Show More
@@ -1,2060 +1,2060 b'' | |||
|
1 | 1 | """Completion for IPython. |
|
2 | 2 | |
|
3 | 3 | This module started as fork of the rlcompleter module in the Python standard |
|
4 | 4 | library. The original enhancements made to rlcompleter have been sent |
|
5 | 5 | upstream and were accepted as of Python 2.3, |
|
6 | 6 | |
|
7 | 7 | This module now support a wide variety of completion mechanism both available |
|
8 | 8 | for normal classic Python code, as well as completer for IPython specific |
|
9 | 9 | Syntax like magics. |
|
10 | 10 | |
|
11 | 11 | Latex and Unicode completion |
|
12 | 12 | ============================ |
|
13 | 13 | |
|
14 | 14 | IPython and compatible frontends not only can complete your code, but can help |
|
15 | 15 | you to input a wide range of characters. In particular we allow you to insert |
|
16 | 16 | a unicode character using the tab completion mechanism. |
|
17 | 17 | |
|
18 | 18 | Forward latex/unicode completion |
|
19 | 19 | -------------------------------- |
|
20 | 20 | |
|
21 | 21 | Forward completion allows you to easily type a unicode character using its latex |
|
22 | 22 | name, or unicode long description. To do so type a backslash follow by the |
|
23 | 23 | relevant name and press tab: |
|
24 | 24 | |
|
25 | 25 | |
|
26 | 26 | Using latex completion: |
|
27 | 27 | |
|
28 | 28 | .. code:: |
|
29 | 29 | |
|
30 | 30 | \\alpha<tab> |
|
31 | 31 | Ξ± |
|
32 | 32 | |
|
33 | 33 | or using unicode completion: |
|
34 | 34 | |
|
35 | 35 | |
|
36 | 36 | .. code:: |
|
37 | 37 | |
|
38 | 38 | \\greek small letter alpha<tab> |
|
39 | 39 | Ξ± |
|
40 | 40 | |
|
41 | 41 | |
|
42 | 42 | Only valid Python identifiers will complete. Combining characters (like arrow or |
|
43 | 43 | dots) are also available, unlike latex they need to be put after the their |
|
44 | 44 | counterpart that is to say, `F\\\\vec<tab>` is correct, not `\\\\vec<tab>F`. |
|
45 | 45 | |
|
46 | 46 | Some browsers are known to display combining characters incorrectly. |
|
47 | 47 | |
|
48 | 48 | Backward latex completion |
|
49 | 49 | ------------------------- |
|
50 | 50 | |
|
51 | 51 | It is sometime challenging to know how to type a character, if you are using |
|
52 | 52 | IPython, or any compatible frontend you can prepend backslash to the character |
|
53 | 53 | and press `<tab>` to expand it to its latex form. |
|
54 | 54 | |
|
55 | 55 | .. code:: |
|
56 | 56 | |
|
57 | 57 | \\Ξ±<tab> |
|
58 | 58 | \\alpha |
|
59 | 59 | |
|
60 | 60 | |
|
61 | 61 | Both forward and backward completions can be deactivated by setting the |
|
62 | 62 | ``Completer.backslash_combining_completions`` option to ``False``. |
|
63 | 63 | |
|
64 | 64 | |
|
65 | 65 | Experimental |
|
66 | 66 | ============ |
|
67 | 67 | |
|
68 | 68 | Starting with IPython 6.0, this module can make use of the Jedi library to |
|
69 | 69 | generate completions both using static analysis of the code, and dynamically |
|
70 | 70 | inspecting multiple namespaces. The APIs attached to this new mechanism is |
|
71 | 71 | unstable and will raise unless use in an :any:`provisionalcompleter` context |
|
72 | 72 | manager. |
|
73 | 73 | |
|
74 | 74 | You will find that the following are experimental: |
|
75 | 75 | |
|
76 | 76 | - :any:`provisionalcompleter` |
|
77 | 77 | - :any:`IPCompleter.completions` |
|
78 | 78 | - :any:`Completion` |
|
79 | 79 | - :any:`rectify_completions` |
|
80 | 80 | |
|
81 | 81 | .. note:: |
|
82 | 82 | |
|
83 | 83 | better name for :any:`rectify_completions` ? |
|
84 | 84 | |
|
85 | 85 | We welcome any feedback on these new API, and we also encourage you to try this |
|
86 | 86 | module in debug mode (start IPython with ``--Completer.debug=True``) in order |
|
87 | 87 | to have extra logging information is :any:`jedi` is crashing, or if current |
|
88 | 88 | IPython completer pending deprecations are returning results not yet handled |
|
89 | 89 | by :any:`jedi` |
|
90 | 90 | |
|
91 | 91 | Using Jedi for tab completion allow snippets like the following to work without |
|
92 | 92 | having to execute any code: |
|
93 | 93 | |
|
94 | 94 | >>> myvar = ['hello', 42] |
|
95 | 95 | ... myvar[1].bi<tab> |
|
96 | 96 | |
|
97 | 97 | Tab completion will be able to infer that ``myvar[1]`` is a real number without |
|
98 | 98 | executing any code unlike the previously available ``IPCompleter.greedy`` |
|
99 | 99 | option. |
|
100 | 100 | |
|
101 | 101 | Be sure to update :any:`jedi` to the latest stable version or to try the |
|
102 | 102 | current development version to get better completions. |
|
103 | 103 | """ |
|
104 | 104 | |
|
105 | 105 | |
|
106 | 106 | # Copyright (c) IPython Development Team. |
|
107 | 107 | # Distributed under the terms of the Modified BSD License. |
|
108 | 108 | # |
|
109 | 109 | # Some of this code originated from rlcompleter in the Python standard library |
|
110 | 110 | # Copyright (C) 2001 Python Software Foundation, www.python.org |
|
111 | 111 | |
|
112 | 112 | |
|
113 | 113 | import __main__ |
|
114 | 114 | import builtins as builtin_mod |
|
115 | 115 | import glob |
|
116 | 116 | import time |
|
117 | 117 | import inspect |
|
118 | 118 | import itertools |
|
119 | 119 | import keyword |
|
120 | 120 | import os |
|
121 | 121 | import re |
|
122 | 122 | import sys |
|
123 | 123 | import unicodedata |
|
124 | 124 | import string |
|
125 | 125 | import warnings |
|
126 | 126 | |
|
127 | 127 | from contextlib import contextmanager |
|
128 | 128 | from importlib import import_module |
|
129 | 129 | from typing import Iterator, List, Tuple, Iterable, Union |
|
130 | 130 | from types import SimpleNamespace |
|
131 | 131 | |
|
132 | 132 | from traitlets.config.configurable import Configurable |
|
133 | 133 | from IPython.core.error import TryNext |
|
134 | 134 | from IPython.core.inputsplitter import ESC_MAGIC |
|
135 | 135 | from IPython.core.latex_symbols import latex_symbols, reverse_latex_symbol |
|
136 | 136 | from IPython.core.oinspect import InspectColors |
|
137 | 137 | from IPython.utils import generics |
|
138 | 138 | from IPython.utils.dir2 import dir2, get_real_method |
|
139 | 139 | from IPython.utils.process import arg_split |
|
140 | 140 | from traitlets import Bool, Enum, observe, Int |
|
141 | 141 | |
|
142 | 142 | # skip module docstests |
|
143 | 143 | skip_doctest = True |
|
144 | 144 | |
|
145 | 145 | try: |
|
146 | 146 | import jedi |
|
147 | 147 | jedi.settings.case_insensitive_completion = False |
|
148 | 148 | import jedi.api.helpers |
|
149 | 149 | import jedi.api.classes |
|
150 | 150 | JEDI_INSTALLED = True |
|
151 | 151 | except ImportError: |
|
152 | 152 | JEDI_INSTALLED = False |
|
153 | 153 | #----------------------------------------------------------------------------- |
|
154 | 154 | # Globals |
|
155 | 155 | #----------------------------------------------------------------------------- |
|
156 | 156 | |
|
157 | 157 | # Public API |
|
158 | 158 | __all__ = ['Completer','IPCompleter'] |
|
159 | 159 | |
|
160 | 160 | if sys.platform == 'win32': |
|
161 | 161 | PROTECTABLES = ' ' |
|
162 | 162 | else: |
|
163 | 163 | PROTECTABLES = ' ()[]{}?=\\|;:\'#*"^&' |
|
164 | 164 | |
|
165 | 165 | # Protect against returning an enormous number of completions which the frontend |
|
166 | 166 | # may have trouble processing. |
|
167 | 167 | MATCHES_LIMIT = 500 |
|
168 | 168 | |
|
169 | 169 | _deprecation_readline_sentinel = object() |
|
170 | 170 | |
|
171 | 171 | |
|
172 | 172 | class ProvisionalCompleterWarning(FutureWarning): |
|
173 | 173 | """ |
|
174 | 174 | Exception raise by an experimental feature in this module. |
|
175 | 175 | |
|
176 | 176 | Wrap code in :any:`provisionalcompleter` context manager if you |
|
177 | 177 | are certain you want to use an unstable feature. |
|
178 | 178 | """ |
|
179 | 179 | pass |
|
180 | 180 | |
|
181 | 181 | warnings.filterwarnings('error', category=ProvisionalCompleterWarning) |
|
182 | 182 | |
|
183 | 183 | @contextmanager |
|
184 | 184 | def provisionalcompleter(action='ignore'): |
|
185 | 185 | """ |
|
186 | 186 | |
|
187 | 187 | |
|
188 | 188 | This contest manager has to be used in any place where unstable completer |
|
189 | 189 | behavior and API may be called. |
|
190 | 190 | |
|
191 | 191 | >>> with provisionalcompleter(): |
|
192 | 192 | ... completer.do_experimetal_things() # works |
|
193 | 193 | |
|
194 | 194 | >>> completer.do_experimental_things() # raises. |
|
195 | 195 | |
|
196 | 196 | .. note:: Unstable |
|
197 | 197 | |
|
198 | 198 | By using this context manager you agree that the API in use may change |
|
199 | 199 | without warning, and that you won't complain if they do so. |
|
200 | 200 | |
|
201 | 201 | You also understand that if the API is not to you liking you should report |
|
202 | 202 | a bug to explain your use case upstream and improve the API and will loose |
|
203 | 203 | credibility if you complain after the API is make stable. |
|
204 | 204 | |
|
205 | 205 | We'll be happy to get your feedback , feature request and improvement on |
|
206 | 206 | any of the unstable APIs ! |
|
207 | 207 | """ |
|
208 | 208 | with warnings.catch_warnings(): |
|
209 | 209 | warnings.filterwarnings(action, category=ProvisionalCompleterWarning) |
|
210 | 210 | yield |
|
211 | 211 | |
|
212 | 212 | |
|
213 | 213 | def has_open_quotes(s): |
|
214 | 214 | """Return whether a string has open quotes. |
|
215 | 215 | |
|
216 | 216 | This simply counts whether the number of quote characters of either type in |
|
217 | 217 | the string is odd. |
|
218 | 218 | |
|
219 | 219 | Returns |
|
220 | 220 | ------- |
|
221 | 221 | If there is an open quote, the quote character is returned. Else, return |
|
222 | 222 | False. |
|
223 | 223 | """ |
|
224 | 224 | # We check " first, then ', so complex cases with nested quotes will get |
|
225 | 225 | # the " to take precedence. |
|
226 | 226 | if s.count('"') % 2: |
|
227 | 227 | return '"' |
|
228 | 228 | elif s.count("'") % 2: |
|
229 | 229 | return "'" |
|
230 | 230 | else: |
|
231 | 231 | return False |
|
232 | 232 | |
|
233 | 233 | |
|
234 | 234 | def protect_filename(s, protectables=PROTECTABLES): |
|
235 | 235 | """Escape a string to protect certain characters.""" |
|
236 | 236 | if set(s) & set(protectables): |
|
237 | 237 | if sys.platform == "win32": |
|
238 | 238 | return '"' + s + '"' |
|
239 | 239 | else: |
|
240 | 240 | return "".join(("\\" + c if c in protectables else c) for c in s) |
|
241 | 241 | else: |
|
242 | 242 | return s |
|
243 | 243 | |
|
244 | 244 | |
|
245 | 245 | def expand_user(path:str) -> Tuple[str, bool, str]: |
|
246 | 246 | """Expand ``~``-style usernames in strings. |
|
247 | 247 | |
|
248 | 248 | This is similar to :func:`os.path.expanduser`, but it computes and returns |
|
249 | 249 | extra information that will be useful if the input was being used in |
|
250 | 250 | computing completions, and you wish to return the completions with the |
|
251 | 251 | original '~' instead of its expanded value. |
|
252 | 252 | |
|
253 | 253 | Parameters |
|
254 | 254 | ---------- |
|
255 | 255 | path : str |
|
256 | 256 | String to be expanded. If no ~ is present, the output is the same as the |
|
257 | 257 | input. |
|
258 | 258 | |
|
259 | 259 | Returns |
|
260 | 260 | ------- |
|
261 | 261 | newpath : str |
|
262 | 262 | Result of ~ expansion in the input path. |
|
263 | 263 | tilde_expand : bool |
|
264 | 264 | Whether any expansion was performed or not. |
|
265 | 265 | tilde_val : str |
|
266 | 266 | The value that ~ was replaced with. |
|
267 | 267 | """ |
|
268 | 268 | # Default values |
|
269 | 269 | tilde_expand = False |
|
270 | 270 | tilde_val = '' |
|
271 | 271 | newpath = path |
|
272 | 272 | |
|
273 | 273 | if path.startswith('~'): |
|
274 | 274 | tilde_expand = True |
|
275 | 275 | rest = len(path)-1 |
|
276 | 276 | newpath = os.path.expanduser(path) |
|
277 | 277 | if rest: |
|
278 | 278 | tilde_val = newpath[:-rest] |
|
279 | 279 | else: |
|
280 | 280 | tilde_val = newpath |
|
281 | 281 | |
|
282 | 282 | return newpath, tilde_expand, tilde_val |
|
283 | 283 | |
|
284 | 284 | |
|
285 | 285 | def compress_user(path:str, tilde_expand:bool, tilde_val:str) -> str: |
|
286 | 286 | """Does the opposite of expand_user, with its outputs. |
|
287 | 287 | """ |
|
288 | 288 | if tilde_expand: |
|
289 | 289 | return path.replace(tilde_val, '~') |
|
290 | 290 | else: |
|
291 | 291 | return path |
|
292 | 292 | |
|
293 | 293 | |
|
294 | 294 | def completions_sorting_key(word): |
|
295 | 295 | """key for sorting completions |
|
296 | 296 | |
|
297 | 297 | This does several things: |
|
298 | 298 | |
|
299 | 299 | - Demote any completions starting with underscores to the end |
|
300 | 300 | - Insert any %magic and %%cellmagic completions in the alphabetical order |
|
301 | 301 | by their name |
|
302 | 302 | """ |
|
303 | 303 | prio1, prio2 = 0, 0 |
|
304 | 304 | |
|
305 | 305 | if word.startswith('__'): |
|
306 | 306 | prio1 = 2 |
|
307 | 307 | elif word.startswith('_'): |
|
308 | 308 | prio1 = 1 |
|
309 | 309 | |
|
310 | 310 | if word.endswith('='): |
|
311 | 311 | prio1 = -1 |
|
312 | 312 | |
|
313 | 313 | if word.startswith('%%'): |
|
314 | 314 | # If there's another % in there, this is something else, so leave it alone |
|
315 | 315 | if not "%" in word[2:]: |
|
316 | 316 | word = word[2:] |
|
317 | 317 | prio2 = 2 |
|
318 | 318 | elif word.startswith('%'): |
|
319 | 319 | if not "%" in word[1:]: |
|
320 | 320 | word = word[1:] |
|
321 | 321 | prio2 = 1 |
|
322 | 322 | |
|
323 | 323 | return prio1, word, prio2 |
|
324 | 324 | |
|
325 | 325 | |
|
326 | 326 | class _FakeJediCompletion: |
|
327 | 327 | """ |
|
328 | 328 | This is a workaround to communicate to the UI that Jedi has crashed and to |
|
329 | 329 | report a bug. Will be used only id :any:`IPCompleter.debug` is set to true. |
|
330 | 330 | |
|
331 | 331 | Added in IPython 6.0 so should likely be removed for 7.0 |
|
