),
     ["get_ipython().set_next_input('a = b + c');"
      "get_ipython().run_line_magic('pinfo', 'c')\n"]
 )
 
 HELP_CONTINUED_LINE = ("""\
 a = \\
 zip?
 """.splitlines(keepends=True), (1, 0),
 [r"get_ipython().set_next_input('a = \\\nzip');get_ipython().run_line_magic('pinfo', 'zip')" + "\n"]
 )
 
 HELP_MULTILINE = ("""\
 (a,
 b) = zip?
 """.splitlines(keepends=True), (1, 0),
 [r"get_ipython().set_next_input('(a,\nb) = zip');get_ipython().run_line_magic('pinfo', 'zip')" + "\n"]
 )
 
+HELP_UNICODE = (
+    ["π.foo?\n"], (1, 0),
+    ["get_ipython().run_line_magic('pinfo', 'π.foo')\n"]
+)
+
 
 def null_cleanup_transformer(lines):
     """
     A cleanup transform that returns an empty list.
     """
     return []
 
 def check_make_token_by_line_never_ends_empty():
     """
     Check that not sequence of single or double characters ends up leading to en empty list of tokens
     """
     from string import printable
     for c in printable:
         nt.assert_not_equal(make_tokens_by_line(c)[-1], [])
         for k in printable:
             nt.assert_not_equal(make_tokens_by_line(c+k)[-1], [])
 
 def check_find(transformer, case, match=True):
     sample, expected_start, _  = case
     tbl = make_tokens_by_line(sample)
     res = transformer.find(tbl)
     if match:
         # start_line is stored 0-indexed, expected values are 1-indexed
         nt.assert_equal((res.start_line+1, res.start_col), expected_start)
         return res
     else:
         nt.assert_is(res, None)
 
 def check_transform(transformer_cls, case):
     lines, start, expected = case
     transformer = transformer_cls(start)
     nt.assert_equal(transformer.transform(lines), expected)
 
 def test_continued_line():
     lines = MULTILINE_MAGIC_ASSIGN[0]
     nt.assert_equal(ipt2.find_end_of_continued_line(lines, 1), 2)
 
     nt.assert_equal(ipt2.assemble_continued_line(lines, (1, 5), 2), "foo    bar")
 
 def test_find_assign_magic():
     check_find(ipt2.MagicAssign, MULTILINE_MAGIC_ASSIGN)
     check_find(ipt2.MagicAssign, MULTILINE_SYSTEM_ASSIGN, match=False)
     check_find(ipt2.MagicAssign, MULTILINE_SYSTEM_ASSIGN_AFTER_DEDENT, match=False)
 
 def test_transform_assign_magic():
     check_transform(ipt2.MagicAssign, MULTILINE_MAGIC_ASSIGN)
 
 def test_find_assign_system():
     check_find(ipt2.SystemAssign, MULTILINE_SYSTEM_ASSIGN)
     check_find(ipt2.SystemAssign, MULTILINE_SYSTEM_ASSIGN_AFTER_DEDENT)
     check_find(ipt2.SystemAssign, (["a =  !ls\n"], (1, 5), None))
     check_find(ipt2.SystemAssign, (["a=!ls\n"], (1, 2), None))
     check_find(ipt2.SystemAssign, MULTILINE_MAGIC_ASSIGN, match=False)
 
 def test_transform_assign_system():
     check_transform(ipt2.SystemAssign, MULTILINE_SYSTEM_ASSIGN)
     check_transform(ipt2.SystemAssign, MULTILINE_SYSTEM_ASSIGN_AFTER_DEDENT)
 
 def test_find_magic_escape():
     check_find(ipt2.EscapedCommand, MULTILINE_MAGIC)
     check_find(ipt2.EscapedCommand, INDENTED_MAGIC)
     check_find(ipt2.EscapedCommand, MULTILINE_MAGIC_ASSIGN, match=False)
 
 def test_transform_magic_escape():
     check_transform(ipt2.EscapedCommand, MULTILINE_MAGIC)
     check_transform(ipt2.EscapedCommand, INDENTED_MAGIC)
 
 def test_find_autocalls():
     for case in [AUTOCALL_QUOTE, AUTOCALL_QUOTE2, AUTOCALL_PAREN]:
         print("Testing %r" % case[0])
         check_find(ipt2.EscapedCommand, case)
 
 def test_transform_autocall():
     for case in [AUTOCALL_QUOTE, AUTOCALL_QUOTE2, AUTOCALL_PAREN]:
         print("Testing %r" % case[0])
         check_transform(ipt2.EscapedCommand, case)
 
 def test_find_help():
     for case in [SIMPLE_HELP, DETAILED_HELP, MAGIC_HELP, HELP_IN_EXPR]:
         check_find(ipt2.HelpEnd, case)
 
     tf = check_find(ipt2.HelpEnd, HELP_CONTINUED_LINE)
     nt.assert_equal(tf.q_line, 1)
     nt.assert_equal(tf.q_col, 3)
 
     tf = check_find(ipt2.HelpEnd, HELP_MULTILINE)
     nt.assert_equal(tf.q_line, 1)
     nt.assert_equal(tf.q_col, 8)
 
     # ? in a comment does not trigger help
     check_find(ipt2.HelpEnd, (["foo # bar?\n"], None, None), match=False)
     # Nor in a string
     check_find(ipt2.HelpEnd, (["foo = '''bar?\n"], None, None), match=False)
 
 def test_transform_help():
     tf = ipt2.HelpEnd((1, 0), (1, 9))
     nt.assert_equal(tf.transform(HELP_IN_EXPR[0]), HELP_IN_EXPR[2])
 
     tf = ipt2.HelpEnd((1, 0), (2, 3))
     nt.assert_equal(tf.transform(HELP_CONTINUED_LINE[0]), HELP_CONTINUED_LINE[2])
 
     tf = ipt2.HelpEnd((1, 0), (2, 8))
     nt.assert_equal(tf.transform(HELP_MULTILINE[0]), HELP_MULTILINE[2])
 