332 | 332 | |
|
333 | 333 | """ |
|
334 | 334 | |
|
335 | 335 | def __init__(self, name): |
|
336 | 336 | |
|
337 | 337 | self.name = name |
|
338 | 338 | self.complete = name |
|
339 | 339 | self.type = 'crashed' |
|
340 | 340 | self.name_with_symbols = name |
|
341 | 341 | self.signature = '' |
|
342 | 342 | self._origin = 'fake' |
|
343 | 343 | |
|
344 | 344 | def __repr__(self): |
|
345 | 345 | return '<Fake completion object jedi has crashed>' |
|
346 | 346 | |
|
347 | 347 | |
|
348 | 348 | class Completion: |
|
349 | 349 | """ |
|
350 | 350 | Completion object used and return by IPython completers. |
|
351 | 351 | |
|
352 | 352 | .. warning:: Unstable |
|
353 | 353 | |
|
354 | 354 | This function is unstable, API may change without warning. |
|
355 | 355 | It will also raise unless use in proper context manager. |
|
356 | 356 | |
|
357 | 357 | This act as a middle ground :any:`Completion` object between the |
|
358 | 358 | :any:`jedi.api.classes.Completion` object and the Prompt Toolkit completion |
|
359 | 359 | object. While Jedi need a lot of information about evaluator and how the |
|
360 | 360 | code should be ran/inspected, PromptToolkit (and other frontend) mostly |
|
361 | 361 | need user facing information. |
|
362 | 362 | |
|
363 | 363 | - Which range should be replaced replaced by what. |
|
364 | 364 | - Some metadata (like completion type), or meta information to displayed to |
|
365 | 365 | the use user. |
|
366 | 366 | |
|
367 | 367 | For debugging purpose we can also store the origin of the completion (``jedi``, |
|
368 | 368 | ``IPython.python_matches``, ``IPython.magics_matches``...). |
|
369 | 369 | """ |
|
370 | 370 | |
|
371 | 371 | __slots__ = ['start', 'end', 'text', 'type', 'signature', '_origin'] |
|
372 | 372 | |
|
373 | 373 | def __init__(self, start: int, end: int, text: str, *, type: str=None, _origin='', signature='') -> None: |
|
374 | 374 | warnings.warn("``Completion`` is a provisional API (as of IPython 6.0). " |
|
375 | 375 | "It may change without warnings. " |
|
376 | 376 | "Use in corresponding context manager.", |
|
377 | 377 | category=ProvisionalCompleterWarning, stacklevel=2) |
|
378 | 378 | |
|
379 | 379 | self.start = start |
|
380 | 380 | self.end = end |
|
381 | 381 | self.text = text |
|
382 | 382 | self.type = type |
|
383 | 383 | self.signature = signature |
|
384 | 384 | self._origin = _origin |
|
385 | 385 | |
|
386 | 386 | def __repr__(self): |
|
387 | 387 | return '<Completion start=%s end=%s text=%r type=%r, signature=%r,>' % \ |
|
388 | 388 | (self.start, self.end, self.text, self.type or '?', self.signature or '?') |
|
389 | 389 | |
|
390 | 390 | def __eq__(self, other)->Bool: |
|
391 | 391 | """ |
|
392 | 392 | Equality and hash do not hash the type (as some completer may not be |
|
393 | 393 | able to infer the type), but are use to (partially) de-duplicate |
|
394 | 394 | completion. |
|
395 | 395 | |
|
396 | 396 | Completely de-duplicating completion is a bit tricker that just |
|
397 | 397 | comparing as it depends on surrounding text, which Completions are not |
|
398 | 398 | aware of. |
|
399 | 399 | """ |
|
400 | 400 | return self.start == other.start and \ |
|
401 | 401 | self.end == other.end and \ |
|
402 | 402 | self.text == other.text |
|
403 | 403 | |
|
404 | 404 | def __hash__(self): |
|
405 | 405 | return hash((self.start, self.end, self.text)) |
|
406 | 406 | |
|
407 | 407 | |
|
408 | 408 | _IC = Iterable[Completion] |
|
409 | 409 | |
|
410 | 410 | |
|
411 | 411 | def _deduplicate_completions(text: str, completions: _IC)-> _IC: |
|
412 | 412 | """ |
|
413 | 413 | Deduplicate a set of completions. |
|
414 | 414 | |
|
415 | 415 | .. warning:: Unstable |
|
416 | 416 | |
|
417 | 417 | This function is unstable, API may change without warning. |
|
418 | 418 | |
|
419 | 419 | Parameters |
|
420 | 420 | ---------- |
|
421 | 421 | text: str |
|
422 | 422 | text that should be completed. |
|
423 | 423 | completions: Iterator[Completion] |
|
424 | 424 | iterator over the completions to deduplicate |
|
425 | 425 | |
|
426 | 426 | Yields |
|
427 | 427 | ------ |
|
428 | 428 | `Completions` objects |
|
429 | 429 | |
|
430 | 430 | |
|
431 | 431 | Completions coming from multiple sources, may be different but end up having |
|
432 | 432 | the same effect when applied to ``text``. If this is the case, this will |
|
433 | 433 | consider completions as equal and only emit the first encountered. |
|
434 | 434 | |
|
435 | 435 | Not folded in `completions()` yet for debugging purpose, and to detect when |
|
436 | 436 | the IPython completer does return things that Jedi does not, but should be |
|
437 | 437 | at some point. |
|
438 | 438 | """ |
|
439 | 439 | completions = list(completions) |
|
440 | 440 | if not completions: |
|
441 | 441 | return |
|
442 | 442 | |
|
443 | 443 | new_start = min(c.start for c in completions) |
|
444 | 444 | new_end = max(c.end for c in completions) |
|
445 | 445 | |
|
446 | 446 | seen = set() |
|
447 | 447 | for c in completions: |
|
448 | 448 | new_text = text[new_start:c.start] + c.text + text[c.end:new_end] |
|
449 | 449 | if new_text not in seen: |
|
450 | 450 | yield c |
|
451 | 451 | seen.add(new_text) |
|
452 | 452 | |
|
453 | 453 | |
|
454 | 454 | def rectify_completions(text: str, completions: _IC, *, _debug=False)->_IC: |
|
455 | 455 | """ |
|
456 | 456 | Rectify a set of completions to all have the same ``start`` and ``end`` |
|
457 | 457 | |
|
458 | 458 | .. warning:: Unstable |
|
459 | 459 | |
|
460 | 460 | This function is unstable, API may change without warning. |
|
461 | 461 | It will also raise unless use in proper context manager. |
|
462 | 462 | |
|
463 | 463 | Parameters |
|
464 | 464 | ---------- |
|
465 | 465 | text: str |
|
466 | 466 | text that should be completed. |
|
467 | 467 | completions: Iterator[Completion] |
|
468 | 468 | iterator over the completions to rectify |
|
469 | 469 | |
|
470 | 470 | |
|
471 | 471 | :any:`jedi.api.classes.Completion` s returned by Jedi may not have the same start and end, though |
|
472 | 472 | the Jupyter Protocol requires them to behave like so. This will readjust |
|
473 | 473 | the completion to have the same ``start`` and ``end`` by padding both |
|
474 | 474 | extremities with surrounding text. |
|
475 | 475 | |
|
476 | 476 | During stabilisation should support a ``_debug`` option to log which |
|
477 | 477 | completion are return by the IPython completer and not found in Jedi in |
|
478 | 478 | order to make upstream bug report. |
|
479 | 479 | """ |
|
480 | 480 | warnings.warn("`rectify_completions` is a provisional API (as of IPython 6.0). " |
|
481 | 481 | "It may change without warnings. " |
|
482 | 482 | "Use in corresponding context manager.", |
|
483 | 483 | category=ProvisionalCompleterWarning, stacklevel=2) |
|
484 | 484 | |
|
485 | 485 | completions = list(completions) |
|
486 | 486 | if not completions: |
|
487 | 487 | return |
|
488 | 488 | starts = (c.start for c in completions) |
|
489 | 489 | ends = (c.end for c in completions) |
|
490 | 490 | |
|
491 | 491 | new_start = min(starts) |
|
492 | 492 | new_end = max(ends) |
|
493 | 493 | |
|
494 | 494 | seen_jedi = set() |
|
495 | 495 | seen_python_matches = set() |
|
496 | 496 | for c in completions: |
|
497 | 497 | new_text = text[new_start:c.start] + c.text + text[c.end:new_end] |
|
498 | 498 | if c._origin == 'jedi': |
|
499 | 499 | seen_jedi.add(new_text) |
|
500 | 500 | elif c._origin == 'IPCompleter.python_matches': |
|
501 | 501 | seen_python_matches.add(new_text) |
|
502 | 502 | yield Completion(new_start, new_end, new_text, type=c.type, _origin=c._origin, signature=c.signature) |
|
503 | 503 | diff = seen_python_matches.difference(seen_jedi) |
|
504 | 504 | if diff and _debug: |
|
505 | 505 | print('IPython.python matches have extras:', diff) |
|
506 | 506 | |
|
507 | 507 | |
|
508 | 508 | if sys.platform == 'win32': |
|
509 | 509 | DELIMS = ' \t\n`!@#$^&*()=+[{]}|;\'",<>?' |
|
510 | 510 | else: |
|
511 | 511 | DELIMS = ' \t\n`!@#$^&*()=+[{]}\\|;:\'",<>?' |
|
512 | 512 | |
|
513 | 513 | GREEDY_DELIMS = ' =\r\n' |
|
514 | 514 | |
|
515 | 515 | |
|
516 | 516 | class CompletionSplitter(object): |
|
517 | 517 | """An object to split an input line in a manner similar to readline. |
|
518 | 518 | |
|
519 | 519 | By having our own implementation, we can expose readline-like completion in |
|
520 | 520 | a uniform manner to all frontends. This object only needs to be given the |
|
521 | 521 | line of text to be split and the cursor position on said line, and it |
|
522 | 522 | returns the 'word' to be completed on at the cursor after splitting the |
|
523 | 523 | entire line. |
|
524 | 524 | |
|
525 | 525 | What characters are used as splitting delimiters can be controlled by |
|
526 | 526 | setting the ``delims`` attribute (this is a property that internally |
|
527 | 527 | automatically builds the necessary regular expression)""" |
|
528 | 528 | |
|
529 | 529 | # Private interface |
|
530 | 530 | |
|
531 | 531 | # A string of delimiter characters. The default value makes sense for |
|
532 | 532 | # IPython's most typical usage patterns. |
|
533 | 533 | _delims = DELIMS |
|
534 | 534 | |
|
535 | 535 | # The expression (a normal string) to be compiled into a regular expression |
|
536 | 536 | # for actual splitting. We store it as an attribute mostly for ease of |
|
537 | 537 | # debugging, since this type of code can be so tricky to debug. |
|
538 | 538 | _delim_expr = None |
|
539 | 539 | |
|
540 | 540 | # The regular expression that does the actual splitting |
|
541 | 541 | _delim_re = None |
|
542 | 542 | |
|
543 | 543 | def __init__(self, delims=None): |
|
544 | 544 | delims = CompletionSplitter._delims if delims is None else delims |
|
545 | 545 | self.delims = delims |
|
546 | 546 | |
|
547 | 547 | @property |
|
548 | 548 | def delims(self): |
|
549 | 549 | """Return the string of delimiter characters.""" |
|
550 | 550 | return self._delims |
|
551 | 551 | |
|
552 | 552 | @delims.setter |
|
553 | 553 | def delims(self, delims): |
|
554 | 554 | """Set the delimiters for line splitting.""" |
|
555 | 555 | expr = '[' + ''.join('\\'+ c for c in delims) + ']' |
|
556 | 556 | self._delim_re = re.compile(expr) |
|
557 | 557 | self._delims = delims |
|
558 | 558 | self._delim_expr = expr |
|
559 | 559 | |
|
560 | 560 | def split_line(self, line, cursor_pos=None): |
|
561 | 561 | """Split a line of text with a cursor at the given position. |
|
562 | 562 | """ |
|
563 | 563 | l = line if cursor_pos is None else line[:cursor_pos] |
|
564 | 564 | return self._delim_re.split(l)[-1] |
|
565 | 565 | |
|
566 | 566 | |
|
567 | 567 | |
|
568 | 568 | class Completer(Configurable): |
|
569 | 569 | |
|
570 | 570 | greedy = Bool(False, |
|
571 | 571 | help="""Activate greedy completion |
|
572 | 572 | PENDING DEPRECTION. this is now mostly taken care of with Jedi. |
|
573 | 573 | |
|
574 | 574 | This will enable completion on elements of lists, results of function calls, etc., |
|
575 | 575 | but can be unsafe because the code is actually evaluated on TAB. |
|
576 | 576 | """ |
|
577 | 577 | ).tag(config=True) |
|
578 | 578 | |
|
579 | 579 | use_jedi = Bool(default_value=False, |
|
580 | 580 | help="Experimental: Use Jedi to generate autocompletions. " |
|
581 | 581 | "Off by default.").tag(config=True) |
|
582 | 582 | |
|
583 | 583 | jedi_compute_type_timeout = Int(default_value=400, |
|
584 | 584 | help="""Experimental: restrict time (in milliseconds) during which Jedi can compute types. |
|
585 | 585 | Set to 0 to stop computing types. Non-zero value lower than 100ms may hurt |
|
586 | 586 | performance by preventing jedi to build its cache. |
|
587 | 587 | """).tag(config=True) |
|
588 | 588 | |
|
589 | 589 | debug = Bool(default_value=False, |
|
590 | 590 | help='Enable debug for the Completer. Mostly print extra ' |
|
591 | 591 | 'information for experimental jedi integration.')\ |
|
592 | 592 | .tag(config=True) |
|
593 | 593 | |
|
594 | 594 | backslash_combining_completions = Bool(True, |
|
595 | 595 | help="Enable unicode completions, e.g. \\alpha<tab> . " |
|
596 | 596 | "Includes completion of latex commands, unicode names, and expanding " |
|
597 | 597 | "unicode characters back to latex commands.").tag(config=True) |
|
598 | 598 | |
|
599 | 599 | |
|
600 | 600 | |
|
601 | 601 | def __init__(self, namespace=None, global_namespace=None, **kwargs): |
|
602 | 602 | """Create a new completer for the command line. |
|
603 | 603 | |
|
604 | 604 | Completer(namespace=ns, global_namespace=ns2) -> completer instance. |
|
605 | 605 | |
|
606 | 606 | If unspecified, the default namespace where completions are performed |
|
607 | 607 | is __main__ (technically, __main__.__dict__). Namespaces should be |
|
608 | 608 | given as dictionaries. |
|
609 | 609 | |
|
610 | 610 | An optional second namespace can be given. This allows the completer |
|
611 | 611 | to handle cases where both the local and global scopes need to be |
|
612 | 612 | distinguished. |
|
613 | 613 | """ |
|
614 | 614 | |
|
615 | 615 | # Don't bind to namespace quite yet, but flag whether the user wants a |
|
616 | 616 | # specific namespace or to use __main__.__dict__. This will allow us |
|
617 | 617 | # to bind to __main__.__dict__ at completion time, not now. |
|
618 | 618 | if namespace is None: |
|
619 | 619 | self.use_main_ns = True |
|
620 | 620 | else: |
|
621 | 621 | self.use_main_ns = False |
|
622 | 622 | self.namespace = namespace |
|
623 | 623 | |
|
624 | 624 | # The global namespace, if given, can be bound directly |
|
625 | 625 | if global_namespace is None: |
|
626 | 626 | self.global_namespace = {} |
|
627 | 627 | else: |
|
628 | 628 | self.global_namespace = global_namespace |
|
629 | 629 | |
|
630 | 630 | super(Completer, self).__init__(**kwargs) |
|
631 | 631 | |
|
632 | 632 | def complete(self, text, state): |
|