+    tf = ipt2.HelpEnd((1, 0), (1, 0))
+    nt.assert_equal(tf.transform(HELP_UNICODE[0]), HELP_UNICODE[2])
+
 def test_find_assign_op_dedent():
     """
     be careful that empty token like dedent are not counted as parens
     """
     class Tk:
         def __init__(self, s):
             self.string = s
 
     nt.assert_equal(_find_assign_op([Tk(s) for s in ('','a','=','b')]), 2)
     nt.assert_equal(_find_assign_op([Tk(s) for s in ('','(', 'a','=','b', ')', '=' ,'5')]), 6)
 
 def test_check_complete():
     cc = ipt2.TransformerManager().check_complete
     nt.assert_equal(cc("a = 1"), ('complete', None))
     nt.assert_equal(cc("for a in range(5):"), ('incomplete', 4))
     nt.assert_equal(cc("for a in range(5):\n    if a > 0:"), ('incomplete', 8))
     nt.assert_equal(cc("raise = 2"), ('invalid', None))
     nt.assert_equal(cc("a = [1,\n2,"), ('incomplete', 0))
     nt.assert_equal(cc(")"), ('incomplete', 0))
     nt.assert_equal(cc("\\\r\n"), ('incomplete', 0))
     nt.assert_equal(cc("a = '''\n   hi"), ('incomplete', 3))
     nt.assert_equal(cc("def a():\n x=1\n global x"), ('invalid', None))
     nt.assert_equal(cc("a \\ "), ('invalid', None))  # Nothing allowed after backslash
     nt.assert_equal(cc("1\\\n+2"), ('complete', None))
     nt.assert_equal(cc("exit"), ('complete', None))
 
     example = dedent("""
         if True:
             a=1""" )
 
     nt.assert_equal(cc(example), ('incomplete', 4))
     nt.assert_equal(cc(example+'\n'), ('complete', None))
     nt.assert_equal(cc(example+'\n    '), ('complete', None))
 
     # no need to loop on all the letters/numbers.
     short = '12abAB'+string.printable[62:]
     for c in short:
         # test does not raise:
         cc(c)
         for k in short:
             cc(c+k)
 
     nt.assert_equal(cc("def f():\n  x=0\n  \\\n  "), ('incomplete', 2))
 
 def test_check_complete_II():
     """
     Test that multiple line strings are properly handled.
 
     Separate test function for convenience
 
     """
     cc = ipt2.TransformerManager().check_complete
     nt.assert_equal(cc('''def foo():\n    """'''), ('incomplete', 4))
 
 
 def test_null_cleanup_transformer():
     manager = ipt2.TransformerManager()
     manager.cleanup_transforms.insert(0, null_cleanup_transformer)
     nt.assert_is(manager.transform_cell(""), "")

IPython/core/tests/test_magic.py

@@ -1,1223 +1,1225 b''
 # -*- coding: utf-8 -*-
 """Tests for various magic functions.
 
 Needs to be run by nose (to make ipython session available).
 """
 
 import io
 import os
 import re
 import sys
 import warnings
 from textwrap import dedent
 from unittest import TestCase
 from unittest import mock
 from importlib import invalidate_caches
 from io import StringIO
 
 import nose.tools as nt
 
 import shlex
 
 from IPython import get_ipython
 from IPython.core import magic
 from IPython.core.error import UsageError
 from IPython.core.magic import (Magics, magics_class, line_magic,
                                 cell_magic,
                                 register_line_magic, register_cell_magic)
 from IPython.core.magics import execution, script, code, logging, osm
 from IPython.testing import decorators as dec
 from IPython.testing import tools as tt
 from IPython.utils.io import capture_output
 from IPython.utils.tempdir import (TemporaryDirectory,
                                     TemporaryWorkingDirectory)
 from IPython.utils.process import find_cmd
 
 
 @magic.magics_class
 class DummyMagics(magic.Magics): pass
 
 def test_extract_code_ranges():
     instr = "1 3 5-6 7-9 10:15 17: :10 10- -13 :"
     expected = [(0, 1), 
                 (2, 3),
                 (4, 6),
                 (6, 9),
                 (9, 14),
                 (16, None),
                 (None, 9),
                 (9, None),
                 (None, 13),
                 (None, None)]
     actual = list(code.extract_code_ranges(instr))
     nt.assert_equal(actual, expected)
 
 def test_extract_symbols():
     source = """import foo\na = 10\ndef b():\n    return 42\n\n\nclass A: pass\n\n\n"""
     symbols_args = ["a", "b", "A", "A,b", "A,a", "z"]
     expected = [([], ['a']),
                 (["def b():\n    return 42\n"], []),
                 (["class A: pass\n"], []),
                 (["class A: pass\n", "def b():\n    return 42\n"], []),
                 (["class A: pass\n"], ['a']),
                 ([], ['z'])]
     for symbols, exp in zip(symbols_args, expected):
         nt.assert_equal(code.extract_symbols(source, symbols), exp)
 
 
 def test_extract_symbols_raises_exception_with_non_python_code():
     source = ("=begin A Ruby program :)=end\n"
               "def hello\n"
               "puts 'Hello world'\n"
               "end")
     with nt.assert_raises(SyntaxError):
         code.extract_symbols(source, "hello")
 
 
 def test_magic_not_found():
     # magic not found raises UsageError
     with nt.assert_raises(UsageError):
         _ip.magic('doesntexist')
 
     # ensure result isn't success when a magic isn't found
     result = _ip.run_cell('%doesntexist')
     assert isinstance(result.error_in_exec, UsageError)
 
 
 def test_cell_magic_not_found():
     # magic not found raises UsageError
     with nt.assert_raises(UsageError):
         _ip.run_cell_magic('doesntexist', 'line', 'cell')
 
     # ensure result isn't success when a magic isn't found
     result = _ip.run_cell('%%doesntexist')
     assert isinstance(result.error_in_exec, UsageError)
 
 
 def test_magic_error_status():
     def fail(shell):
         1/0
     _ip.register_magic_function(fail)
     result = _ip.run_cell('%fail')
     assert isinstance(result.error_in_exec, ZeroDivisionError)
 
 
 def test_config():
     """ test that config magic does not raise
     can happen if Configurable init is moved too early into
     Magics.__init__ as then a Config object will be registered as a
     magic.
     """
     ## should not raise.
     _ip.magic('config')
 
 def test_config_available_configs():
     """ test that config magic prints available configs in unique and
     sorted order. """
     with capture_output() as captured:
         _ip.magic('config')
 
     stdout = captured.stdout
     config_classes = stdout.strip().split('\n')[1:]
     nt.assert_list_equal(config_classes, sorted(set(config_classes)))
 
 def test_config_print_class():
     """ test that config with a classname prints the class's options. """
     with capture_output() as captured:
         _ip.magic('config TerminalInteractiveShell')
 
     stdout = captured.stdout
     if not re.match("TerminalInteractiveShell.* options", stdout.splitlines()[0]):
         print(stdout)
         raise AssertionError("1st line of stdout not like "
                              "'TerminalInteractiveShell.* options'")
 
 def test_rehashx():
     # clear up everything
     _ip.alias_manager.clear_aliases()
     del _ip.db['syscmdlist']
     
     _ip.magic('rehashx')
     # Practically ALL ipython development systems will have more than 10 aliases
 
     nt.assert_true(len(_ip.alias_manager.aliases) > 10)
     for name, cmd in _ip.alias_manager.aliases:
         # we must strip dots from alias names
         nt.assert_not_in('.', name)
 
     # rehashx must fill up syscmdlist
     scoms = _ip.db['syscmdlist']
     nt.assert_true(len(scoms) > 10)
             
 
 
 def test_magic_parse_options():
     """Test that we don't mangle paths when parsing magic options."""
     ip = get_ipython()
     path = 'c:\\x'
     m = DummyMagics(ip)
     opts = m.parse_options('-f %s' % path,'f:')[0]
     # argv splitting is os-dependent
     if os.name == 'posix':
         expected = 'c:x'
     else:
         expected = path
     nt.assert_equal(opts['f'], expected)
 
 def test_magic_parse_long_options():
     """Magic.parse_options can handle --foo=bar long options"""
     ip = get_ipython()
     m = DummyMagics(ip)
     opts, _ = m.parse_options('--foo --bar=bubble', 'a', 'foo', 'bar=')
     nt.assert_in('foo', opts)
     nt.assert_in('bar', opts)
     nt.assert_equal(opts['bar'], "bubble")
 
 
 @dec.skip_without('sqlite3')
 def doctest_hist_f():
     """Test %hist -f with temporary filename.
 
     In [9]: import tempfile
 
     In [10]: tfile = tempfile.mktemp('.py','tmp-ipython-')
 
     In [11]: %hist -nl -f $tfile 3
 
     In [13]: import os; os.unlink(tfile)
     """
 
 
 @dec.skip_without('sqlite3')
 def doctest_hist_r():
     """Test %hist -r
 
     XXX - This test is not recording the output correctly.  For some reason, in
     testing mode the raw history isn't getting populated.  No idea why.
     Disabling the output checking for now, though at least we do run it.
 
     In [1]: 'hist' in _ip.lsmagic()
     Out[1]: True
 
     In [2]: x=1
 
     In [3]: %hist -rl 2
     x=1 # random
     %hist -r 2
     """
 
 
 @dec.skip_without('sqlite3')
 def doctest_hist_op():
     """Test %hist -op
 
     In [1]: class b(float):
        ...:     pass
        ...: 
 
     In [2]: class s(object):
        ...:     def __str__(self):
        ...:         return 's'
        ...: 
 
     In [3]: 
 
     In [4]: class r(b):
        ...:     def __repr__(self):
        ...:         return 'r'
        ...: 
 
     In [5]: class sr(s,r): pass
        ...: 
 
     In [6]: 
 
     In [7]: bb=b()
 
     In [8]: ss=s()
 
     In [9]: rr=r()
 
     In [10]: ssrr=sr()
 
     In [11]: 4.5
     Out[11]: 4.5
 
     In [12]: str(ss)
     Out[12]: 's'
 
     In [13]: 
 
     In [14]: %hist -op
     >>> class b:
     ...     pass
     ... 
     >>> class s(b):
     ...     def __str__(self):
     ...         return 's'
     ... 
     >>> 
     >>> class r(b):
     ...     def __repr__(self):
     ...         return 'r'
     ... 
     >>> class sr(s,r): pass
     >>> 
     >>> bb=b()
     >>> ss=s()
     >>> rr=r()
     >>> ssrr=sr()
     >>> 4.5
     4.5
     >>> str(ss)
     's'
     >>> 
     """
 
 def test_hist_pof():
     ip = get_ipython()
     ip.run_cell(u"1+2", store_history=True)
     #raise Exception(ip.history_manager.session_number)
     #raise Exception(list(ip.history_manager._get_range_session()))
     with TemporaryDirectory() as td:
         tf = os.path.join(td, 'hist.py')
         ip.run_line_magic('history', '-pof %s' % tf)
         assert os.path.isfile(tf)
 
 
 @dec.skip_without('sqlite3')
 def test_macro():
     ip = get_ipython()
     ip.history_manager.reset()   # Clear any existing history.
     cmds = ["a=1", "def b():\n  return a**2", "print(a,b())"]
     for i, cmd in enumerate(cmds, start=1):
         ip.history_manager.store_inputs(i, cmd)
     ip.magic("macro test 1-3")
     nt.assert_equal(ip.user_ns["test"].value, "\n".join(cmds)+"\n")
     
     # List macros
     nt.assert_in("test", ip.magic("macro"))
 
 
 @dec.skip_without('sqlite3')
 def test_macro_run():
     """Test that we can run a multi-line macro successfully."""
     ip = get_ipython()
     ip.history_manager.reset()
     cmds = ["a=10", "a+=1", "print(a)", "%macro test 2-3"]
     for cmd in cmds:
         ip.run_cell(cmd, store_history=True)
     nt.assert_equal(ip.user_ns["test"].value, "a+=1\nprint(a)\n")
     with tt.AssertPrints("12"):
         ip.run_cell("test")
     with tt.AssertPrints("13"):
         ip.run_cell("test")
 
 
 def test_magic_magic():
     """Test %magic"""
     ip = get_ipython()
     with capture_output() as captured:
         ip.magic("magic")
     
     stdout = captured.stdout
     nt.assert_in('%magic', stdout)
     nt.assert_in('IPython', stdout)
     nt.assert_in('Available', stdout)
 
 
 @dec.skipif_not_numpy
 def test_numpy_reset_array_undec():
     "Test '%reset array' functionality"
     _ip.ex('import numpy as np')
     _ip.ex('a = np.empty(2)')
     nt.assert_in('a', _ip.user_ns)
     _ip.magic('reset -f array')
     nt.assert_not_in('a', _ip.user_ns)
 
 def test_reset_out():
     "Test '%reset out' magic"
     _ip.run_cell("parrot = 'dead'", store_history=True)
     # test '%reset -f out', make an Out prompt
     _ip.run_cell("parrot", store_history=True)
     nt.assert_true('dead' in [_ip.user_ns[x] for x in ('_','__','___')])
     _ip.magic('reset -f out')
     nt.assert_false('dead' in [_ip.user_ns[x] for x in ('_','__','___')])
     nt.assert_equal(len(_ip.user_ns['Out']), 0)
 
 def test_reset_in():
     "Test '%reset in' magic"
     # test '%reset -f in'
     _ip.run_cell("parrot", store_history=True)
     nt.assert_true('parrot' in [_ip.user_ns[x] for x in ('_i','_ii','_iii')])
     _ip.magic('%reset -f in')
     nt.assert_false('parrot' in [_ip.user_ns[x] for x in ('_i','_ii','_iii')])
     nt.assert_equal(len(set(_ip.user_ns['In'])), 1)
 
 def test_reset_dhist():
     "Test '%reset dhist' magic"
     _ip.run_cell("tmp = [d for d in _dh]") # copy before clearing
     _ip.magic('cd ' + os.path.dirname(nt.__file__))
     _ip.magic('cd -')
     nt.assert_true(len(_ip.user_ns['_dh']) > 0)
     _ip.magic('reset -f dhist')
     nt.assert_equal(len(_ip.user_ns['_dh']), 0)
     _ip.run_cell("_dh = [d for d in tmp]") #restore
 
 def test_reset_in_length():
     "Test that '%reset in' preserves In[] length"
     _ip.run_cell("print 'foo'")
     _ip.run_cell("reset -f in")
     nt.assert_equal(len(_ip.user_ns['In']), _ip.displayhook.prompt_count+1)
 
 class TestResetErrors(TestCase):
 
     def test_reset_redefine(self):
 