633 | 633 | """Return the next possible completion for 'text'. |
|
634 | 634 | |
|
635 | 635 | This is called successively with state == 0, 1, 2, ... until it |
|
636 | 636 | returns None. The completion should begin with 'text'. |
|
637 | 637 | |
|
638 | 638 | """ |
|
639 | 639 | if self.use_main_ns: |
|
640 | 640 | self.namespace = __main__.__dict__ |
|
641 | 641 | |
|
642 | 642 | if state == 0: |
|
643 | 643 | if "." in text: |
|
644 | 644 | self.matches = self.attr_matches(text) |
|
645 | 645 | else: |
|
646 | 646 | self.matches = self.global_matches(text) |
|
647 | 647 | try: |
|
648 | 648 | return self.matches[state] |
|
649 | 649 | except IndexError: |
|
650 | 650 | return None |
|
651 | 651 | |
|
652 | 652 | def global_matches(self, text): |
|
653 | 653 | """Compute matches when text is a simple name. |
|
654 | 654 | |
|
655 | 655 | Return a list of all keywords, built-in functions and names currently |
|
656 | 656 | defined in self.namespace or self.global_namespace that match. |
|
657 | 657 | |
|
658 | 658 | """ |
|
659 | 659 | matches = [] |
|
660 | 660 | match_append = matches.append |
|
661 | 661 | n = len(text) |
|
662 | 662 | for lst in [keyword.kwlist, |
|
663 | 663 | builtin_mod.__dict__.keys(), |
|
664 | 664 | self.namespace.keys(), |
|
665 | 665 | self.global_namespace.keys()]: |
|
666 | 666 | for word in lst: |
|
667 | 667 | if word[:n] == text and word != "__builtins__": |
|
668 | 668 | match_append(word) |
|
669 | 669 | |
|
670 | 670 | snake_case_re = re.compile(r"[^_]+(_[^_]+)+?\Z") |
|
671 | 671 | for lst in [self.namespace.keys(), |
|
672 | 672 | self.global_namespace.keys()]: |
|
673 | 673 | shortened = {"_".join([sub[0] for sub in word.split('_')]) : word |
|
674 | 674 | for word in lst if snake_case_re.match(word)} |
|
675 | 675 | for word in shortened.keys(): |
|
676 | 676 | if word[:n] == text and word != "__builtins__": |
|
677 | 677 | match_append(shortened[word]) |
|
678 | 678 | return matches |
|
679 | 679 | |
|
680 | 680 | def attr_matches(self, text): |
|
681 | 681 | """Compute matches when text contains a dot. |
|
682 | 682 | |
|
683 | 683 | Assuming the text is of the form NAME.NAME....[NAME], and is |
|
684 | 684 | evaluatable in self.namespace or self.global_namespace, it will be |
|
685 | 685 | evaluated and its attributes (as revealed by dir()) are used as |
|
686 | 686 | possible completions. (For class instances, class members are |
|
687 | 687 | also considered.) |
|
688 | 688 | |
|
689 | 689 | WARNING: this can still invoke arbitrary C code, if an object |
|
690 | 690 | with a __getattr__ hook is evaluated. |
|
691 | 691 | |
|
692 | 692 | """ |
|
693 | 693 | |
|
694 | 694 | # Another option, seems to work great. Catches things like ''.<tab> |
|
695 | 695 | m = re.match(r"(\S+(\.\w+)*)\.(\w*)$", text) |
|
696 | 696 | |
|
697 | 697 | if m: |
|
698 | 698 | expr, attr = m.group(1, 3) |
|
699 | 699 | elif self.greedy: |
|
700 | 700 | m2 = re.match(r"(.+)\.(\w*)$", self.line_buffer) |
|
701 | 701 | if not m2: |
|
702 | 702 | return [] |
|
703 | 703 | expr, attr = m2.group(1,2) |
|
704 | 704 | else: |
|
705 | 705 | return [] |
|
706 | 706 | |
|
707 | 707 | try: |
|
708 | 708 | obj = eval(expr, self.namespace) |
|
709 | 709 | except: |
|
710 | 710 | try: |
|
711 | 711 | obj = eval(expr, self.global_namespace) |
|
712 | 712 | except: |
|
713 | 713 | return [] |
|
714 | 714 | |
|
715 | 715 | if self.limit_to__all__ and hasattr(obj, '__all__'): |
|
716 | 716 | words = get__all__entries(obj) |
|
717 | 717 | else: |
|
718 | 718 | words = dir2(obj) |
|
719 | 719 | |
|
720 | 720 | try: |
|
721 | 721 | words = generics.complete_object(obj, words) |
|
722 | 722 | except TryNext: |
|
723 | 723 | pass |
|
724 | 724 | except AssertionError: |
|
725 | 725 | raise |
|
726 | 726 | except Exception: |
|
727 | 727 | # Silence errors from completion function |
|
728 | 728 | #raise # dbg |
|
729 | 729 | pass |
|
730 | 730 | # Build match list to return |
|
731 | 731 | n = len(attr) |
|
732 | 732 | return [u"%s.%s" % (expr, w) for w in words if w[:n] == attr ] |
|
733 | 733 | |
|
734 | 734 | |
|
735 | 735 | def get__all__entries(obj): |
|
736 | 736 | """returns the strings in the __all__ attribute""" |
|
737 | 737 | try: |
|
738 | 738 | words = getattr(obj, '__all__') |
|
739 | 739 | except: |
|
740 | 740 | return [] |
|
741 | 741 | |
|
742 | 742 | return [w for w in words if isinstance(w, str)] |
|
743 | 743 | |
|
744 | 744 | |
|
745 | 745 | def match_dict_keys(keys: List[str], prefix: str, delims: str): |
|
746 | 746 | """Used by dict_key_matches, matching the prefix to a list of keys |
|
747 | 747 | |
|
748 | 748 | Parameters |
|
749 | 749 | ========== |
|
750 | 750 | keys: |
|
751 | 751 | list of keys in dictionary currently being completed. |
|
752 | 752 | prefix: |
|
753 | 753 | Part of the text already typed by the user. e.g. `mydict[b'fo` |
|
754 | 754 | delims: |
|
755 | 755 | String of delimiters to consider when finding the current key. |
|
756 | 756 | |
|
757 | 757 | Returns |
|
758 | 758 | ======= |
|
759 | 759 | |
|
760 | 760 | A tuple of three elements: ``quote``, ``token_start``, ``matched``, with |
|
761 | 761 | ``quote`` being the quote that need to be used to close current string. |
|
762 | 762 | ``token_start`` the position where the replacement should start occurring, |
|
763 | 763 | ``matches`` a list of replacement/completion |
|
764 | 764 | |
|
765 | 765 | """ |
|
766 | 766 | if not prefix: |
|
767 | 767 | return None, 0, [repr(k) for k in keys |
|
768 | 768 | if isinstance(k, (str, bytes))] |
|
769 | 769 | quote_match = re.search('["\']', prefix) |
|
770 | 770 | quote = quote_match.group() |
|
771 | 771 | try: |
|
772 | 772 | prefix_str = eval(prefix + quote, {}) |
|
773 | 773 | except Exception: |
|
774 | 774 | return None, 0, [] |
|
775 | 775 | |
|
776 | 776 | pattern = '[^' + ''.join('\\' + c for c in delims) + ']*$' |
|
777 | 777 | token_match = re.search(pattern, prefix, re.UNICODE) |
|
778 | 778 | token_start = token_match.start() |
|
779 | 779 | token_prefix = token_match.group() |
|
780 | 780 | |
|
781 | 781 | matched = [] |
|
782 | 782 | for key in keys: |
|
783 | 783 | try: |
|
784 | 784 | if not key.startswith(prefix_str): |
|
785 | 785 | continue |
|
786 | 786 | except (AttributeError, TypeError, UnicodeError): |
|
787 | 787 | # Python 3+ TypeError on b'a'.startswith('a') or vice-versa |
|
788 | 788 | continue |
|
789 | 789 | |
|
790 | 790 | # reformat remainder of key to begin with prefix |
|
791 | 791 | rem = key[len(prefix_str):] |
|
792 | 792 | # force repr wrapped in ' |
|
793 | 793 | rem_repr = repr(rem + '"') if isinstance(rem, str) else repr(rem + b'"') |
|
794 | 794 | if rem_repr.startswith('u') and prefix[0] not in 'uU': |
|
795 | 795 | # Found key is unicode, but prefix is Py2 string. |
|
796 | 796 | # Therefore attempt to interpret key as string. |
|
797 | 797 | try: |
|
798 | 798 | rem_repr = repr(rem.encode('ascii') + '"') |
|
799 | 799 | except UnicodeEncodeError: |
|
800 | 800 | continue |
|
801 | 801 | |
|
802 | 802 | rem_repr = rem_repr[1 + rem_repr.index("'"):-2] |
|
803 | 803 | if quote == '"': |
|
804 | 804 | # The entered prefix is quoted with ", |
|
805 | 805 | # but the match is quoted with '. |
|
806 | 806 | # A contained " hence needs escaping for comparison: |
|
807 | 807 | rem_repr = rem_repr.replace('"', '\\"') |
|
808 | 808 | |
|
809 | 809 | # then reinsert prefix from start of token |
|
810 | 810 | matched.append('%s%s' % (token_prefix, rem_repr)) |
|
811 | 811 | return quote, token_start, matched |
|
812 | 812 | |
|
813 | 813 | |
|
814 | 814 | def cursor_to_position(text:str, line:int, column:int)->int: |
|
815 | 815 | """ |
|
816 | 816 | |
|
817 | 817 | Convert the (line,column) position of the cursor in text to an offset in a |
|
818 | 818 | string. |
|
819 | 819 | |
|
820 | 820 | Parameters |
|
821 | 821 | ---------- |
|
822 | 822 | |
|
823 | 823 | text : str |
|
824 | 824 | The text in which to calculate the cursor offset |
|
825 | 825 | line : int |
|
826 | 826 | Line of the cursor; 0-indexed |
|
827 | 827 | column : int |
|
828 | 828 | Column of the cursor 0-indexed |
|
829 | 829 | |
|
830 | 830 | Return |
|
831 | 831 | ------ |
|
832 | 832 | Position of the cursor in ``text``, 0-indexed. |
|
833 | 833 | |
|
834 | 834 | See Also |
|
835 | 835 | -------- |
|
836 | 836 | position_to_cursor: reciprocal of this function |
|
837 | 837 | |
|
838 | 838 | """ |
|
839 | 839 | lines = text.split('\n') |
|
840 | 840 | assert line <= len(lines), '{} <= {}'.format(str(line), str(len(lines))) |
|
841 | 841 | |
|
842 | 842 | return sum(len(l) + 1 for l in lines[:line]) + column |
|
843 | 843 | |
|
844 | 844 | def position_to_cursor(text:str, offset:int)->Tuple[int, int]: |
|
845 | 845 | """ |
|
846 | 846 | Convert the position of the cursor in text (0 indexed) to a line |
|
847 | 847 | number(0-indexed) and a column number (0-indexed) pair |
|
848 | 848 | |
|
849 | 849 | Position should be a valid position in ``text``. |
|
850 | 850 | |
|
851 | 851 | Parameters |
|
852 | 852 | ---------- |
|
853 | 853 | |
|
854 | 854 | text : str |
|
855 | 855 | The text in which to calculate the cursor offset |
|
856 | 856 | offset : int |
|
857 | 857 | Position of the cursor in ``text``, 0-indexed. |
|
858 | 858 | |
|
859 | 859 | Return |
|
860 | 860 | ------ |
|
861 | 861 | (line, column) : (int, int) |
|
862 | 862 | Line of the cursor; 0-indexed, column of the cursor 0-indexed |
|
863 | 863 | |
|
864 | 864 | |
|
865 | 865 | See Also |
|
866 | 866 | -------- |
|
867 | 867 | cursor_to_position : reciprocal of this function |
|
868 | 868 | |
|
869 | 869 | |
|
870 | 870 | """ |
|
871 | 871 | |
|
872 | 872 | assert 0 <= offset <= len(text) , "0 <= %s <= %s" % (offset , len(text)) |
|
873 | 873 | |
|
874 | 874 | before = text[:offset] |
|
875 | 875 | blines = before.split('\n') # ! splitnes trim trailing \n |
|
876 | 876 | line = before.count('\n') |
|
877 | 877 | col = len(blines[-1]) |
|
878 | 878 | return line, col |
|
879 | 879 | |
|
880 | 880 | |
|
881 | 881 | def _safe_isinstance(obj, module, class_name): |
|
882 | 882 | """Checks if obj is an instance of module.class_name if loaded |
|
883 | 883 | """ |
|
884 | 884 | return (module in sys.modules and |
|
885 | 885 | isinstance(obj, getattr(import_module(module), class_name))) |
|
886 | 886 | |
|
887 | 887 | |
|
888 | 888 | def back_unicode_name_matches(text): |
|
889 | 889 | u"""Match unicode characters back to unicode name |
|
890 | 890 | |
|
891 | 891 | This does ``β`` -> ``\\snowman`` |
|
892 | 892 | |
|
893 | 893 | Note that snowman is not a valid python3 combining character but will be expanded. |
|
894 | 894 | Though it will not recombine back to the snowman character by the completion machinery. |
|
895 | 895 | |
|
896 | 896 | This will not either back-complete standard sequences like \\n, \\b ... |
|
897 | 897 | |
|
898 | 898 | Used on Python 3 only. |
|
899 | 899 | """ |
|
900 | 900 | if len(text)<2: |
|
901 | 901 | return u'', () |
|
902 | 902 | maybe_slash = text[-2] |
|
903 | 903 | if maybe_slash != '\\': |
|
904 | 904 | return u'', () |
|
905 | 905 | |
|
906 | 906 | char = text[-1] |
|
907 | 907 | # no expand on quote for completion in strings. |
|
908 | 908 | # nor backcomplete standard ascii keys |
|
909 | 909 | if char in string.ascii_letters or char in ['"',"'"]: |
|
910 | 910 | return u'', () |
|
911 | 911 | try : |
|
912 | 912 | unic = unicodedata.name(char) |
|
913 | 913 | return '\\'+char,['\\'+unic] |
|
914 | 914 | except KeyError: |
|
915 | 915 | pass |
|
916 | 916 | return u'', () |
|
917 | 917 | |
|
918 | 918 | def back_latex_name_matches(text:str): |
|
919 | 919 | """Match latex characters back to unicode name |
|
920 | 920 | |
|
921 | 921 | This does ``\\β΅`` -> ``\\aleph`` |
|
922 | 922 | |
|
923 | 923 | Used on Python 3 only. |
|
924 | 924 | """ |
|
925 | 925 | if len(text)<2: |
|
926 | 926 | return u'', () |
|
927 | 927 | maybe_slash = text[-2] |
|
928 | 928 | if maybe_slash != '\\': |
|
929 | 929 | return u'', () |
|
930 | 930 | |
|
931 | 931 | |
|
932 | 932 | char = text[-1] |
|
933 | 933 | # no expand on quote for completion in strings. |
|
934 | 934 | # nor backcomplete standard ascii keys |
|
935 | 935 | if char in string.ascii_letters or char in ['"',"'"]: |
|
936 | 936 | return u'', () |
|
937 | 937 | try : |
|
938 | 938 | latex = reverse_latex_symbol[char] |
|
939 | 939 | # '\\' replace the \ as well |
|
940 | 940 | return '\\'+char,[latex] |
|
941 | 941 | except KeyError: |
|
942 | 942 | pass |
|
943 | 943 | return u'', () |
|
944 | 944 | |
|
945 | 945 | |
|
946 | 946 | def _formatparamchildren(parameter) -> str: |
|
947 | 947 | """ |
|
948 | 948 | Get parameter name and value from Jedi Private API |
|
949 | 949 | |
|
950 | 950 | Jedi does not expose a simple way to get `param=value` from its API. |
|
951 | 951 | |
|
952 | 952 | Parameter |
|
953 | 953 | ========= |
|
954 | 954 | |
|
955 | 955 | parameter: |
|
956 | 956 | Jedi's function `Param` |
|
957 | 957 | |
|
958 | 958 | Returns |
|
959 | 959 | ======= |
|
960 | 960 | |
|
961 | 961 | A string like 'a', 'b=1', '*args', '**kwargs' |
|
962 | 962 | |
|
963 | 963 | |
|
964 | 964 | """ |
|
965 | 965 | description = parameter.description |
|
966 | 966 | if not description.startswith('param '): |
|
967 | 967 | raise ValueError('Jedi function parameter description have change format.' |
|
968 | 968 | 'Expected "param ...", found %r".' % description) |
|
969 | 969 | return description[6:] |
|
970 | 970 | |
|
971 | 971 | def _make_signature(completion)-> str: |
|
972 | 972 | """ |
|
973 | 973 | Make the signature from a jedi completion |
|
974 | 974 | |
|
975 | 975 | Parameter |
|
976 | 976 | ========= |
|
977 | 977 | |
|
978 | 978 | completion: jedi.Completion |
|
979 | 979 | object does not complete a function type |
|
980 | 980 | |
|
981 | 981 | Returns |
|
982 | 982 | ======= |
|
983 | 983 | |
|
984 | 984 | a string consisting of the function signature, with the parenthesis but |
|
985 | 985 | without the function name. example: |
|
986 | 986 | `(a, *args, b=1, **kwargs)` |
|
987 | 987 | |
|
988 | 988 | """ |
|
989 | 989 | |
|
990 | 990 | return '(%s)'% ', '.join([f for f in (_formatparamchildren(p) for p in completion.params) if f]) |
|
991 | 991 | |
|
992 | 992 | class IPCompleter(Completer): |
|
993 | 993 | """Extension of the completer class with IPython-specific features""" |
|
994 | 994 | |
|
995 | 995 | @observe('greedy') |
|
996 | 996 | def _greedy_changed(self, change): |
|
997 | 997 | """update the splitter and readline delims when greedy is changed""" |
|
998 | 998 | if change['new']: |
|
999 | 999 | self.splitter.delims = GREEDY_DELIMS |
|
1000 | 1000 | else: |
|
1001 | 1001 | self.splitter.delims = DELIMS |
|
1002 | 1002 | |
|
1003 | 1003 | merge_completions = Bool(True, |
|
1004 | 1004 | help="""Whether to merge completion results into a single list |
|
1005 | 1005 | |
|
1006 | 1006 | If False, only the completion results from the first non-empty |
|
1007 | 1007 | completer will be returned. |
|
1008 | 1008 | """ |
|
1009 | 1009 | ).tag(config=True) |
|
1010 | 1010 | omit__names = Enum((0,1,2), default_value=2, |
|
1011 | 1011 | help="""Instruct the completer to omit private method names |
|
1012 | 1012 | |
|
1013 | 1013 | Specifically, when completing on ``object.<tab>``. |
|
1014 | 1014 | |
|
1015 | 1015 | When 2 [default]: all names that start with '_' will be excluded. |
|
1016 | 1016 | |
|
1017 | 1017 | When 1: all 'magic' names (``__foo__``) will be excluded. |
|
1018 | 1018 | |
|
1019 | 1019 | When 0: nothing will be excluded. |
|
1020 | 1020 | """ |
|
1021 | 1021 | ).tag(config=True) |
|
1022 | 1022 | limit_to__all__ = Bool(False, |
|
1023 | 1023 | help=""" |
|
1024 | 1024 | DEPRECATED as of version 5.0. |
|
1025 | 1025 | |
|
1026 | 1026 | Instruct the completer to use __all__ for the completion |
|
1027 | 1027 | |
|
1028 | 1028 | Specifically, when completing on ``object.<tab>``. |
|
1029 | 1029 | |
|
1030 | 1030 | When True: only those names in obj.__all__ will be included. |
|
1031 | 1031 | |
|
1032 | 1032 | When False [default]: the __all__ attribute is ignored |
|
1033 | 1033 | """, |
|
1034 | 1034 | ).tag(config=True) |
|
1035 | 1035 | |
|
1036 | 1036 | @observe('limit_to__all__') |
|
1037 | 1037 | def _limit_to_all_changed(self, change): |
|
1038 | 1038 | warnings.warn('`IPython.core.IPCompleter.limit_to__all__` configuration ' |
|
1039 | 1039 | 'value has been deprecated since IPython 5.0, will be made to have ' |
|
1040 | 1040 | 'no effects and then removed in future version of IPython.', |
|
1041 | 1041 | UserWarning) |
|
1042 | 1042 | |
|
1043 | 1043 | def __init__(self, shell=None, namespace=None, global_namespace=None, |
|
1044 | 1044 | use_readline=_deprecation_readline_sentinel, config=None, **kwargs): |
|
1045 | 1045 | """IPCompleter() -> completer |
|
1046 | 1046 | |
|
1047 | 1047 | Return a completer object. |
|
1048 | 1048 | |
|
1049 | 1049 | Parameters |
|
1050 | 1050 | ---------- |
|
1051 | 1051 | |
|
1052 | 1052 | shell |
|
1053 | 1053 | a pointer to the ipython shell itself. This is needed |
|
1054 | 1054 | because this completer knows about magic functions, and those can |
|
1055 | 1055 | only be accessed via the ipython instance. |
|
1056 | 1056 | |
|
1057 | 1057 | namespace : dict, optional |
|
1058 | 1058 | an optional dict where completions are performed. |
|
1059 | 1059 | |
|
1060 | 1060 | global_namespace : dict, optional |
|
1061 | 1061 | secondary optional dict for completions, to |
|
1062 | 1062 | handle cases (such as IPython embedded inside functions) where |
|
1063 | 1063 | both Python scopes are visible. |
|
1064 | 1064 | |
|
1065 | 1065 | use_readline : bool, optional |
|
1066 | 1066 | DEPRECATED, ignored since IPython 6.0, will have no effects |
|
1067 | 1067 | """ |
|
1068 | 1068 | |
|
1069 | 1069 | self.magic_escape = ESC_MAGIC |
|
1070 | 1070 | self.splitter = CompletionSplitter() |
|
1071 | 1071 | |
|
1072 | 1072 | if use_readline is not _deprecation_readline_sentinel: |
|
1073 | 1073 | warnings.warn('The `use_readline` parameter is deprecated and ignored since IPython 6.0.', |
|
1074 | 1074 | DeprecationWarning, stacklevel=2) |
|
1075 | 1075 | |
|
1076 | 1076 | # _greedy_changed() depends on splitter and readline being defined: |
|
1077 | 1077 | Completer.__init__(self, namespace=namespace, global_namespace=global_namespace, |
|
1078 | 1078 | config=config, **kwargs) |
|
1079 | 1079 | |
|
1080 | 1080 | # List where completion matches will be stored |
|
1081 | 1081 | self.matches = [] |
|
1082 | 1082 | self.shell = shell |
|
1083 | 1083 | # Regexp to split filenames with spaces in them |
|
1084 | 1084 | self.space_name_re = re.compile(r'([^\\] )') |
|
1085 | 1085 | # Hold a local ref. to glob.glob for speed |
|
1086 | 1086 | self.glob = glob.glob |
|
1087 | 1087 | |
|
1088 | 1088 | # Determine if we are running on 'dumb' terminals, like (X)Emacs |
|
1089 | 1089 | # buffers, to avoid completion problems. |
|
1090 | 1090 | term = os.environ.get('TERM','xterm') |
|
1091 | 1091 | self.dumb_terminal = term in ['dumb','emacs'] |
|
1092 | 1092 | |
|
1093 | 1093 | # Special handling of backslashes needed in win32 platforms |
|
1094 | 1094 | if sys.platform == "win32": |
|
1095 | 1095 | self.clean_glob = self._clean_glob_win32 |
|
1096 | 1096 | else: |
|
1097 | 1097 | self.clean_glob = self._clean_glob |
|
1098 | 1098 | |
|
1099 | 1099 | #regexp to parse docstring for function signature |
|
1100 | 1100 | self.docstring_sig_re = re.compile(r'^[\w|\s.]+\(([^)]*)\).*') |
|
1101 | 1101 | self.docstring_kwd_re = re.compile(r'[\s|\[]*(\w+)(?:\s*=\s*.*)') |
|
1102 | 1102 | #use this if positional argument name is also needed |
|
1103 | 1103 | #= re.compile(r'[\s|\[]*(\w+)(?:\s*=?\s*.*)') |
|
1104 | 1104 | |
|
1105 | 1105 | self.magic_arg_matchers = [ |
|
1106 | 1106 | self.magic_config_matches, |
|
1107 | 1107 | self.magic_color_matches, |
|
1108 | 1108 | ] |
|
1109 | 1109 | |
|
1110 | 1110 | # This is set externally by InteractiveShell |
|
1111 | 1111 | self.custom_completers = None |
|
1112 | 1112 | |
|
1113 | 1113 | @property |
|
1114 | 1114 | def matchers(self): |
|
1115 | 1115 | """All active matcher routines for completion""" |
|
1116 | 1116 | if self.use_jedi: |
|
1117 | 1117 | return [ |
|
1118 | 1118 | self.file_matches, |
|
1119 | 1119 | self.magic_matches, |
|
1120 | 1120 | self.dict_key_matches, |
|
1121 | 1121 | ] |
|
1122 | 1122 | else: |
|
1123 | 1123 | return [ |
|
1124 | 1124 | self.python_matches, |
|
1125 | 1125 | self.file_matches, |
|
1126 | 1126 | self.magic_matches, |
|
1127 | 1127 | self.python_func_kw_matches, |
|
1128 | 1128 | self.dict_key_matches, |
|
1129 | 1129 | ] |
|
1130 | 1130 | |
|
1131 | 1131 | def all_completions(self, text): |
|
1132 | 1132 | """ |
|
1133 | 1133 | Wrapper around the complete method for the benefit of emacs. |
|
1134 | 1134 | """ |
|
1135 | 1135 | return self.complete(text)[1] |
|
1136 | 1136 | |
|
1137 | 1137 | def _clean_glob(self, text): |
|
1138 | 1138 | return self.glob("%s*" % text) |
|
1139 | 1139 | |
|
1140 | 1140 | def _clean_glob_win32(self,text): |
|
1141 | 1141 | return [f.replace("\\","/") |
|
1142 | 1142 | for f in self.glob("%s*" % text)] |
|
1143 | 1143 | |
|
1144 | 1144 | def file_matches(self, text): |
|
1145 | 1145 | """Match filenames, expanding ~USER type strings. |
|
1146 | 1146 | |
|
1147 | 1147 | Most of the seemingly convoluted logic in this completer is an |
|
1148 | 1148 | attempt to handle filenames with spaces in them. And yet it's not |
|
1149 | 1149 | quite perfect, because Python's readline doesn't expose all of the |
|
1150 | 1150 | GNU readline details needed for this to be done correctly. |
|
1151 | 1151 | |
|
1152 | 1152 | For a filename with a space in it, the printed completions will be |
|
1153 | 1153 | only the parts after what's already been typed (instead of the |
|
1154 | 1154 | full completions, as is normally done). I don't think with the |
|
1155 | 1155 | current (as of Python 2.3) Python readline it's possible to do |
|
1156 | 1156 | better.""" |
|
1157 | 1157 | |
|
1158 | 1158 | # chars that require escaping with backslash - i.e. chars |
|
1159 | 1159 | # that readline treats incorrectly as delimiters, but we |
|
1160 | 1160 | # don't want to treat as delimiters in filename matching |
|
1161 | 1161 | # when escaped with backslash |
|
1162 | 1162 | if text.startswith('!'): |
|
1163 | 1163 | text = text[1:] |
|
1164 | 1164 | text_prefix = u'!' |
|
1165 | 1165 | else: |
|
1166 | 1166 | text_prefix = u'' |
|
1167 | 1167 | |
|
1168 | 1168 | text_until_cursor = self.text_until_cursor |
|
1169 | 1169 | # track strings with open quotes |
|
1170 | 1170 | open_quotes = has_open_quotes(text_until_cursor) |
|
1171 | 1171 | |
|
1172 | 1172 | if '(' in text_until_cursor or '[' in text_until_cursor: |
|
1173 | 1173 | lsplit = text |
|
1174 | 1174 | else: |
|
1175 | 1175 | try: |
|
1176 | 1176 | # arg_split ~ shlex.split, but with unicode bugs fixed by us |
|
1177 | 1177 | lsplit = arg_split(text_until_cursor)[-1] |
|
1178 | 1178 | except ValueError: |
|
1179 | 1179 | # typically an unmatched ", or backslash without escaped char. |
|
1180 | 1180 | if open_quotes: |
|
1181 | 1181 | lsplit = text_until_cursor.split(open_quotes)[-1] |
|
1182 | 1182 | else: |
|
1183 | 1183 | return [] |
|
1184 | 1184 | except IndexError: |
|
1185 | 1185 | # tab pressed on empty line |
|
1186 | 1186 | lsplit = "" |
|
1187 | 1187 | |
|
1188 | 1188 | if not open_quotes and lsplit != protect_filename(lsplit): |
|
1189 | 1189 | # if protectables are found, do matching on the whole escaped name |
|
1190 | 1190 | has_protectables = True |
|
1191 | 1191 | text0,text = text,lsplit |
|
1192 | 1192 | else: |
|
1193 | 1193 | has_protectables = False |
|
1194 | 1194 | text = os.path.expanduser(text) |
|
1195 | 1195 | |
|
1196 | 1196 | if text == "": |
|
1197 | 1197 | return [text_prefix + protect_filename(f) for f in self.glob("*")] |
|
1198 | 1198 | |
|
1199 | 1199 | # Compute the matches from the filesystem |
|
1200 | 1200 | if sys.platform == 'win32': |
|
1201 | 1201 | m0 = self.clean_glob(text) |
|
1202 | 1202 | else: |
|
1203 | 1203 | m0 = self.clean_glob(text.replace('\\', '')) |
|
1204 | 1204 | |
|
1205 | 1205 | if has_protectables: |
|
1206 | 1206 | # If we had protectables, we need to revert our changes to the |
|
1207 | 1207 | # beginning of filename so that we don't double-write the part |
|
1208 | 1208 | # of the filename we have so far |
|
1209 | 1209 | len_lsplit = len(lsplit) |
|
1210 | 1210 | matches = [text_prefix + text0 + |
|
1211 | 1211 | protect_filename(f[len_lsplit:]) for f in m0] |
|
1212 | 1212 | else: |
|
1213 | 1213 | if open_quotes: |
|
1214 | 1214 | # if we have a string with an open quote, we don't need to |
|
1215 | 1215 | # protect the names beyond the quote (and we _shouldn't_, as |
|
1216 | 1216 | # it would cause bugs when the filesystem call is made). |
|
1217 | 1217 | matches = m0 if sys.platform == "win32" else\ |
|
1218 | 1218 | [protect_filename(f, open_quotes) for f in m0] |
|
1219 | 1219 | else: |
|
1220 | 1220 | matches = [text_prefix + |
|
1221 | 1221 | protect_filename(f) for f in m0] |
|
1222 | 1222 | |
|
1223 | 1223 | # Mark directories in input list by appending '/' to their names. |
|
1224 | 1224 | return [x+'/' if os.path.isdir(x) else x for x in matches] |
|
1225 | 1225 | |
|
1226 | 1226 | def magic_matches(self, text): |
|
1227 | 1227 | """Match magics""" |
|
1228 | 1228 | # Get all shell magics now rather than statically, so magics loaded at |
|
1229 | 1229 | # runtime show up too. |
|
1230 | 1230 | lsm = self.shell.magics_manager.lsmagic() |
|
1231 | 1231 | line_magics = lsm['line'] |
|
1232 | 1232 | cell_magics = lsm['cell'] |
|
1233 | 1233 | pre = self.magic_escape |
|
1234 | 1234 | pre2 = pre+pre |
|
1235 | 1235 | |
|
1236 | 1236 | explicit_magic = text.startswith(pre) |
|
1237 | 1237 | |
|
1238 | 1238 | # Completion logic: |
|
1239 | 1239 | # - user gives %%: only do cell magics |
|
1240 | 1240 | # - user gives %: do both line and cell magics |
|
1241 | 1241 | # - no prefix: do both |
|
1242 | 1242 | # In other words, line magics are skipped if the user gives %% explicitly |
|
1243 | 1243 | # |
|
1244 | 1244 | # We also exclude magics that match any currently visible names: |
|
1245 | 1245 | # https://github.com/ipython/ipython/issues/4877, unless the user has |
|
1246 | 1246 | # typed a %: |
|
1247 | 1247 | # https://github.com/ipython/ipython/issues/10754 |
|
1248 | 1248 | bare_text = text.lstrip(pre) |
|
1249 | 1249 | global_matches = self.global_matches(bare_text) |
|
1250 | 1250 | if not explicit_magic: |
|
1251 | 1251 | def matches(magic): |
|
1252 | 1252 | """ |
|
1253 | 1253 | Filter magics, in particular remove magics that match |
|
1254 | 1254 | a name present in global namespace. |
|
1255 | 1255 | """ |
|
1256 | 1256 | return ( magic.startswith(bare_text) and |
|
1257 | 1257 | magic not in global_matches ) |
|
1258 | 1258 | else: |
|
1259 | 1259 | def matches(magic): |
|
1260 | 1260 | return magic.startswith(bare_text) |
|
1261 | 1261 | |
|
1262 | 1262 | comp = [ pre2+m for m in cell_magics if matches(m)] |
|
1263 | 1263 | if not text.startswith(pre2): |
|
1264 | 1264 | comp += [ pre+m for m in line_magics if matches(m)] |
|
1265 | 1265 | |
|
1266 | 1266 | return comp |
|
1267 | 1267 | |
|
1268 | 1268 | def magic_config_matches(self, text:str) -> List[str]: |
|
1269 | 1269 | """ Match class names and attributes for %config magic """ |
|
1270 | 1270 | texts = text.strip().split() |
|
1271 | 1271 | |
|
1272 | 1272 | if len(texts) > 0 and (texts[0] == 'config' or texts[0] == '%config'): |
|
1273 | 1273 | # get all configuration classes |
|
1274 | 1274 | classes = sorted(set([ c for c in self.shell.configurables |
|
1275 | 1275 | if c.__class__.class_traits(config=True) |