         @magics_class
         class KernelMagics(Magics):
               @line_magic
               def less(self, shell): pass
 
         _ip.register_magics(KernelMagics)
 
         with self.assertLogs() as cm:
             # hack, we want to just capture logs, but assertLogs fails if not
             # logs get produce.
             # so log one things we ignore.
             import logging as log_mod
             log = log_mod.getLogger()
             log.info('Nothing')
             # end hack.
             _ip.run_cell("reset -f")
 
         assert len(cm.output) == 1
         for out in cm.output:
             assert "Invalid alias" not in out
 
 def test_tb_syntaxerror():
     """test %tb after a SyntaxError"""
     ip = get_ipython()
     ip.run_cell("for")
     
     # trap and validate stdout
     save_stdout = sys.stdout
     try:
         sys.stdout = StringIO()
         ip.run_cell("%tb")
         out = sys.stdout.getvalue()
     finally:
         sys.stdout = save_stdout
     # trim output, and only check the last line
     last_line = out.rstrip().splitlines()[-1].strip()
     nt.assert_equal(last_line, "SyntaxError: invalid syntax")
 
 
 def test_time():
     ip = get_ipython()
     
     with tt.AssertPrints("Wall time: "):
         ip.run_cell("%time None")
     
     ip.run_cell("def f(kmjy):\n"
                 "    %time print (2*kmjy)")
     
     with tt.AssertPrints("Wall time: "):
         with tt.AssertPrints("hihi", suppress=False):
             ip.run_cell("f('hi')")
 
 def test_time_last_not_expression():
     ip.run_cell("%%time\n"
                 "var_1 = 1\n"
                 "var_2 = 2\n")
     assert ip.user_ns['var_1'] == 1
     del ip.user_ns['var_1']
     assert ip.user_ns['var_2'] == 2
     del ip.user_ns['var_2']
     
 
 @dec.skip_win32
 def test_time2():
     ip = get_ipython()
     
     with tt.AssertPrints("CPU times: user "):
         ip.run_cell("%time None")
 
 def test_time3():
     """Erroneous magic function calls, issue gh-3334"""
     ip = get_ipython()
     ip.user_ns.pop('run', None)
     
     with tt.AssertNotPrints("not found", channel='stderr'):
         ip.run_cell("%%time\n"
                     "run = 0\n"
                     "run += 1")
 
 def test_multiline_time():
     """Make sure last statement from time return a value."""
     ip = get_ipython()
     ip.user_ns.pop('run', None)
 
     ip.run_cell(dedent("""\
         %%time
         a = "ho"
         b = "hey"
         a+b
         """))
     nt.assert_equal(ip.user_ns_hidden['_'], 'hohey')
 
 def test_time_local_ns():
     """
     Test that local_ns is actually global_ns when running a cell magic
     """
     ip = get_ipython()
     ip.run_cell("%%time\n"
                 "myvar = 1")
     nt.assert_equal(ip.user_ns['myvar'], 1)
     del ip.user_ns['myvar']
 
 def test_doctest_mode():
     "Toggle doctest_mode twice, it should be a no-op and run without error"
     _ip.magic('doctest_mode')
     _ip.magic('doctest_mode')
 
 
 def test_parse_options():
     """Tests for basic options parsing in magics."""
     # These are only the most minimal of tests, more should be added later.  At
     # the very least we check that basic text/unicode calls work OK.
     m = DummyMagics(_ip)
     nt.assert_equal(m.parse_options('foo', '')[1], 'foo')
     nt.assert_equal(m.parse_options(u'foo', '')[1], u'foo')
 
 
 def test_dirops():
     """Test various directory handling operations."""
     # curpath = lambda :os.path.splitdrive(os.getcwd())[1].replace('\\','/')
     curpath = os.getcwd
     startdir = os.getcwd()
     ipdir = os.path.realpath(_ip.ipython_dir)
     try:
         _ip.magic('cd "%s"' % ipdir)
         nt.assert_equal(curpath(), ipdir)
         _ip.magic('cd -')
         nt.assert_equal(curpath(), startdir)
         _ip.magic('pushd "%s"' % ipdir)
         nt.assert_equal(curpath(), ipdir)
         _ip.magic('popd')
         nt.assert_equal(curpath(), startdir)
     finally:
         os.chdir(startdir)
 
 
 def test_cd_force_quiet():
     """Test OSMagics.cd_force_quiet option"""
     _ip.config.OSMagics.cd_force_quiet = True
     osmagics = osm.OSMagics(shell=_ip)
 
     startdir = os.getcwd()
     ipdir = os.path.realpath(_ip.ipython_dir)
 
     try:
         with tt.AssertNotPrints(ipdir):
             osmagics.cd('"%s"' % ipdir)
         with tt.AssertNotPrints(startdir):
             osmagics.cd('-')
     finally:
         os.chdir(startdir)
 
 
 def test_xmode():
     # Calling xmode three times should be a no-op
     xmode = _ip.InteractiveTB.mode
     for i in range(4):
         _ip.magic("xmode")
     nt.assert_equal(_ip.InteractiveTB.mode, xmode)
     
 def test_reset_hard():
     monitor = []
     class A(object):
         def __del__(self):
             monitor.append(1)
         def __repr__(self):
             return "<A instance>"
             
     _ip.user_ns["a"] = A()
     _ip.run_cell("a")
     
     nt.assert_equal(monitor, [])
     _ip.magic("reset -f")
     nt.assert_equal(monitor, [1])
     
 class TestXdel(tt.TempFileMixin):
     def test_xdel(self):
         """Test that references from %run are cleared by xdel."""
         src = ("class A(object):\n"
                "    monitor = []\n"
                "    def __del__(self):\n"
                "        self.monitor.append(1)\n"
                "a = A()\n")
         self.mktmp(src)
         # %run creates some hidden references...
         _ip.magic("run %s" % self.fname)
         # ... as does the displayhook.
         _ip.run_cell("a")
         
         monitor = _ip.user_ns["A"].monitor
         nt.assert_equal(monitor, [])
         
         _ip.magic("xdel a")
         
         # Check that a's __del__ method has been called.
         nt.assert_equal(monitor, [1])
 
 def doctest_who():
     """doctest for %who
     
     In [1]: %reset -f
     
     In [2]: alpha = 123
     
     In [3]: beta = 'beta'
     