|
1276 | 1276 | ]), key=lambda x: x.__class__.__name__) |
|
1277 | 1277 | classnames = [ c.__class__.__name__ for c in classes ] |
|
1278 | 1278 | |
|
1279 | 1279 | # return all classnames if config or %config is given |
|
1280 | 1280 | if len(texts) == 1: |
|
1281 | 1281 | return classnames |
|
1282 | 1282 | |
|
1283 | 1283 | # match classname |
|
1284 | 1284 | classname_texts = texts[1].split('.') |
|
1285 | 1285 | classname = classname_texts[0] |
|
1286 | 1286 | classname_matches = [ c for c in classnames |
|
1287 | 1287 | if c.startswith(classname) ] |
|
1288 | 1288 | |
|
1289 | 1289 | # return matched classes or the matched class with attributes |
|
1290 | 1290 | if texts[1].find('.') < 0: |
|
1291 | 1291 | return classname_matches |
|
1292 | 1292 | elif len(classname_matches) == 1 and \ |
|
1293 | 1293 | classname_matches[0] == classname: |
|
1294 | 1294 | cls = classes[classnames.index(classname)].__class__ |
|
1295 | 1295 | help = cls.class_get_help() |
|
1296 | 1296 | # strip leading '--' from cl-args: |
|
1297 | 1297 | help = re.sub(re.compile(r'^--', re.MULTILINE), '', help) |
|
1298 | 1298 | return [ attr.split('=')[0] |
|
1299 | 1299 | for attr in help.strip().splitlines() |
|
1300 | 1300 | if attr.startswith(texts[1]) ] |
|
1301 | 1301 | return [] |
|
1302 | 1302 | |
|
1303 | 1303 | def magic_color_matches(self, text:str) -> List[str] : |
|
1304 | 1304 | """ Match color schemes for %colors magic""" |
|
1305 | 1305 | texts = text.split() |
|
1306 | 1306 | if text.endswith(' '): |
|
1307 | 1307 | # .split() strips off the trailing whitespace. Add '' back |
|
1308 | 1308 | # so that: '%colors ' -> ['%colors', ''] |
|
1309 | 1309 | texts.append('') |
|
1310 | 1310 | |
|
1311 | 1311 | if len(texts) == 2 and (texts[0] == 'colors' or texts[0] == '%colors'): |
|
1312 | 1312 | prefix = texts[1] |
|
1313 | 1313 | return [ color for color in InspectColors.keys() |
|
1314 | 1314 | if color.startswith(prefix) ] |
|
1315 | 1315 | return [] |
|
1316 | 1316 | |
|
1317 | 1317 | def _jedi_matches(self, cursor_column:int, cursor_line:int, text:str): |
|
1318 | 1318 | """ |
|
1319 | 1319 | |
|
1320 | 1320 | Return a list of :any:`jedi.api.Completions` object from a ``text`` and |
|
1321 | 1321 | cursor position. |
|
1322 | 1322 | |
|
1323 | 1323 | Parameters |
|
1324 | 1324 | ---------- |
|
1325 | 1325 | cursor_column : int |
|
1326 | 1326 | column position of the cursor in ``text``, 0-indexed. |
|
1327 | 1327 | cursor_line : int |
|
1328 | 1328 | line position of the cursor in ``text``, 0-indexed |
|
1329 | 1329 | text : str |
|
1330 | 1330 | text to complete |
|
1331 | 1331 | |
|
1332 | 1332 | Debugging |
|
1333 | 1333 | --------- |
|
1334 | 1334 | |
|
1335 | 1335 | If ``IPCompleter.debug`` is ``True`` may return a :any:`_FakeJediCompletion` |
|
1336 | 1336 | object containing a string with the Jedi debug information attached. |
|
1337 | 1337 | """ |
|
1338 | 1338 | namespaces = [self.namespace] |
|
1339 | 1339 | if self.global_namespace is not None: |
|
1340 | 1340 | namespaces.append(self.global_namespace) |
|
1341 | 1341 | |
|
1342 | 1342 | completion_filter = lambda x:x |
|
1343 | 1343 | offset = cursor_to_position(text, cursor_line, cursor_column) |
|
1344 | 1344 | # filter output if we are completing for object members |
|
1345 | 1345 | if offset: |
|
1346 | 1346 | pre = text[offset-1] |
|
1347 | 1347 | if pre == '.': |
|
1348 | 1348 | if self.omit__names == 2: |
|
1349 | 1349 | completion_filter = lambda c:not c.name.startswith('_') |
|
1350 | 1350 | elif self.omit__names == 1: |
|
1351 | 1351 | completion_filter = lambda c:not (c.name.startswith('__') and c.name.endswith('__')) |
|
1352 | 1352 | elif self.omit__names == 0: |
|
1353 | 1353 | completion_filter = lambda x:x |
|
1354 | 1354 | else: |
|
1355 | 1355 | raise ValueError("Don't understand self.omit__names == {}".format(self.omit__names)) |
|
1356 | 1356 | |
|
1357 | 1357 | interpreter = jedi.Interpreter( |
|
1358 | 1358 | text[:offset], namespaces, column=cursor_column, line=cursor_line + 1) |
|
1359 | 1359 | try_jedi = True |
|
1360 | 1360 | |
|
1361 | 1361 | try: |
|
1362 | 1362 | # should we check the type of the node is Error ? |
|
1363 | 1363 | try: |
|
1364 | 1364 | # jedi < 0.11 |
|
1365 | 1365 | from jedi.parser.tree import ErrorLeaf |
|
1366 | 1366 | except ImportError: |
|
1367 | 1367 | # jedi >= 0.11 |
|
1368 | 1368 | from parso.tree import ErrorLeaf |
|
1369 | 1369 | |
|
1370 | 1370 | next_to_last_tree = interpreter._get_module().tree_node.children[-2] |
|
1371 | 1371 | completing_string = False |
|
1372 | 1372 | if isinstance(next_to_last_tree, ErrorLeaf): |
|
1373 | 1373 | completing_string = next_to_last_tree.value.lstrip()[0] in {'"', "'"} |
|
1374 | 1374 | # if we are in a string jedi is likely not the right candidate for |
|
1375 | 1375 | # now. Skip it. |
|
1376 | 1376 | try_jedi = not completing_string |
|
1377 | 1377 | except Exception as e: |
|
1378 | 1378 | # many of things can go wrong, we are using private API just don't crash. |
|
1379 | 1379 | if self.debug: |
|
1380 | 1380 | print("Error detecting if completing a non-finished string :", e, '|') |
|
1381 | 1381 | |
|
1382 | 1382 | if not try_jedi: |
|
1383 | 1383 | return [] |
|
1384 | 1384 | try: |
|
1385 | 1385 | return filter(completion_filter, interpreter.completions()) |
|
1386 | 1386 | except Exception as e: |
|
1387 | 1387 | if self.debug: |
|
1388 | 1388 | return [_FakeJediCompletion('Oops Jedi has crashed, please report a bug with the following:\n"""\n%s\ns"""' % (e))] |
|
1389 | 1389 | else: |
|
1390 | 1390 | return [] |
|
1391 | 1391 | |
|
1392 | 1392 | def python_matches(self, text): |
|
1393 | 1393 | """Match attributes or global python names""" |
|
1394 | 1394 | if "." in text: |
|
1395 | 1395 | try: |
|
1396 | 1396 | matches = self.attr_matches(text) |
|
1397 | 1397 | if text.endswith('.') and self.omit__names: |
|
1398 | 1398 | if self.omit__names == 1: |
|
1399 | 1399 | # true if txt is _not_ a __ name, false otherwise: |
|
1400 | 1400 | no__name = (lambda txt: |
|
1401 | 1401 | re.match(r'.*\.__.*?__',txt) is None) |
|
1402 | 1402 | else: |
|
1403 | 1403 | # true if txt is _not_ a _ name, false otherwise: |
|
1404 | 1404 | no__name = (lambda txt: |
|
1405 | 1405 | re.match(r'\._.*?',txt[txt.rindex('.'):]) is None) |
|
1406 | 1406 | matches = filter(no__name, matches) |
|
1407 | 1407 | except NameError: |
|
1408 | 1408 | # catches <undefined attributes>.<tab> |
|
1409 | 1409 | matches = [] |
|
1410 | 1410 | else: |
|
1411 | 1411 | matches = self.global_matches(text) |
|
1412 | 1412 | return matches |
|
1413 | 1413 | |
|
1414 | 1414 | def _default_arguments_from_docstring(self, doc): |
|
1415 | 1415 | """Parse the first line of docstring for call signature. |
|
1416 | 1416 | |
|
1417 | 1417 | Docstring should be of the form 'min(iterable[, key=func])\n'. |
|
1418 | 1418 | It can also parse cython docstring of the form |
|
1419 | 1419 | 'Minuit.migrad(self, int ncall=10000, resume=True, int nsplit=1)'. |
|
1420 | 1420 | """ |
|
1421 | 1421 | if doc is None: |
|
1422 | 1422 | return [] |
|
1423 | 1423 | |
|
1424 | 1424 | #care only the firstline |
|
1425 | 1425 | line = doc.lstrip().splitlines()[0] |
|
1426 | 1426 | |
|
1427 | 1427 | #p = re.compile(r'^[\w|\s.]+\(([^)]*)\).*') |
|
1428 | 1428 | #'min(iterable[, key=func])\n' -> 'iterable[, key=func]' |
|
1429 | 1429 | sig = self.docstring_sig_re.search(line) |
|
1430 | 1430 | if sig is None: |
|
1431 | 1431 | return [] |
|
1432 | 1432 | # iterable[, key=func]' -> ['iterable[' ,' key=func]'] |
|
1433 | 1433 | sig = sig.groups()[0].split(',') |
|
1434 | 1434 | ret = [] |
|
1435 | 1435 | for s in sig: |
|
1436 | 1436 | #re.compile(r'[\s|\[]*(\w+)(?:\s*=\s*.*)') |
|
1437 | 1437 | ret += self.docstring_kwd_re.findall(s) |
|
1438 | 1438 | return ret |
|
1439 | 1439 | |
|
1440 | 1440 | def _default_arguments(self, obj): |
|
1441 | 1441 | """Return the list of default arguments of obj if it is callable, |
|
1442 | 1442 | or empty list otherwise.""" |
|
1443 | 1443 | call_obj = obj |
|
1444 | 1444 | ret = [] |
|
1445 | 1445 | if inspect.isbuiltin(obj): |
|
1446 | 1446 | pass |
|
1447 | 1447 | elif not (inspect.isfunction(obj) or inspect.ismethod(obj)): |
|
1448 | 1448 | if inspect.isclass(obj): |
|
1449 | 1449 | #for cython embedsignature=True the constructor docstring |
|
1450 | 1450 | #belongs to the object itself not __init__ |
|
1451 | 1451 | ret += self._default_arguments_from_docstring( |
|
1452 | 1452 | getattr(obj, '__doc__', '')) |
|
1453 | 1453 | # for classes, check for __init__,__new__ |
|
1454 | 1454 | call_obj = (getattr(obj, '__init__', None) or |
|
1455 | 1455 | getattr(obj, '__new__', None)) |
|
1456 | 1456 | # for all others, check if they are __call__able |
|
1457 | 1457 | elif hasattr(obj, '__call__'): |
|
1458 | 1458 | call_obj = obj.__call__ |
|
1459 | 1459 | ret += self._default_arguments_from_docstring( |
|
1460 | 1460 | getattr(call_obj, '__doc__', '')) |
|
1461 | 1461 | |
|
1462 | 1462 | _keeps = (inspect.Parameter.KEYWORD_ONLY, |
|
1463 | 1463 | inspect.Parameter.POSITIONAL_OR_KEYWORD) |
|
1464 | 1464 | |
|
1465 | 1465 | try: |
|
1466 | 1466 | sig = inspect.signature(call_obj) |
|
1467 | 1467 | ret.extend(k for k, v in sig.parameters.items() if |
|
1468 | 1468 | v.kind in _keeps) |
|
1469 | 1469 | except ValueError: |
|
1470 | 1470 | pass |
|
1471 | 1471 | |
|
1472 | 1472 | return list(set(ret)) |
|
1473 | 1473 | |
|
1474 | 1474 | def python_func_kw_matches(self,text): |
|
1475 | 1475 | """Match named parameters (kwargs) of the last open function""" |
|
1476 | 1476 | |
|
1477 | 1477 | if "." in text: # a parameter cannot be dotted |
|
1478 | 1478 | return [] |
|
1479 | 1479 | try: regexp = self.__funcParamsRegex |
|
1480 | 1480 | except AttributeError: |
|
1481 | 1481 | regexp = self.__funcParamsRegex = re.compile(r''' |
|
1482 | 1482 | '.*?(?<!\\)' | # single quoted strings or |
|
1483 | 1483 | ".*?(?<!\\)" | # double quoted strings or |
|
1484 | 1484 | \w+ | # identifier |
|
1485 | 1485 | \S # other characters |
|
1486 | 1486 | ''', re.VERBOSE | re.DOTALL) |
|
1487 | 1487 | # 1. find the nearest identifier that comes before an unclosed |
|
1488 | 1488 | # parenthesis before the cursor |
|
1489 | 1489 | # e.g. for "foo (1+bar(x), pa<cursor>,a=1)", the candidate is "foo" |
|
1490 | 1490 | tokens = regexp.findall(self.text_until_cursor) |
|
1491 | 1491 | iterTokens = reversed(tokens); openPar = 0 |
|
1492 | 1492 | |
|
1493 | 1493 | for token in iterTokens: |
|
1494 | 1494 | if token == ')': |
|
1495 | 1495 | openPar -= 1 |
|
1496 | 1496 | elif token == '(': |
|
1497 | 1497 | openPar += 1 |
|
1498 | 1498 | if openPar > 0: |
|
1499 | 1499 | # found the last unclosed parenthesis |
|
1500 | 1500 | break |
|
1501 | 1501 | else: |
|
1502 | 1502 | return [] |
|
1503 | 1503 | # 2. Concatenate dotted names ("foo.bar" for "foo.bar(x, pa" ) |
|
1504 | 1504 | ids = [] |
|
1505 | 1505 | isId = re.compile(r'\w+$').match |
|
1506 | 1506 | |
|
1507 | 1507 | while True: |
|
1508 | 1508 | try: |
|
1509 | 1509 | ids.append(next(iterTokens)) |
|
1510 | 1510 | if not isId(ids[-1]): |
|
1511 | 1511 | ids.pop(); break |
|
1512 | 1512 | if not next(iterTokens) == '.': |
|
1513 | 1513 | break |
|
1514 | 1514 | except StopIteration: |
|
1515 | 1515 | break |
|
1516 | 1516 | |
|
1517 | 1517 | # Find all named arguments already assigned to, as to avoid suggesting |
|
1518 | 1518 | # them again |
|
1519 | 1519 | usedNamedArgs = set() |
|
1520 | 1520 | par_level = -1 |
|
1521 | 1521 | for token, next_token in zip(tokens, tokens[1:]): |
|
1522 | 1522 | if token == '(': |
|
1523 | 1523 | par_level += 1 |
|
1524 | 1524 | elif token == ')': |
|
1525 | 1525 | par_level -= 1 |
|
1526 | 1526 | |
|
1527 | 1527 | if par_level != 0: |
|
1528 | 1528 | continue |
|
1529 | 1529 | |
|
1530 | 1530 | if next_token != '=': |
|
1531 | 1531 | continue |
|
1532 | 1532 | |
|
1533 | 1533 | usedNamedArgs.add(token) |
|
1534 | 1534 | |
|
1535 | 1535 | # lookup the candidate callable matches either using global_matches |
|
1536 | 1536 | # or attr_matches for dotted names |
|
1537 | 1537 | if len(ids) == 1: |
|
1538 | 1538 | callableMatches = self.global_matches(ids[0]) |
|
1539 | 1539 | else: |
|
1540 | 1540 | callableMatches = self.attr_matches('.'.join(ids[::-1])) |
|
1541 | 1541 | argMatches = [] |
|
1542 | 1542 | for callableMatch in callableMatches: |
|
1543 | 1543 | try: |
|
1544 | 1544 | namedArgs = self._default_arguments(eval(callableMatch, |
|
1545 | 1545 | self.namespace)) |
|
1546 | 1546 | except: |
|
1547 | 1547 | continue |
|
1548 | 1548 | |
|
1549 | 1549 | # Remove used named arguments from the list, no need to show twice |
|
1550 | 1550 | for namedArg in set(namedArgs) - usedNamedArgs: |
|
1551 | 1551 | if namedArg.startswith(text): |
|
1552 | 1552 | argMatches.append(u"%s=" %namedArg) |
|
1553 | 1553 | return argMatches |
|
1554 | 1554 | |
|
1555 | 1555 | def dict_key_matches(self, text): |
|
1556 | 1556 | "Match string keys in a dictionary, after e.g. 'foo[' " |
|
1557 | 1557 | def get_keys(obj): |
|
1558 | 1558 | # Objects can define their own completions by defining an |
|
1559 | 1559 | # _ipy_key_completions_() method. |
|
1560 | 1560 | method = get_real_method(obj, '_ipython_key_completions_') |
|
1561 | 1561 | if method is not None: |
|
1562 | 1562 | return method() |
|
1563 | 1563 | |
|
1564 | 1564 | # Special case some common in-memory dict-like types |
|
1565 | 1565 | if isinstance(obj, dict) or\ |
|
1566 | 1566 | _safe_isinstance(obj, 'pandas', 'DataFrame'): |
|
1567 | 1567 | try: |
|
1568 | 1568 | return list(obj.keys()) |
|
1569 | 1569 | except Exception: |
|
1570 | 1570 | return [] |
|
1571 | 1571 | elif _safe_isinstance(obj, 'numpy', 'ndarray') or\ |
|
1572 | 1572 | _safe_isinstance(obj, 'numpy', 'void'): |
|
1573 | 1573 | return obj.dtype.names or [] |
|
1574 | 1574 | return [] |
|
1575 | 1575 | |
|
1576 | 1576 | try: |
|
1577 | 1577 | regexps = self.__dict_key_regexps |
|
1578 | 1578 | except AttributeError: |
|
1579 | 1579 | dict_key_re_fmt = r'''(?x) |
|
1580 | 1580 | ( # match dict-referring expression wrt greedy setting |
|
1581 | 1581 | %s |
|
1582 | 1582 | ) |
|
1583 | 1583 | \[ # open bracket |
|
1584 | 1584 | \s* # and optional whitespace |
|
1585 | 1585 | ([uUbB]? # string prefix (r not handled) |
|
1586 | 1586 | (?: # unclosed string |
|
1587 | 1587 | '(?:[^']|(?<!\\)\\')* |
|
1588 | 1588 | | |
|
1589 | 1589 | "(?:[^"]|(?<!\\)\\")* |
|
1590 | 1590 | ) |
|
1591 | 1591 | )? |
|
1592 | 1592 | $ |
|
1593 | 1593 | ''' |
|
1594 | 1594 | regexps = self.__dict_key_regexps = { |
|
1595 | False: re.compile(dict_key_re_fmt % ''' | |
|
1595 | False: re.compile(dict_key_re_fmt % r''' | |
|
1596 | 1596 | # identifiers separated by . |
|
1597 | 1597 | (?!\d)\w+ |
|
1598 | 1598 | (?:\.(?!\d)\w+)* |
|
1599 | 1599 | '''), |
|
1600 | 1600 | True: re.compile(dict_key_re_fmt % ''' |
|
1601 | 1601 | .+ |
|
1602 | 1602 | ''') |
|
1603 | 1603 | } |
|
1604 | 1604 | |
|
1605 | 1605 | match = regexps[self.greedy].search(self.text_until_cursor) |
|
1606 | 1606 | if match is None: |
|
1607 | 1607 | return [] |
|
1608 | 1608 | |
|
1609 | 1609 | expr, prefix = match.groups() |
|
1610 | 1610 | try: |
|
1611 | 1611 | obj = eval(expr, self.namespace) |
|
1612 | 1612 | except Exception: |
|
1613 | 1613 | try: |
|
1614 | 1614 | obj = eval(expr, self.global_namespace) |
|
1615 | 1615 | except Exception: |
|
1616 | 1616 | return [] |
|
1617 | 1617 | |
|
1618 | 1618 | keys = get_keys(obj) |
|
1619 | 1619 | if not keys: |
|
1620 | 1620 | return keys |
|
1621 | 1621 | closing_quote, token_offset, matches = match_dict_keys(keys, prefix, self.splitter.delims) |
|
1622 | 1622 | if not matches: |
|
1623 | 1623 | return matches |
|
1624 | 1624 | |
|
1625 | 1625 | # get the cursor position of |
|
1626 | 1626 | # - the text being completed |
|
1627 | 1627 | # - the start of the key text |
|
1628 | 1628 | # - the start of the completion |
|
1629 | 1629 | text_start = len(self.text_until_cursor) - len(text) |
|
1630 | 1630 | if prefix: |
|
1631 | 1631 | key_start = match.start(2) |
|
1632 | 1632 | completion_start = key_start + token_offset |
|
1633 | 1633 | else: |
|
1634 | 1634 | key_start = completion_start = match.end() |
|
1635 | 1635 | |
|
1636 | 1636 | # grab the leading prefix, to make sure all completions start with `text` |
|
1637 | 1637 | if text_start > key_start: |
|
1638 | 1638 | leading = '' |
|
1639 | 1639 | else: |
|
1640 | 1640 | leading = text[text_start:completion_start] |
|
1641 | 1641 | |
|
1642 | 1642 | # the index of the `[` character |
|
1643 | 1643 | bracket_idx = match.end(1) |
|
1644 | 1644 | |
|
1645 | 1645 | # append closing quote and bracket as appropriate |
|
1646 | 1646 | # this is *not* appropriate if the opening quote or bracket is outside |
|
1647 | 1647 | # the text given to this method |
|
1648 | 1648 | suf = '' |
|
1649 | 1649 | continuation = self.line_buffer[len(self.text_until_cursor):] |
|
1650 | 1650 | if key_start > text_start and closing_quote: |
|
1651 | 1651 | # quotes were opened inside text, maybe close them |
|
1652 | 1652 | if continuation.startswith(closing_quote): |
|
1653 | 1653 | continuation = continuation[len(closing_quote):] |
|
1654 | 1654 | else: |
|
1655 | 1655 | suf += closing_quote |
|
1656 | 1656 | if bracket_idx > text_start: |
|
1657 | 1657 | # brackets were opened inside text, maybe close them |
|
1658 | 1658 | if not continuation.startswith(']'): |
|
1659 | 1659 | suf += ']' |
|
1660 | 1660 | |
|
1661 | 1661 | return [leading + k + suf for k in matches] |
|
1662 | 1662 | |
|
1663 | 1663 | def unicode_name_matches(self, text): |
|
1664 | 1664 | u"""Match Latex-like syntax for unicode characters base |
|
1665 | 1665 | on the name of the character. |
|
1666 | 1666 | |
|
1667 | 1667 | This does ``\\GREEK SMALL LETTER ETA`` -> ``Ξ·`` |
|
1668 | 1668 | |
|
1669 | 1669 | Works only on valid python 3 identifier, or on combining characters that |
|
1670 | 1670 | will combine to form a valid identifier. |
|
1671 | 1671 | |
|
1672 | 1672 | Used on Python 3 only. |
|
1673 | 1673 | """ |
|
1674 | 1674 | slashpos = text.rfind('\\') |
|
1675 | 1675 | if slashpos > -1: |
|
1676 | 1676 | s = text[slashpos+1:] |
|
1677 | 1677 | try : |
|
1678 | 1678 | unic = unicodedata.lookup(s) |
|
1679 | 1679 | # allow combining chars |
|
1680 | 1680 | if ('a'+unic).isidentifier(): |
|
1681 | 1681 | return '\\'+s,[unic] |
|
1682 | 1682 | except KeyError: |
|
1683 | 1683 | pass |
|
1684 | 1684 | return u'', [] |
|
1685 | 1685 | |
|
1686 | 1686 | |
|
1687 | 1687 | def latex_matches(self, text): |
|
1688 | 1688 | u"""Match Latex syntax for unicode characters. |
|
1689 | 1689 | |
|
1690 | 1690 | This does both ``\\alp`` -> ``\\alpha`` and ``\\alpha`` -> ``Ξ±`` |
|
1691 | 1691 | |
|
1692 | 1692 | Used on Python 3 only. |
|
1693 | 1693 | """ |
|
1694 | 1694 | slashpos = text.rfind('\\') |
|
1695 | 1695 | if slashpos > -1: |
|
1696 | 1696 | s = text[slashpos:] |
|
1697 | 1697 | if s in latex_symbols: |
|
1698 | 1698 | # Try to complete a full latex symbol to unicode |
|
1699 | 1699 | # \\alpha -> Ξ± |
|
1700 | 1700 | return s, [latex_symbols[s]] |
|
1701 | 1701 | else: |
|
1702 | 1702 | # If a user has partially typed a latex symbol, give them |
|
1703 | 1703 | # a full list of options \al -> [\aleph, \alpha] |
|
1704 | 1704 | matches = [k for k in latex_symbols if k.startswith(s)] |
|
1705 | 1705 | return s, matches |
|
1706 | 1706 | return u'', [] |
|
1707 | 1707 | |
|
1708 | 1708 | def dispatch_custom_completer(self, text): |
|
1709 | 1709 | if not self.custom_completers: |
|
1710 | 1710 | return |
|
1711 | 1711 | |
|
1712 | 1712 | line = self.line_buffer |
|
1713 | 1713 | if not line.strip(): |
|
1714 | 1714 | return None |
|
1715 | 1715 | |
|
1716 | 1716 | # Create a little structure to pass all the relevant information about |
|
1717 | 1717 | # the current completion to any custom completer. |
|
1718 | 1718 | event = SimpleNamespace() |
|
1719 | 1719 | event.line = line |
|
1720 | 1720 | event.symbol = text |
|
1721 | 1721 | cmd = line.split(None,1)[0] |
|
1722 | 1722 | event.command = cmd |
|
1723 | 1723 | event.text_until_cursor = self.text_until_cursor |
|
1724 | 1724 | |
|
1725 | 1725 | # for foo etc, try also to find completer for %foo |
|
1726 | 1726 | if not cmd.startswith(self.magic_escape): |
|
1727 | 1727 | try_magic = self.custom_completers.s_matches( |
|
1728 | 1728 | self.magic_escape + cmd) |
|
1729 | 1729 | else: |
|
1730 | 1730 | try_magic = [] |
|
1731 | 1731 | |
|
1732 | 1732 | for c in itertools.chain(self.custom_completers.s_matches(cmd), |
|
1733 | 1733 | try_magic, |
|
1734 | 1734 | self.custom_completers.flat_matches(self.text_until_cursor)): |
|
1735 | 1735 | try: |
|
1736 | 1736 | res = c(event) |
|
1737 | 1737 | if res: |
|
1738 | 1738 | # first, try case sensitive match |
|
1739 | 1739 | withcase = [r for r in res if r.startswith(text)] |
|
1740 | 1740 | if withcase: |
|
1741 | 1741 | return withcase |
|
1742 | 1742 | # if none, then case insensitive ones are ok too |
|
1743 | 1743 | text_low = text.lower() |
|
1744 | 1744 | return [r for r in res if r.lower().startswith(text_low)] |
|
1745 | 1745 | except TryNext: |
|
1746 | 1746 | pass |
|
1747 | 1747 | except KeyboardInterrupt: |
|
1748 | 1748 | """ |
|
1749 | 1749 | If custom completer take too long, |
|
1750 | 1750 | let keyboard interrupt abort and return nothing. |
|
1751 | 1751 | """ |
|
1752 | 1752 | break |
|
1753 | 1753 | |
|
1754 | 1754 | return None |
|
1755 | 1755 | |
|
1756 | 1756 | def completions(self, text: str, offset: int)->Iterator[Completion]: |
|
1757 | 1757 | """ |
|
1758 | 1758 | Returns an iterator over the possible completions |
|
1759 | 1759 | |
|
1760 | 1760 | .. warning:: Unstable |
|
1761 | 1761 | |
|
1762 | 1762 | This function is unstable, API may change without warning. |
|
1763 | 1763 | It will also raise unless use in proper context manager. |
|
1764 | 1764 | |
|
1765 | 1765 | Parameters |
|
1766 | 1766 | ---------- |
|
1767 | 1767 | |
|
1768 | 1768 | text:str |
|
1769 | 1769 | Full text of the current input, multi line string. |
|
1770 | 1770 | offset:int |
|
1771 | 1771 | Integer representing the position of the cursor in ``text``. Offset |
|
1772 | 1772 | is 0-based indexed. |
|
1773 | 1773 | |
|
1774 | 1774 | Yields |
|
1775 | 1775 | ------ |
|
1776 | 1776 | :any:`Completion` object |
|
1777 | 1777 | |
|
1778 | 1778 | |
|
1779 | 1779 | The cursor on a text can either be seen as being "in between" |
|
1780 | 1780 | characters or "On" a character depending on the interface visible to |
|
1781 | 1781 | the user. For consistency the cursor being on "in between" characters X |
|
1782 | 1782 | and Y is equivalent to the cursor being "on" character Y, that is to say |
|
1783 | 1783 | the character the cursor is on is considered as being after the cursor. |
|
1784 | 1784 | |
|
1785 | 1785 | Combining characters may span more that one position in the |
|
1786 | 1786 | text. |
|
1787 | 1787 | |
|
1788 | 1788 | |
|
1789 | 1789 | .. note:: |
|
1790 | 1790 | |
|
1791 | 1791 | If ``IPCompleter.debug`` is :any:`True` will yield a ``--jedi/ipython--`` |
|
1792 | 1792 | fake Completion token to distinguish completion returned by Jedi |
|
1793 | 1793 | and usual IPython completion. |
|
1794 | 1794 | |
|
1795 | 1795 | .. note:: |
|
1796 | 1796 | |
|
1797 | 1797 | Completions are not completely deduplicated yet. If identical |
|
1798 | 1798 | completions are coming from different sources this function does not |
|
1799 | 1799 | ensure that each completion object will only be present once. |
|
1800 | 1800 | """ |
|
1801 | 1801 | warnings.warn("_complete is a provisional API (as of IPython 6.0). " |
|
1802 | 1802 | "It may change without warnings. " |
|
1803 | 1803 | "Use in corresponding context manager.", |
|
1804 | 1804 | category=ProvisionalCompleterWarning, stacklevel=2) |
|
1805 | 1805 | |
|
1806 | 1806 | seen = set() |
|
1807 | 1807 | try: |
|
1808 | 1808 | for c in self._completions(text, offset, _timeout=self.jedi_compute_type_timeout/1000): |
|
1809 | 1809 | if c and (c in seen): |
|
1810 | 1810 | continue |
|
1811 | 1811 | yield c |
|
1812 | 1812 | seen.add(c) |
|
1813 | 1813 | except KeyboardInterrupt: |
|
1814 | 1814 | """if completions take too long and users send keyboard interrupt, |
|
1815 | 1815 | do not crash and return ASAP. """ |
|
1816 | 1816 | pass |
|
1817 | 1817 | |
|
1818 | 1818 | def _completions(self, full_text: str, offset: int, *, _timeout)->Iterator[Completion]: |
|
1819 | 1819 | """ |
|
1820 | 1820 | Core completion module.Same signature as :any:`completions`, with the |
|
1821 | 1821 | extra `timeout` parameter (in seconds). |
|
1822 | 1822 | |
|
1823 | 1823 | |
|
1824 | 1824 | Computing jedi's completion ``.type`` can be quite expensive (it is a |
|
1825 | 1825 | lazy property) and can require some warm-up, more warm up than just |
|
1826 | 1826 | computing the ``name`` of a completion. The warm-up can be : |
|
1827 | 1827 | |
|
1828 | 1828 | - Long warm-up the first time a module is encountered after |
|
1829 | 1829 | install/update: actually build parse/inference tree. |
|
1830 | 1830 | |
|
1831 | 1831 | - first time the module is encountered in a session: load tree from |
|
1832 | 1832 | disk. |
|
1833 | 1833 | |
|
1834 | 1834 | We don't want to block completions for tens of seconds so we give the |
|
1835 | 1835 | completer a "budget" of ``_timeout`` seconds per invocation to compute |
|
1836 | 1836 | completions types, the completions that have not yet been computed will |
|
1837 | 1837 | be marked as "unknown" an will have a chance to be computed next round |
|
1838 | 1838 | are things get cached. |
|
1839 | 1839 | |
|
1840 | 1840 | Keep in mind that Jedi is not the only thing treating the completion so |
|
1841 | 1841 | keep the timeout short-ish as if we take more than 0.3 second we still |
|
1842 | 1842 | have lots of processing to do. |
|
1843 | 1843 | |
|
1844 | 1844 | """ |
|
1845 | 1845 | deadline = time.monotonic() + _timeout |
|
1846 | 1846 | |
|
1847 | 1847 | |
|
1848 | 1848 | before = full_text[:offset] |
|
1849 | 1849 | cursor_line, cursor_column = position_to_cursor(full_text, offset) |
|
1850 | 1850 | |
|
1851 | 1851 | matched_text, matches, matches_origin, jedi_matches = self._complete( |
|
1852 | 1852 | full_text=full_text, cursor_line=cursor_line, cursor_pos=cursor_column) |
|
1853 | 1853 | |
|
1854 | 1854 | iter_jm = iter(jedi_matches) |
|
1855 | 1855 | if _timeout: |
|
1856 | 1856 | for jm in iter_jm: |
|
1857 | 1857 | try: |
|
1858 | 1858 | type_ = jm.type |
|
1859 | 1859 | except Exception: |
|
1860 | 1860 | if self.debug: |
|
1861 | 1861 | print("Error in Jedi getting type of ", jm) |
|
1862 | 1862 | type_ = None |
|
1863 | 1863 | delta = len(jm.name_with_symbols) - len(jm.complete) |
|
1864 | 1864 | if type_ == 'function': |
|
1865 | 1865 | signature = _make_signature(jm) |
|
1866 | 1866 | else: |
|
1867 | 1867 | signature = '' |
|
1868 | 1868 | yield Completion(start=offset - delta, |
|
1869 | 1869 | end=offset, |
|
1870 | 1870 | text=jm.name_with_symbols, |
|
1871 | 1871 | type=type_, |
|
1872 | 1872 | signature=signature, |
|
1873 | 1873 | _origin='jedi') |
|
1874 | 1874 | |
|
1875 | 1875 | if time.monotonic() > deadline: |
|
1876 | 1876 | break |
|
1877 | 1877 | |
|
1878 | 1878 | for jm in iter_jm: |
|
1879 | 1879 | delta = len(jm.name_with_symbols) - len(jm.complete) |
|
1880 | 1880 | yield Completion(start=offset - delta, |
|
1881 | 1881 | end=offset, |
|
1882 | 1882 | text=jm.name_with_symbols, |
|
1883 | 1883 | type='<unknown>', # don't compute type for speed |
|
1884 | 1884 | _origin='jedi', |
|
1885 | 1885 | signature='') |
|
1886 | 1886 | |
|
1887 | 1887 | |
|
1888 | 1888 | start_offset = before.rfind(matched_text) |
|
1889 | 1889 | |
|