     In [4]: %who int
     alpha
     
     In [5]: %who str
     beta
     
     In [6]: %whos
     Variable   Type    Data/Info
     ----------------------------
     alpha      int     123
     beta       str     beta
     
     In [7]: %who_ls
     Out[7]: ['alpha', 'beta']
     """
 
 def test_whos():
     """Check that whos is protected against objects where repr() fails."""
     class A(object):
         def __repr__(self):
             raise Exception()
     _ip.user_ns['a'] = A()
     _ip.magic("whos")
 
 def doctest_precision():
     """doctest for %precision
     
     In [1]: f = get_ipython().display_formatter.formatters['text/plain']
     
     In [2]: %precision 5
     Out[2]: '%.5f'
     
     In [3]: f.float_format
     Out[3]: '%.5f'
     
     In [4]: %precision %e
     Out[4]: '%e'
     
     In [5]: f(3.1415927)
     Out[5]: '3.141593e+00'
     """
 
 def test_psearch():
     with tt.AssertPrints("dict.fromkeys"):
         _ip.run_cell("dict.fr*?")
+    with tt.AssertPrints("π.is_integer"):
+        _ip.run_cell("π = 3.14;\nπ.is_integ*?")
 
 def test_timeit_shlex():
     """test shlex issues with timeit (#1109)"""
     _ip.ex("def f(*a,**kw): pass")
     _ip.magic('timeit -n1 "this is a bug".count(" ")')
     _ip.magic('timeit -r1 -n1 f(" ", 1)')
     _ip.magic('timeit -r1 -n1 f(" ", 1, " ", 2, " ")')
     _ip.magic('timeit -r1 -n1 ("a " + "b")')
     _ip.magic('timeit -r1 -n1 f("a " + "b")')
     _ip.magic('timeit -r1 -n1 f("a " + "b ")')
 
 
 def test_timeit_special_syntax():
     "Test %%timeit with IPython special syntax"
     @register_line_magic
     def lmagic(line):
         ip = get_ipython()
         ip.user_ns['lmagic_out'] = line
 
     # line mode test
     _ip.run_line_magic('timeit', '-n1 -r1 %lmagic my line')
     nt.assert_equal(_ip.user_ns['lmagic_out'], 'my line')
     # cell mode test
     _ip.run_cell_magic('timeit', '-n1 -r1', '%lmagic my line2')
     nt.assert_equal(_ip.user_ns['lmagic_out'], 'my line2')
 
 def test_timeit_return():
     """
     test whether timeit -o return object
     """
 
     res = _ip.run_line_magic('timeit','-n10 -r10 -o 1')
     assert(res is not None)
 
 def test_timeit_quiet():
     """
     test quiet option of timeit magic
     """
     with tt.AssertNotPrints("loops"):
         _ip.run_cell("%timeit -n1 -r1 -q 1")
 
 def test_timeit_return_quiet():
     with tt.AssertNotPrints("loops"):
         res = _ip.run_line_magic('timeit', '-n1 -r1 -q -o 1')
     assert (res is not None)
 
 def test_timeit_invalid_return():
     with nt.assert_raises_regex(SyntaxError, "outside function"):
         _ip.run_line_magic('timeit', 'return')
 
 @dec.skipif(execution.profile is None)
 def test_prun_special_syntax():
     "Test %%prun with IPython special syntax"
     @register_line_magic
     def lmagic(line):
         ip = get_ipython()
         ip.user_ns['lmagic_out'] = line
 
     # line mode test
     _ip.run_line_magic('prun', '-q %lmagic my line')
     nt.assert_equal(_ip.user_ns['lmagic_out'], 'my line')
     # cell mode test
     _ip.run_cell_magic('prun', '-q', '%lmagic my line2')
     nt.assert_equal(_ip.user_ns['lmagic_out'], 'my line2')
 
 @dec.skipif(execution.profile is None)
 def test_prun_quotes():
     "Test that prun does not clobber string escapes (GH #1302)"
     _ip.magic(r"prun -q x = '\t'")
     nt.assert_equal(_ip.user_ns['x'], '\t')
 
 def test_extension():
     # Debugging information for failures of this test
     print('sys.path:')
     for p in sys.path:
         print(' ', p)
     print('CWD', os.getcwd())
 
     nt.assert_raises(ImportError, _ip.magic, "load_ext daft_extension")
     daft_path = os.path.join(os.path.dirname(__file__), "daft_extension")
     sys.path.insert(0, daft_path)
     try:
         _ip.user_ns.pop('arq', None)
         invalidate_caches()   # Clear import caches
         _ip.magic("load_ext daft_extension")
         nt.assert_equal(_ip.user_ns['arq'], 185)
         _ip.magic("unload_ext daft_extension")
         assert 'arq' not in _ip.user_ns
     finally:
         sys.path.remove(daft_path)
 
 
 def test_notebook_export_json():
     _ip = get_ipython()
     _ip.history_manager.reset()   # Clear any existing history.
     cmds = [u"a=1", u"def b():\n  return a**2", u"print('noël, été', b())"]
     for i, cmd in enumerate(cmds, start=1):
         _ip.history_manager.store_inputs(i, cmd)
     with TemporaryDirectory() as td:
         outfile = os.path.join(td, "nb.ipynb")
         _ip.magic("notebook -e %s" % outfile)
 
 
 class TestEnv(TestCase):
 
     def test_env(self):
         env = _ip.magic("env")
         self.assertTrue(isinstance(env, dict))
 
     def test_env_secret(self):
         env = _ip.magic("env")
         hidden = "<hidden>"
         with mock.patch.dict(
             os.environ,
             {
                 "API_KEY": "abc123",
                 "SECRET_THING": "ssshhh",
                 "JUPYTER_TOKEN": "",
                 "VAR": "abc"
             }
         ):
             env = _ip.magic("env")
         assert env["API_KEY"] == hidden
         assert env["SECRET_THING"] == hidden
         assert env["JUPYTER_TOKEN"] == hidden
         assert env["VAR"] == "abc"
 
     def test_env_get_set_simple(self):
         env = _ip.magic("env var val1")
         self.assertEqual(env, None)
         self.assertEqual(os.environ['var'], 'val1')
         self.assertEqual(_ip.magic("env var"), 'val1')
         env = _ip.magic("env var=val2")
         self.assertEqual(env, None)
         self.assertEqual(os.environ['var'], 'val2')
 
     def test_env_get_set_complex(self):
         env = _ip.magic("env var 'val1 '' 'val2")
         self.assertEqual(env, None)
         self.assertEqual(os.environ['var'], "'val1 '' 'val2")
         self.assertEqual(_ip.magic("env var"), "'val1 '' 'val2")
         env = _ip.magic('env var=val2 val3="val4')
         self.assertEqual(env, None)
         self.assertEqual(os.environ['var'], 'val2 val3="val4')
 
     def test_env_set_bad_input(self):
         self.assertRaises(UsageError, lambda: _ip.magic("set_env var"))
 
     def test_env_set_whitespace(self):
         self.assertRaises(UsageError, lambda: _ip.magic("env var A=B"))
 
 
 class CellMagicTestCase(TestCase):
 
     def check_ident(self, magic):
         # Manually called, we get the result
         out = _ip.run_cell_magic(magic, 'a', 'b')
         nt.assert_equal(out, ('a','b'))
         # Via run_cell, it goes into the user's namespace via displayhook
         _ip.run_cell('%%' + magic +' c\nd\n')
         nt.assert_equal(_ip.user_ns['_'], ('c','d\n'))
 
     def test_cell_magic_func_deco(self):
         "Cell magic using simple decorator"
         @register_cell_magic
         def cellm(line, cell):
             return line, cell
 
         self.check_ident('cellm')
 
     def test_cell_magic_reg(self):
         "Cell magic manually registered"
         def cellm(line, cell):
             return line, cell
 
         _ip.register_magic_function(cellm, 'cell', 'cellm2')
         self.check_ident('cellm2')
 
     def test_cell_magic_class(self):
         "Cell magics declared via a class"
         @magics_class
         class MyMagics(Magics):
 
             @cell_magic
             def cellm3(self, line, cell):
                 return line, cell
 
         _ip.register_magics(MyMagics)
         self.check_ident('cellm3')
 
     def test_cell_magic_class2(self):
         "Cell magics declared via a class, #2"
         @magics_class
         class MyMagics2(Magics):
 
             @cell_magic('cellm4')
             def cellm33(self, line, cell):
                 return line, cell
             
         _ip.register_magics(MyMagics2)
         self.check_ident('cellm4')
         # Check that nothing is registered as 'cellm33'
         c33 = _ip.find_cell_magic('cellm33')
         nt.assert_equal(c33, None)
 
 def test_file():
     """Basic %%writefile"""
     ip = get_ipython()
     with TemporaryDirectory() as td:
         fname = os.path.join(td, 'file1')
         ip.run_cell_magic("writefile", fname, u'\n'.join([
             'line1',
             'line2',
         ]))
         with open(fname) as f:
             s = f.read()
         nt.assert_in('line1\n', s)
         nt.assert_in('line2', s)
 
 @dec.skip_win32
 def test_file_single_quote():
     """Basic %%writefile with embedded single quotes"""
     ip = get_ipython()
     with TemporaryDirectory() as td:
         fname = os.path.join(td, '\'file1\'')
         ip.run_cell_magic("writefile", fname, u'\n'.join([
             'line1',
             'line2',
         ]))
         with open(fname) as f:
             s = f.read()
         nt.assert_in('line1\n', s)
         nt.assert_in('line2', s)
 
 @dec.skip_win32
 def test_file_double_quote():
     """Basic %%writefile with embedded double quotes"""
     ip = get_ipython()
     with TemporaryDirectory() as td:
         fname = os.path.join(td, '"file1"')
         ip.run_cell_magic("writefile", fname, u'\n'.join([
             'line1',
             'line2',
         ]))
         with open(fname) as f:
             s = f.read()
         nt.assert_in('line1\n', s)
         nt.assert_in('line2', s)
 
 def test_file_var_expand():
     """%%writefile $filename"""
     ip = get_ipython()
     with TemporaryDirectory() as td:
         fname = os.path.join(td, 'file1')
         ip.user_ns['filename'] = fname
         ip.run_cell_magic("writefile", '$filename', u'\n'.join([
             'line1',
             'line2',
         ]))
         with open(fname) as f:
             s = f.read()
         nt.assert_in('line1\n', s)
         nt.assert_in('line2', s)
 
 def test_file_unicode():
     """%%writefile with unicode cell"""
     ip = get_ipython()
     with TemporaryDirectory() as td:
         fname = os.path.join(td, 'file1')
         ip.run_cell_magic("writefile", fname, u'\n'.join([
             u'liné1',
             u'liné2',
         ]))
         with io.open(fname, encoding='utf-8') as f:
             s = f.read()
         nt.assert_in(u'liné1\n', s)
         nt.assert_in(u'liné2', s)
 
 def test_file_amend():
     """%%writefile -a amends files"""
     ip = get_ipython()
     with TemporaryDirectory() as td:
         fname = os.path.join(td, 'file2')
         ip.run_cell_magic("writefile", fname, u'\n'.join([
             'line1',
             'line2',
         ]))
         ip.run_cell_magic("writefile", "-a %s" % fname, u'\n'.join([
             'line3',
             'line4',
         ]))
         with open(fname) as f:
             s = f.read()
         nt.assert_in('line1\n', s)
         nt.assert_in('line3\n', s)
 
 def test_file_spaces():
     """%%file with spaces in filename"""
     ip = get_ipython()
     with TemporaryWorkingDirectory() as td:
         fname = "file name"
         ip.run_cell_magic("file", '"%s"'%fname, u'\n'.join([
             'line1',
             'line2',
         ]))
         with open(fname) as f:
             s = f.read()
         nt.assert_in('line1\n', s)
         nt.assert_in('line2', s)
     
 def test_script_config():
     ip = get_ipython()
     ip.config.ScriptMagics.script_magics = ['whoda']
     sm = script.ScriptMagics(shell=ip)
     nt.assert_in('whoda', sm.magics['cell'])
 
 @dec.skip_win32
 def test_script_out():
     ip = get_ipython()
     ip.run_cell_magic("script", "--out output sh", "echo 'hi'")
     nt.assert_equal(ip.user_ns['output'], 'hi\n')
 
 @dec.skip_win32
 def test_script_err():
     ip = get_ipython()
     ip.run_cell_magic("script", "--err error sh", "echo 'hello' >&2")
     nt.assert_equal(ip.user_ns['error'], 'hello\n')
 
 @dec.skip_win32
 def test_script_out_err():
     ip = get_ipython()
     ip.run_cell_magic("script", "--out output --err error sh", "echo 'hi'\necho 'hello' >&2")
     nt.assert_equal(ip.user_ns['output'], 'hi\n')
     nt.assert_equal(ip.user_ns['error'], 'hello\n')
 
 @dec.skip_win32
 def test_script_bg_out():
     ip = get_ipython()
     ip.run_cell_magic("script", "--bg --out output sh", "echo 'hi'")
 
     nt.assert_equal(ip.user_ns['output'].read(), b'hi\n')
     ip.user_ns['output'].close()
 
 @dec.skip_win32
 def test_script_bg_err():
     ip = get_ipython()
     ip.run_cell_magic("script", "--bg --err error sh", "echo 'hello' >&2")
     nt.assert_equal(ip.user_ns['error'].read(), b'hello\n')
     ip.user_ns['error'].close()
 