1890 | 1890 | # TODO: |
|
1891 | 1891 | # Suppress this, right now just for debug. |
|
1892 | 1892 | if jedi_matches and matches and self.debug: |
|
1893 | 1893 | yield Completion(start=start_offset, end=offset, text='--jedi/ipython--', |
|
1894 | 1894 | _origin='debug', type='none', signature='') |
|
1895 | 1895 | |
|
1896 | 1896 | # I'm unsure if this is always true, so let's assert and see if it |
|
1897 | 1897 | # crash |
|
1898 | 1898 | assert before.endswith(matched_text) |
|
1899 | 1899 | for m, t in zip(matches, matches_origin): |
|
1900 | 1900 | yield Completion(start=start_offset, end=offset, text=m, _origin=t, signature='', type='<unknown>') |
|
1901 | 1901 | |
|
1902 | 1902 | |
|
1903 | 1903 | def complete(self, text=None, line_buffer=None, cursor_pos=None): |
|
1904 | 1904 | """Find completions for the given text and line context. |
|
1905 | 1905 | |
|
1906 | 1906 | Note that both the text and the line_buffer are optional, but at least |
|
1907 | 1907 | one of them must be given. |
|
1908 | 1908 | |
|
1909 | 1909 | Parameters |
|
1910 | 1910 | ---------- |
|
1911 | 1911 | text : string, optional |
|
1912 | 1912 | Text to perform the completion on. If not given, the line buffer |
|
1913 | 1913 | is split using the instance's CompletionSplitter object. |
|
1914 | 1914 | |
|
1915 | 1915 | line_buffer : string, optional |
|
1916 | 1916 | If not given, the completer attempts to obtain the current line |
|
1917 | 1917 | buffer via readline. This keyword allows clients which are |
|
1918 | 1918 | requesting for text completions in non-readline contexts to inform |
|
1919 | 1919 | the completer of the entire text. |
|
1920 | 1920 | |
|
1921 | 1921 | cursor_pos : int, optional |
|
1922 | 1922 | Index of the cursor in the full line buffer. Should be provided by |
|
1923 | 1923 | remote frontends where kernel has no access to frontend state. |
|
1924 | 1924 | |
|
1925 | 1925 | Returns |
|
1926 | 1926 | ------- |
|
1927 | 1927 | text : str |
|
1928 | 1928 | Text that was actually used in the completion. |
|
1929 | 1929 | |
|
1930 | 1930 | matches : list |
|
1931 | 1931 | A list of completion matches. |
|
1932 | 1932 | |
|
1933 | 1933 | |
|
1934 | 1934 | .. note:: |
|
1935 | 1935 | |
|
1936 | 1936 | This API is likely to be deprecated and replaced by |
|
1937 | 1937 | :any:`IPCompleter.completions` in the future. |
|
1938 | 1938 | |
|
1939 | 1939 | |
|
1940 | 1940 | """ |
|
1941 | 1941 | warnings.warn('`Completer.complete` is pending deprecation since ' |
|
1942 | 1942 | 'IPython 6.0 and will be replaced by `Completer.completions`.', |
|
1943 | 1943 | PendingDeprecationWarning) |
|
1944 | 1944 | # potential todo, FOLD the 3rd throw away argument of _complete |
|
1945 | 1945 | # into the first 2 one. |
|
1946 | 1946 | return self._complete(line_buffer=line_buffer, cursor_pos=cursor_pos, text=text, cursor_line=0)[:2] |
|
1947 | 1947 | |
|
1948 | 1948 | def _complete(self, *, cursor_line, cursor_pos, line_buffer=None, text=None, |
|
1949 | 1949 | full_text=None) -> Tuple[str, List[str], List[str], Iterable[_FakeJediCompletion]]: |
|
1950 | 1950 | """ |
|
1951 | 1951 | |
|
1952 | 1952 | Like complete but can also returns raw jedi completions as well as the |
|
1953 | 1953 | origin of the completion text. This could (and should) be made much |
|
1954 | 1954 | cleaner but that will be simpler once we drop the old (and stateful) |
|
1955 | 1955 | :any:`complete` API. |
|
1956 | 1956 | |
|
1957 | 1957 | |
|
1958 | 1958 | With current provisional API, cursor_pos act both (depending on the |
|
1959 | 1959 | caller) as the offset in the ``text`` or ``line_buffer``, or as the |
|
1960 | 1960 | ``column`` when passing multiline strings this could/should be renamed |
|
1961 | 1961 | but would add extra noise. |
|
1962 | 1962 | """ |
|
1963 | 1963 | |
|
1964 | 1964 | # if the cursor position isn't given, the only sane assumption we can |
|
1965 | 1965 | # make is that it's at the end of the line (the common case) |
|
1966 | 1966 | if cursor_pos is None: |
|
1967 | 1967 | cursor_pos = len(line_buffer) if text is None else len(text) |
|
1968 | 1968 | |
|
1969 | 1969 | if self.use_main_ns: |
|
1970 | 1970 | self.namespace = __main__.__dict__ |
|
1971 | 1971 | |
|
1972 | 1972 | # if text is either None or an empty string, rely on the line buffer |
|
1973 | 1973 | if (not line_buffer) and full_text: |
|
1974 | 1974 | line_buffer = full_text.split('\n')[cursor_line] |
|
1975 | 1975 | if not text: |
|
1976 | 1976 | text = self.splitter.split_line(line_buffer, cursor_pos) |
|
1977 | 1977 | |
|
1978 | 1978 | if self.backslash_combining_completions: |
|
1979 | 1979 | # allow deactivation of these on windows. |
|
1980 | 1980 | base_text = text if not line_buffer else line_buffer[:cursor_pos] |
|
1981 | 1981 | latex_text, latex_matches = self.latex_matches(base_text) |
|
1982 | 1982 | if latex_matches: |
|
1983 | 1983 | return latex_text, latex_matches, ['latex_matches']*len(latex_matches), () |
|
1984 | 1984 | name_text = '' |
|
1985 | 1985 | name_matches = [] |
|
1986 | 1986 | for meth in (self.unicode_name_matches, back_latex_name_matches, back_unicode_name_matches): |
|
1987 | 1987 | name_text, name_matches = meth(base_text) |
|
1988 | 1988 | if name_text: |
|
1989 | 1989 | return name_text, name_matches[:MATCHES_LIMIT], \ |
|
1990 | 1990 | [meth.__qualname__]*min(len(name_matches), MATCHES_LIMIT), () |
|
1991 | 1991 | |
|
1992 | 1992 | |
|
1993 | 1993 | # If no line buffer is given, assume the input text is all there was |
|
1994 | 1994 | if line_buffer is None: |
|
1995 | 1995 | line_buffer = text |
|
1996 | 1996 | |
|
1997 | 1997 | self.line_buffer = line_buffer |
|
1998 | 1998 | self.text_until_cursor = self.line_buffer[:cursor_pos] |
|
1999 | 1999 | |
|
2000 | 2000 | # Do magic arg matches |
|
2001 | 2001 | for matcher in self.magic_arg_matchers: |
|
2002 | 2002 | matches = list(matcher(line_buffer))[:MATCHES_LIMIT] |
|
2003 | 2003 | if matches: |
|
2004 | 2004 | origins = [matcher.__qualname__] * len(matches) |
|
2005 | 2005 | return text, matches, origins, () |
|
2006 | 2006 | |
|
2007 | 2007 | # Start with a clean slate of completions |
|
2008 | 2008 | matches = [] |
|
2009 | 2009 | custom_res = self.dispatch_custom_completer(text) |
|
2010 | 2010 | # FIXME: we should extend our api to return a dict with completions for |
|
2011 | 2011 | # different types of objects. The rlcomplete() method could then |
|
2012 | 2012 | # simply collapse the dict into a list for readline, but we'd have |
|
2013 | 2013 | # richer completion semantics in other environments. |
|
2014 | 2014 | completions = () |
|
2015 | 2015 | if self.use_jedi: |
|
2016 | 2016 | if not full_text: |
|
2017 | 2017 | full_text = line_buffer |
|
2018 | 2018 | completions = self._jedi_matches( |
|
2019 | 2019 | cursor_pos, cursor_line, full_text) |
|
2020 | 2020 | if custom_res is not None: |
|
2021 | 2021 | # did custom completers produce something? |
|
2022 | 2022 | matches = [(m, 'custom') for m in custom_res] |
|
2023 | 2023 | else: |
|
2024 | 2024 | # Extend the list of completions with the results of each |
|
2025 | 2025 | # matcher, so we return results to the user from all |
|
2026 | 2026 | # namespaces. |
|
2027 | 2027 | if self.merge_completions: |
|
2028 | 2028 | matches = [] |
|
2029 | 2029 | for matcher in self.matchers: |
|
2030 | 2030 | try: |
|
2031 | 2031 | matches.extend([(m, matcher.__qualname__) |
|
2032 | 2032 | for m in matcher(text)]) |
|
2033 | 2033 | except: |
|
2034 | 2034 | # Show the ugly traceback if the matcher causes an |
|
2035 | 2035 | # exception, but do NOT crash the kernel! |
|
2036 | 2036 | sys.excepthook(*sys.exc_info()) |
|
2037 | 2037 | else: |
|
2038 | 2038 | for matcher in self.matchers: |
|
2039 | 2039 | matches = [(m, matcher.__qualname__) |
|
2040 | 2040 | for m in matcher(text)] |
|
2041 | 2041 | if matches: |
|
2042 | 2042 | break |
|
2043 | 2043 | seen = set() |
|
2044 | 2044 | filtered_matches = set() |
|
2045 | 2045 | for m in matches: |
|
2046 | 2046 | t, c = m |
|
2047 | 2047 | if t not in seen: |
|
2048 | 2048 | filtered_matches.add(m) |
|
2049 | 2049 | seen.add(t) |
|
2050 | 2050 | |
|
2051 | 2051 | _filtered_matches = sorted( |
|
2052 | 2052 | set(filtered_matches), key=lambda x: completions_sorting_key(x[0]))\ |
|
2053 | 2053 | [:MATCHES_LIMIT] |
|
2054 | 2054 | |
|
2055 | 2055 | _matches = [m[0] for m in _filtered_matches] |
|
2056 | 2056 | origins = [m[1] for m in _filtered_matches] |
|
2057 | 2057 | |
|
2058 | 2058 | self.matches = _matches |
|
2059 | 2059 | |
|
2060 | 2060 | return text, _matches, origins, completions |
@@ -1,645 +1,645 b'' | |||
|
1 | 1 | # -*- coding: utf-8 -*- |
|
2 | 2 | """ |
|
3 | 3 | Pdb debugger class. |
|
4 | 4 | |
|
5 | 5 | Modified from the standard pdb.Pdb class to avoid including readline, so that |
|
6 | 6 | the command line completion of other programs which include this isn't |
|
7 | 7 | damaged. |
|
8 | 8 | |
|
9 | 9 | In the future, this class will be expanded with improvements over the standard |
|
10 | 10 | pdb. |
|
11 | 11 | |
|
12 | 12 | The code in this file is mainly lifted out of cmd.py in Python 2.2, with minor |
|
13 | 13 | changes. Licensing should therefore be under the standard Python terms. For |
|
14 | 14 | details on the PSF (Python Software Foundation) standard license, see: |
|
15 | 15 | |
|
16 | 16 | https://docs.python.org/2/license.html |
|
17 | 17 | """ |
|
18 | 18 | |
|
19 | 19 | #***************************************************************************** |
|
20 | 20 | # |
|
21 | 21 | # This file is licensed under the PSF license. |
|
22 | 22 | # |
|
23 | 23 | # Copyright (C) 2001 Python Software Foundation, www.python.org |
|
24 | 24 | # Copyright (C) 2005-2006 Fernando Perez. <fperez@colorado.edu> |
|
25 | 25 | # |
|
26 | 26 | # |
|
27 | 27 | #***************************************************************************** |
|
28 | 28 | |
|
29 | 29 | import bdb |
|
30 | 30 | import functools |
|
31 | 31 | import inspect |
|
32 | 32 | import linecache |
|
33 | 33 | import sys |
|
34 | 34 | import warnings |
|
35 | 35 | import re |
|
36 | 36 | |
|
37 | 37 | from IPython import get_ipython |
|
38 | 38 | from IPython.utils import PyColorize |
|
39 | 39 | from IPython.utils import coloransi, py3compat |
|
40 | 40 | from IPython.core.excolors import exception_colors |
|
41 | 41 | from IPython.testing.skipdoctest import skip_doctest |
|
42 | 42 | |
|
43 | 43 | |
|
44 | 44 | prompt = 'ipdb> ' |
|
45 | 45 | |
|
46 | 46 | #We have to check this directly from sys.argv, config struct not yet available |
|
47 | 47 | from pdb import Pdb as OldPdb |
|
48 | 48 | |
|
49 | 49 | # Allow the set_trace code to operate outside of an ipython instance, even if |
|
50 | 50 | # it does so with some limitations. The rest of this support is implemented in |
|
51 | 51 | # the Tracer constructor. |
|
52 | 52 | |
|
53 | 53 | def make_arrow(pad): |
|
54 | 54 | """generate the leading arrow in front of traceback or debugger""" |
|
55 | 55 | if pad >= 2: |
|
56 | 56 | return '-'*(pad-2) + '> ' |
|
57 | 57 | elif pad == 1: |
|
58 | 58 | return '>' |
|
59 | 59 | return '' |
|
60 | 60 | |
|
61 | 61 | |
|
62 | 62 | def BdbQuit_excepthook(et, ev, tb, excepthook=None): |
|
63 | 63 | """Exception hook which handles `BdbQuit` exceptions. |
|
64 | 64 | |
|
65 | 65 | All other exceptions are processed using the `excepthook` |
|
66 | 66 | parameter. |
|
67 | 67 | """ |
|
68 | 68 | warnings.warn("`BdbQuit_excepthook` is deprecated since version 5.1", |
|
69 | 69 | DeprecationWarning, stacklevel=2) |
|
70 | 70 | if et==bdb.BdbQuit: |
|
71 | 71 | print('Exiting Debugger.') |
|
72 | 72 | elif excepthook is not None: |
|
73 | 73 | excepthook(et, ev, tb) |
|
74 | 74 | else: |
|
75 | 75 | # Backwards compatibility. Raise deprecation warning? |
|
76 | 76 | BdbQuit_excepthook.excepthook_ori(et,ev,tb) |
|
77 | 77 | |
|
78 | 78 | |
|
79 | 79 | def BdbQuit_IPython_excepthook(self,et,ev,tb,tb_offset=None): |
|
80 | 80 | warnings.warn( |
|
81 | 81 | "`BdbQuit_IPython_excepthook` is deprecated since version 5.1", |
|
82 | 82 | DeprecationWarning, stacklevel=2) |
|
83 | 83 | print('Exiting Debugger.') |
|
84 | 84 | |
|
85 | 85 | |
|
86 | 86 | class Tracer(object): |
|
87 | 87 | """ |
|
88 | 88 | DEPRECATED |
|
89 | 89 | |
|
90 | 90 | Class for local debugging, similar to pdb.set_trace. |
|
91 | 91 | |
|
92 | 92 | Instances of this class, when called, behave like pdb.set_trace, but |
|
93 | 93 | providing IPython's enhanced capabilities. |
|
94 | 94 | |
|
95 | 95 | This is implemented as a class which must be initialized in your own code |
|
96 | 96 | and not as a standalone function because we need to detect at runtime |
|
97 | 97 | whether IPython is already active or not. That detection is done in the |
|
98 | 98 | constructor, ensuring that this code plays nicely with a running IPython, |
|
99 | 99 | while functioning acceptably (though with limitations) if outside of it. |
|
100 | 100 | """ |
|
101 | 101 | |
|
102 | 102 | @skip_doctest |
|
103 | 103 | def __init__(self, colors=None): |
|
104 | 104 | """ |
|
105 | 105 | DEPRECATED |
|
106 | 106 | |
|
107 | 107 | Create a local debugger instance. |
|
108 | 108 | |
|
109 | 109 | Parameters |
|
110 | 110 | ---------- |
|
111 | 111 | |
|
112 | 112 | colors : str, optional |
|
113 | 113 | The name of the color scheme to use, it must be one of IPython's |
|
114 | 114 | valid color schemes. If not given, the function will default to |
|
115 | 115 | the current IPython scheme when running inside IPython, and to |
|
116 | 116 | 'NoColor' otherwise. |
|
117 | 117 | |
|
118 | 118 | Examples |
|
119 | 119 | -------- |
|
120 | 120 | :: |
|
121 | 121 | |
|
122 | 122 | from IPython.core.debugger import Tracer; debug_here = Tracer() |
|
123 | 123 | |
|
124 | 124 | Later in your code:: |
|
125 | 125 | |
|
126 | 126 | debug_here() # -> will open up the debugger at that point. |