 @dec.skip_win32
 def test_script_bg_out_err():
     ip = get_ipython()
     ip.run_cell_magic("script", "--bg --out output --err error sh", "echo 'hi'\necho 'hello' >&2")
     nt.assert_equal(ip.user_ns['output'].read(), b'hi\n')
     nt.assert_equal(ip.user_ns['error'].read(), b'hello\n')
     ip.user_ns['output'].close()
     ip.user_ns['error'].close()
 
 def test_script_defaults():
     ip = get_ipython()
     for cmd in ['sh', 'bash', 'perl', 'ruby']:
         try:
             find_cmd(cmd)
         except Exception:
             pass
         else:
             nt.assert_in(cmd, ip.magics_manager.magics['cell'])
 
 
 @magics_class
 class FooFoo(Magics):
     """class with both %foo and %%foo magics"""
     @line_magic('foo')
     def line_foo(self, line):
         "I am line foo"
         pass
 
     @cell_magic("foo")
     def cell_foo(self, line, cell):
         "I am cell foo, not line foo"
         pass
 
 def test_line_cell_info():
     """%%foo and %foo magics are distinguishable to inspect"""
     ip = get_ipython()
     ip.magics_manager.register(FooFoo)
     oinfo = ip.object_inspect('foo')
     nt.assert_true(oinfo['found'])
     nt.assert_true(oinfo['ismagic'])
     
     oinfo = ip.object_inspect('%%foo')
     nt.assert_true(oinfo['found'])
     nt.assert_true(oinfo['ismagic'])
     nt.assert_equal(oinfo['docstring'], FooFoo.cell_foo.__doc__)
 
     oinfo = ip.object_inspect('%foo')
     nt.assert_true(oinfo['found'])
     nt.assert_true(oinfo['ismagic'])
     nt.assert_equal(oinfo['docstring'], FooFoo.line_foo.__doc__)
 
 def test_multiple_magics():
     ip = get_ipython()
     foo1 = FooFoo(ip)
     foo2 = FooFoo(ip)
     mm = ip.magics_manager
     mm.register(foo1)
     nt.assert_true(mm.magics['line']['foo'].__self__ is foo1)
     mm.register(foo2)
     nt.assert_true(mm.magics['line']['foo'].__self__ is foo2)
 
 def test_alias_magic():
     """Test %alias_magic."""
     ip = get_ipython()
     mm = ip.magics_manager
 
     # Basic operation: both cell and line magics are created, if possible.
     ip.run_line_magic('alias_magic', 'timeit_alias timeit')
     nt.assert_in('timeit_alias', mm.magics['line'])
     nt.assert_in('timeit_alias', mm.magics['cell'])
 
     # --cell is specified, line magic not created.
     ip.run_line_magic('alias_magic', '--cell timeit_cell_alias timeit')
     nt.assert_not_in('timeit_cell_alias', mm.magics['line'])
     nt.assert_in('timeit_cell_alias', mm.magics['cell'])
 
     # Test that line alias is created successfully.
     ip.run_line_magic('alias_magic', '--line env_alias env')
     nt.assert_equal(ip.run_line_magic('env', ''),
                     ip.run_line_magic('env_alias', ''))
 
     # Test that line alias with parameters passed in is created successfully.
     ip.run_line_magic('alias_magic', '--line history_alias history --params ' + shlex.quote('3'))
     nt.assert_in('history_alias', mm.magics['line'])
 
 
 def test_save():
     """Test %save."""
     ip = get_ipython()
     ip.history_manager.reset()   # Clear any existing history.
     cmds = [u"a=1", u"def b():\n  return a**2", u"print(a, b())"]
     for i, cmd in enumerate(cmds, start=1):
         ip.history_manager.store_inputs(i, cmd)
     with TemporaryDirectory() as tmpdir:
         file = os.path.join(tmpdir, "testsave.py")
         ip.run_line_magic("save", "%s 1-10" % file)
         with open(file) as f:
             content = f.read()
             nt.assert_equal(content.count(cmds[0]), 1)
             nt.assert_in('coding: utf-8', content)
         ip.run_line_magic("save", "-a %s 1-10" % file)
         with open(file) as f:
             content = f.read()
             nt.assert_equal(content.count(cmds[0]), 2)
             nt.assert_in('coding: utf-8', content)
 
 
 def test_store():
     """Test %store."""
     ip = get_ipython()
     ip.run_line_magic('load_ext', 'storemagic')
     
     # make sure the storage is empty
     ip.run_line_magic('store', '-z')
     ip.user_ns['var'] = 42
     ip.run_line_magic('store', 'var')
     ip.user_ns['var'] = 39
     ip.run_line_magic('store', '-r')
     nt.assert_equal(ip.user_ns['var'], 42)
 
     ip.run_line_magic('store', '-d var')
     ip.user_ns['var'] = 39
     ip.run_line_magic('store' , '-r')
     nt.assert_equal(ip.user_ns['var'], 39)
 
 
 def _run_edit_test(arg_s, exp_filename=None,
                         exp_lineno=-1,
                         exp_contents=None,
                         exp_is_temp=None):
     ip = get_ipython()
     M = code.CodeMagics(ip)
     last_call = ['','']
     opts,args = M.parse_options(arg_s,'prxn:')
     filename, lineno, is_temp = M._find_edit_target(ip, args, opts, last_call)
         
     if exp_filename is not None:
         nt.assert_equal(exp_filename, filename)
     if exp_contents is not None:
         with io.open(filename, 'r', encoding='utf-8') as f:
             contents = f.read()
         nt.assert_equal(exp_contents, contents)
     if exp_lineno != -1:
         nt.assert_equal(exp_lineno, lineno)
     if exp_is_temp is not None:
         nt.assert_equal(exp_is_temp, is_temp)
 
 
 def test_edit_interactive():
     """%edit on interactively defined objects"""
     ip = get_ipython()
     n = ip.execution_count
     ip.run_cell(u"def foo(): return 1", store_history=True)
     
     try:
         _run_edit_test("foo")
     except code.InteractivelyDefined as e:
         nt.assert_equal(e.index, n)
     else:
         raise AssertionError("Should have raised InteractivelyDefined")
 
 
 def test_edit_cell():
     """%edit [cell id]"""
     ip = get_ipython()
     
     ip.run_cell(u"def foo(): return 1", store_history=True)
     
     # test
     _run_edit_test("1", exp_contents=ip.user_ns['In'][1], exp_is_temp=True)
 
 def test_bookmark():
     ip = get_ipython()
     ip.run_line_magic('bookmark', 'bmname')
     with tt.AssertPrints('bmname'):
         ip.run_line_magic('bookmark', '-l')
     ip.run_line_magic('bookmark', '-d bmname')
 
 def test_ls_magic():
     ip = get_ipython()
     json_formatter = ip.display_formatter.formatters['application/json']
     json_formatter.enabled = True
     lsmagic = ip.magic('lsmagic')
     with warnings.catch_warnings(record=True) as w:
         j = json_formatter(lsmagic)
     nt.assert_equal(sorted(j), ['cell', 'line'])
     nt.assert_equal(w, []) # no warnings
 
 def test_strip_initial_indent():
     def sii(s):
         lines = s.splitlines()
         return '\n'.join(code.strip_initial_indent(lines))
 
     nt.assert_equal(sii("  a = 1\nb = 2"), "a = 1\nb = 2")
     nt.assert_equal(sii("  a\n    b\nc"), "a\n  b\nc")
     nt.assert_equal(sii("a\n  b"), "a\n  b")
 
 def test_logging_magic_quiet_from_arg():
     _ip.config.LoggingMagics.quiet = False
     lm = logging.LoggingMagics(shell=_ip)
     with TemporaryDirectory() as td:
         try:
             with tt.AssertNotPrints(re.compile("Activating.*")):
                 lm.logstart('-q {}'.format(
                         os.path.join(td, "quiet_from_arg.log")))
         finally:
             _ip.logger.logstop()
 
 def test_logging_magic_quiet_from_config():
     _ip.config.LoggingMagics.quiet = True
     lm = logging.LoggingMagics(shell=_ip)
     with TemporaryDirectory() as td:
         try:
             with tt.AssertNotPrints(re.compile("Activating.*")):
                 lm.logstart(os.path.join(td, "quiet_from_config.log"))
         finally:
             _ip.logger.logstop()
 
 
 def test_logging_magic_not_quiet():
     _ip.config.LoggingMagics.quiet = False
     lm = logging.LoggingMagics(shell=_ip)
     with TemporaryDirectory() as td:
         try:
             with tt.AssertPrints(re.compile("Activating.*")):
                 lm.logstart(os.path.join(td, "not_quiet.log"))
         finally:
             _ip.logger.logstop()
 
 
 def test_time_no_var_expand():
     _ip.user_ns['a'] = 5
     _ip.user_ns['b'] = []
     _ip.magic('time b.append("{a}")')
     assert _ip.user_ns['b'] == ['{a}']
 
 
 # this is slow, put at the end for local testing.
 def test_timeit_arguments():
     "Test valid timeit arguments, should not cause SyntaxError (GH #1269)"
     if sys.version_info < (3,7):
         _ip.magic("timeit -n1 -r1 ('#')")
     else:
         # 3.7 optimize no-op statement like above out, and complain there is
         # nothing in the for loop.
         _ip.magic("timeit -n1 -r1 a=('#')")

IPython/core/tests/test_prefilter.py

@@ -1,117 +1,127 b''
 """Tests for input manipulation machinery."""
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 import nose.tools as nt
 
 from IPython.core.prefilter import AutocallChecker
 
 #-----------------------------------------------------------------------------
 # Tests
 #-----------------------------------------------------------------------------
 
 def test_prefilter():
     """Test user input conversions"""
 
     # pairs of (raw, expected correct) input
     pairs = [ ('2+2','2+2'),
              ]
 
     for raw, correct in pairs:
         nt.assert_equal(ip.prefilter(raw), correct)
 
 def test_prefilter_shadowed():
     def dummy_magic(line): pass
 
     prev_automagic_state = ip.automagic
     ip.automagic = True
     ip.autocall = 0
 
     try:
         # These should not be transformed - they are shadowed by other names
         for name in ['if', 'zip', 'get_ipython']: # keyword, builtin, global
             ip.register_magic_function(dummy_magic, magic_name=name)
             res = ip.prefilter(name+' foo')
             nt.assert_equal(res, name+' foo')
             del ip.magics_manager.magics['line'][name]
 
         # These should be transformed
         for name in ['fi', 'piz', 'nohtypi_teg']:
             ip.register_magic_function(dummy_magic, magic_name=name)
             res = ip.prefilter(name+' foo')
             nt.assert_not_equal(res, name+' foo')
             del ip.magics_manager.magics['line'][name]
 
     finally:
         ip.automagic = prev_automagic_state
 
 def test_autocall_binops():
     """See https://github.com/ipython/ipython/issues/81"""
     ip.magic('autocall 2')
     f = lambda x: x
     ip.user_ns['f'] = f
     try:
         nt.assert_equal(ip.prefilter('f 1'),'f(1)')
         for t in ['f +1', 'f -1']:
             nt.assert_equal(ip.prefilter(t), t)
 
         # Run tests again with a more permissive exclude_regexp, which will
         # allow transformation of binary operations ('f -1' -> 'f(-1)').
         pm = ip.prefilter_manager
         ac = AutocallChecker(shell=pm.shell, prefilter_manager=pm,
                              config=pm.config)
         try:
             ac.priority = 1
             ac.exclude_regexp = r'^[,&^\|\*/]|^is |^not |^in |^and |^or '
             pm.sort_checkers()
 
             nt.assert_equal(ip.prefilter('f -1'), 'f(-1)')
             nt.assert_equal(ip.prefilter('f +1'), 'f(+1)')
         finally:
             pm.unregister_checker(ac)
     finally:
         ip.magic('autocall 0')
         del ip.user_ns['f']
 
 
 def test_issue_114():
     """Check that multiline string literals don't expand as magic
     see http://github.com/ipython/ipython/issues/114"""
 
     template = '"""\n%s\n"""'
     # Store the current value of multi_line_specials and turn it off before
     # running test, since it could be true (case in which the test doesn't make
     # sense, as multiline string literals *will* expand as magic in that case).
     msp = ip.prefilter_manager.multi_line_specials
     ip.prefilter_manager.multi_line_specials = False
     try:
         for mgk in ip.magics_manager.lsmagic()['line']:
             raw = template % mgk
             nt.assert_equal(ip.prefilter(raw), raw)
     finally:
         ip.prefilter_manager.multi_line_specials = msp
 
 
 def test_prefilter_attribute_errors():
     """Capture exceptions thrown by user objects on attribute access.
 
     See http://github.com/ipython/ipython/issues/988."""
 
     class X(object):
         def __getattr__(self, k):
             raise ValueError('broken object')
         def __call__(self, x):
             return x
 
     # Create a callable broken object
     ip.user_ns['x'] = X()
     ip.magic('autocall 2')
     try:
         # Even if x throws an attribute error when looking at its rewrite
         # attribute, we should not crash.  So the test here is simply making
         # the prefilter call and not having an exception.
         ip.prefilter('x 1')
     finally:
         del ip.user_ns['x']
         ip.magic('autocall 0')
+
+
+def test_autocall_should_support_unicode():
+    ip.magic('autocall 2')
+    ip.user_ns['π'] = lambda x: x
+    try:
+        nt.assert_equal(ip.prefilter('π 3'),'π(3)')
+    finally:
+        ip.magic('autocall 0')
+        del ip.user_ns['π']