|
127 | 127 | |
|
128 | 128 | Once the debugger activates, you can use all of its regular commands to |
|
129 | 129 | step through code, set breakpoints, etc. See the pdb documentation |
|
130 | 130 | from the Python standard library for usage details. |
|
131 | 131 | """ |
|
132 | 132 | warnings.warn("`Tracer` is deprecated since version 5.1, directly use " |
|
133 | 133 | "`IPython.core.debugger.Pdb.set_trace()`", |
|
134 | 134 | DeprecationWarning, stacklevel=2) |
|
135 | 135 | |
|
136 | 136 | ip = get_ipython() |
|
137 | 137 | if ip is None: |
|
138 | 138 | # Outside of ipython, we set our own exception hook manually |
|
139 | 139 | sys.excepthook = functools.partial(BdbQuit_excepthook, |
|
140 | 140 | excepthook=sys.excepthook) |
|
141 | 141 | def_colors = 'NoColor' |
|
142 | 142 | else: |
|
143 | 143 | # In ipython, we use its custom exception handler mechanism |
|
144 | 144 | def_colors = ip.colors |
|
145 | 145 | ip.set_custom_exc((bdb.BdbQuit,), BdbQuit_IPython_excepthook) |
|
146 | 146 | |
|
147 | 147 | if colors is None: |
|
148 | 148 | colors = def_colors |
|
149 | 149 | |
|
150 | 150 | # The stdlib debugger internally uses a modified repr from the `repr` |
|
151 | 151 | # module, that limits the length of printed strings to a hardcoded |
|
152 | 152 | # limit of 30 characters. That much trimming is too aggressive, let's |
|
153 | 153 | # at least raise that limit to 80 chars, which should be enough for |
|
154 | 154 | # most interactive uses. |
|
155 | 155 | try: |
|
156 | 156 | try: |
|
157 | 157 | from reprlib import aRepr # Py 3 |
|
158 | 158 | except ImportError: |
|
159 | 159 | from repr import aRepr # Py 2 |
|
160 | 160 | aRepr.maxstring = 80 |
|
161 | 161 | except: |
|
162 | 162 | # This is only a user-facing convenience, so any error we encounter |
|
163 | 163 | # here can be warned about but can be otherwise ignored. These |
|
164 | 164 | # printouts will tell us about problems if this API changes |
|
165 | 165 | import traceback |
|
166 | 166 | traceback.print_exc() |
|
167 | 167 | |
|
168 | 168 | self.debugger = Pdb(colors) |
|
169 | 169 | |
|
170 | 170 | def __call__(self): |
|
171 | 171 | """Starts an interactive debugger at the point where called. |
|
172 | 172 | |
|
173 | 173 | This is similar to the pdb.set_trace() function from the std lib, but |
|
174 | 174 | using IPython's enhanced debugger.""" |
|
175 | 175 | |
|
176 | 176 | self.debugger.set_trace(sys._getframe().f_back) |
|
177 | 177 | |
|
178 | 178 | |
|
179 | RGX_EXTRA_INDENT = re.compile('(?<=\n)\s+') | |
|
179 | RGX_EXTRA_INDENT = re.compile(r'(?<=\n)\s+') | |
|
180 | 180 | |
|
181 | 181 | |
|
182 | 182 | def strip_indentation(multiline_string): |
|
183 | 183 | return RGX_EXTRA_INDENT.sub('', multiline_string) |
|
184 | 184 | |
|
185 | 185 | |
|
186 | 186 | def decorate_fn_with_doc(new_fn, old_fn, additional_text=""): |
|
187 | 187 | """Make new_fn have old_fn's doc string. This is particularly useful |
|
188 | 188 | for the ``do_...`` commands that hook into the help system. |
|
189 | 189 | Adapted from from a comp.lang.python posting |
|
190 | 190 | by Duncan Booth.""" |
|
191 | 191 | def wrapper(*args, **kw): |
|
192 | 192 | return new_fn(*args, **kw) |
|
193 | 193 | if old_fn.__doc__: |
|
194 | 194 | wrapper.__doc__ = strip_indentation(old_fn.__doc__) + additional_text |
|
195 | 195 | return wrapper |
|
196 | 196 | |
|
197 | 197 | |
|
198 | 198 | def _file_lines(fname): |
|
199 | 199 | """Return the contents of a named file as a list of lines. |
|
200 | 200 | |
|
201 | 201 | This function never raises an IOError exception: if the file can't be |
|
202 | 202 | read, it simply returns an empty list.""" |
|
203 | 203 | |
|
204 | 204 | try: |
|
205 | 205 | outfile = open(fname) |
|
206 | 206 | except IOError: |
|
207 | 207 | return [] |
|
208 | 208 | else: |
|
209 | 209 | out = outfile.readlines() |
|
210 | 210 | outfile.close() |
|
211 | 211 | return out |
|
212 | 212 | |
|
213 | 213 | |
|
214 | 214 | class Pdb(OldPdb): |
|
215 | 215 | """Modified Pdb class, does not load readline. |
|
216 | 216 | |
|
217 | 217 | for a standalone version that uses prompt_toolkit, see |
|
218 | 218 | `IPython.terminal.debugger.TerminalPdb` and |
|
219 | 219 | `IPython.terminal.debugger.set_trace()` |
|
220 | 220 | """ |
|
221 | 221 | |
|
222 | 222 | def __init__(self, color_scheme=None, completekey=None, |
|
223 | 223 | stdin=None, stdout=None, context=5): |
|
224 | 224 | |
|
225 | 225 | # Parent constructor: |
|
226 | 226 | try: |
|
227 | 227 | self.context = int(context) |
|
228 | 228 | if self.context <= 0: |
|
229 | 229 | raise ValueError("Context must be a positive integer") |
|
230 | 230 | except (TypeError, ValueError): |
|
231 | 231 | raise ValueError("Context must be a positive integer") |
|
232 | 232 | |
|
233 | 233 | OldPdb.__init__(self, completekey, stdin, stdout) |
|
234 | 234 | |
|
235 | 235 | # IPython changes... |
|
236 | 236 | self.shell = get_ipython() |
|
237 | 237 | |
|
238 | 238 | if self.shell is None: |
|
239 | 239 | save_main = sys.modules['__main__'] |
|
240 | 240 | # No IPython instance running, we must create one |
|
241 | 241 | from IPython.terminal.interactiveshell import \ |
|
242 | 242 | TerminalInteractiveShell |
|
243 | 243 | self.shell = TerminalInteractiveShell.instance() |
|
244 | 244 | # needed by any code which calls __import__("__main__") after |
|
245 | 245 | # the debugger was entered. See also #9941. |
|
246 | 246 | sys.modules['__main__'] = save_main |
|
247 | 247 | |
|
248 | 248 | if color_scheme is not None: |
|
249 | 249 | warnings.warn( |
|
250 | 250 | "The `color_scheme` argument is deprecated since version 5.1", |
|
251 | 251 | DeprecationWarning, stacklevel=2) |
|
252 | 252 | else: |
|
253 | 253 | color_scheme = self.shell.colors |
|
254 | 254 | |
|
255 | 255 | self.aliases = {} |
|
256 | 256 | |
|
257 | 257 | # Create color table: we copy the default one from the traceback |
|
258 | 258 | # module and add a few attributes needed for debugging |
|
259 | 259 | self.color_scheme_table = exception_colors() |
|
260 | 260 | |
|
261 | 261 | # shorthands |
|
262 | 262 | C = coloransi.TermColors |
|
263 | 263 | cst = self.color_scheme_table |
|
264 | 264 | |
|
265 | 265 | cst['NoColor'].colors.prompt = C.NoColor |
|
266 | 266 | cst['NoColor'].colors.breakpoint_enabled = C.NoColor |
|
267 | 267 | cst['NoColor'].colors.breakpoint_disabled = C.NoColor |
|
268 | 268 | |
|
269 | 269 | cst['Linux'].colors.prompt = C.Green |
|
270 | 270 | cst['Linux'].colors.breakpoint_enabled = C.LightRed |
|
271 | 271 | cst['Linux'].colors.breakpoint_disabled = C.Red |
|
272 | 272 | |
|
273 | 273 | cst['LightBG'].colors.prompt = C.Blue |
|
274 | 274 | cst['LightBG'].colors.breakpoint_enabled = C.LightRed |
|
275 | 275 | cst['LightBG'].colors.breakpoint_disabled = C.Red |
|
276 | 276 | |
|
277 | 277 | cst['Neutral'].colors.prompt = C.Blue |
|
278 | 278 | cst['Neutral'].colors.breakpoint_enabled = C.LightRed |
|
279 | 279 | cst['Neutral'].colors.breakpoint_disabled = C.Red |
|
280 | 280 | |
|
281 | 281 | |
|
282 | 282 | # Add a python parser so we can syntax highlight source while |
|
283 | 283 | # debugging. |
|
284 | 284 | self.parser = PyColorize.Parser(style=color_scheme) |
|
285 | 285 | self.set_colors(color_scheme) |
|
286 | 286 | |
|
287 | 287 | # Set the prompt - the default prompt is '(Pdb)' |
|
288 | 288 | self.prompt = prompt |
|
289 | 289 | |
|
290 | 290 | def set_colors(self, scheme): |
|
291 | 291 | """Shorthand access to the color table scheme selector method.""" |
|
292 | 292 | self.color_scheme_table.set_active_scheme(scheme) |
|
293 | 293 | self.parser.style = scheme |
|
294 | 294 | |
|
295 | 295 | def interaction(self, frame, traceback): |
|
296 | 296 | try: |
|
297 | 297 | OldPdb.interaction(self, frame, traceback) |
|
298 | 298 | except KeyboardInterrupt: |
|
299 | 299 | sys.stdout.write('\n' + self.shell.get_exception_only()) |
|
300 | 300 | |
|
301 | 301 | def new_do_up(self, arg): |
|
302 | 302 | OldPdb.do_up(self, arg) |
|
303 | 303 | do_u = do_up = decorate_fn_with_doc(new_do_up, OldPdb.do_up) |
|
304 | 304 | |
|
305 | 305 | def new_do_down(self, arg): |
|
306 | 306 | OldPdb.do_down(self, arg) |
|
307 | 307 | |
|
308 | 308 | do_d = do_down = decorate_fn_with_doc(new_do_down, OldPdb.do_down) |
|
309 | 309 | |
|
310 | 310 | def new_do_frame(self, arg): |
|
311 | 311 | OldPdb.do_frame(self, arg) |
|
312 | 312 | |
|
313 | 313 | def new_do_quit(self, arg): |
|
314 | 314 | |
|
315 | 315 | if hasattr(self, 'old_all_completions'): |
|
316 | 316 | self.shell.Completer.all_completions=self.old_all_completions |
|
317 | 317 | |
|
318 | 318 | return OldPdb.do_quit(self, arg) |
|
319 | 319 | |
|
320 | 320 | do_q = do_quit = decorate_fn_with_doc(new_do_quit, OldPdb.do_quit) |
|
321 | 321 | |
|
322 | 322 | def new_do_restart(self, arg): |
|
323 | 323 | """Restart command. In the context of ipython this is exactly the same |
|
324 | 324 | thing as 'quit'.""" |
|
325 | 325 | self.msg("Restart doesn't make sense here. Using 'quit' instead.") |
|
326 | 326 | return self.do_quit(arg) |
|
327 | 327 | |
|
328 | 328 | def print_stack_trace(self, context=None): |
|
329 | 329 | if context is None: |
|
330 | 330 | context = self.context |
|
331 | 331 | try: |
|
332 | 332 | context=int(context) |
|
333 | 333 | if context <= 0: |
|
334 | 334 | raise ValueError("Context must be a positive integer") |
|
335 | 335 | except (TypeError, ValueError): |
|
336 | 336 | raise ValueError("Context must be a positive integer") |
|
337 | 337 | try: |
|
338 | 338 | for frame_lineno in self.stack: |
|
339 | 339 | self.print_stack_entry(frame_lineno, context=context) |
|
340 | 340 | except KeyboardInterrupt: |
|
341 | 341 | pass |
|
342 | 342 | |
|
343 | 343 | def print_stack_entry(self,frame_lineno, prompt_prefix='\n-> ', |
|
344 | 344 | context=None): |
|
345 | 345 | if context is None: |
|
346 | 346 | context = self.context |
|
347 | 347 | try: |
|
348 | 348 | context=int(context) |
|
349 | 349 | if context <= 0: |
|
350 | 350 | raise ValueError("Context must be a positive integer") |
|
351 | 351 | except (TypeError, ValueError): |
|
352 | 352 | raise ValueError("Context must be a positive integer") |
|
353 | 353 | print(self.format_stack_entry(frame_lineno, '', context)) |
|
354 | 354 | |
|
355 | 355 | # vds: >> |
|
356 | 356 | frame, lineno = frame_lineno |
|
357 | 357 | filename = frame.f_code.co_filename |
|
358 | 358 | self.shell.hooks.synchronize_with_editor(filename, lineno, 0) |
|
359 | 359 | # vds: << |
|
360 | 360 | |
|
361 | 361 | def format_stack_entry(self, frame_lineno, lprefix=': ', context=None): |
|
362 | 362 | if context is None: |
|
363 | 363 | context = self.context |
|
364 | 364 | try: |
|
365 | 365 | context=int(context) |
|
366 | 366 | if context <= 0: |
|
367 | 367 | print("Context must be a positive integer") |
|
368 | 368 | except (TypeError, ValueError): |
|
369 | 369 | print("Context must be a positive integer") |
|
370 | 370 | try: |
|
371 | 371 | import reprlib # Py 3 |
|
372 | 372 | except ImportError: |
|
373 | 373 | import repr as reprlib # Py 2 |
|
374 | 374 | |
|
375 | 375 | ret = [] |
|
376 | 376 | |
|
377 | 377 | Colors = self.color_scheme_table.active_colors |
|
378 | 378 | ColorsNormal = Colors.Normal |
|
379 | 379 | tpl_link = u'%s%%s%s' % (Colors.filenameEm, ColorsNormal) |
|
380 | 380 | tpl_call = u'%s%%s%s%%s%s' % (Colors.vName, Colors.valEm, ColorsNormal) |
|
381 | 381 | tpl_line = u'%%s%s%%s %s%%s' % (Colors.lineno, ColorsNormal) |
|
382 | 382 | tpl_line_em = u'%%s%s%%s %s%%s%s' % (Colors.linenoEm, Colors.line, |
|
383 | 383 | ColorsNormal) |
|
384 | 384 | |
|
385 | 385 | frame, lineno = frame_lineno |
|
386 | 386 | |
|
387 | 387 | return_value = '' |
|
388 | 388 | if '__return__' in frame.f_locals: |
|
389 | 389 | rv = frame.f_locals['__return__'] |
|
390 | 390 | #return_value += '->' |
|
391 | 391 | return_value += reprlib.repr(rv) + '\n' |
|
392 | 392 | ret.append(return_value) |
|
393 | 393 | |
|
394 | 394 | #s = filename + '(' + `lineno` + ')' |
|
395 | 395 | filename = self.canonic(frame.f_code.co_filename) |
|
396 | 396 | link = tpl_link % py3compat.cast_unicode(filename) |
|
397 | 397 | |
|
398 | 398 | if frame.f_code.co_name: |
|
399 | 399 | func = frame.f_code.co_name |
|
400 | 400 | else: |
|
401 | 401 | func = "<lambda>" |
|
402 | 402 | |
|
403 | 403 | call = '' |
|
404 | 404 | if func != '?': |
|
405 | 405 | if '__args__' in frame.f_locals: |
|
406 | 406 | args = reprlib.repr(frame.f_locals['__args__']) |
|
407 | 407 | else: |
|
408 | 408 | args = '()' |
|
409 | 409 | call = tpl_call % (func, args) |
|
410 | 410 | |
|
411 | 411 | # The level info should be generated in the same format pdb uses, to |
|
412 | 412 | # avoid breaking the pdbtrack functionality of python-mode in *emacs. |
|
413 | 413 | if frame is self.curframe: |
|
414 | 414 | ret.append('> ') |
|
415 | 415 | else: |
|
416 | 416 | ret.append(' ') |
|
417 | 417 | ret.append(u'%s(%s)%s\n' % (link,lineno,call)) |
|
418 | 418 | |
|
419 | 419 | start = lineno - 1 - context//2 |
|
420 | 420 | lines = linecache.getlines(filename) |
|
421 | 421 | start = min(start, len(lines) - context) |
|
422 | 422 | start = max(start, 0) |
|
423 | 423 | lines = lines[start : start + context] |
|
424 | 424 | |
|
425 | 425 | for i,line in enumerate(lines): |
|
426 | 426 | show_arrow = (start + 1 + i == lineno) |
|
427 | 427 | linetpl = (frame is self.curframe or show_arrow) \ |
|
428 | 428 | and tpl_line_em \ |
|
429 | 429 | or tpl_line |
|
430 | 430 | ret.append(self.__format_line(linetpl, filename, |
|
431 | 431 | start + 1 + i, line, |
|
432 | 432 | arrow = show_arrow) ) |
|
433 | 433 | return ''.join(ret) |
|
434 | 434 | |
|
435 | 435 | def __format_line(self, tpl_line, filename, lineno, line, arrow = False): |
|
436 | 436 | bp_mark = "" |
|
437 | 437 | bp_mark_color = "" |
|
438 | 438 | |
|
439 | 439 | new_line, err = self.parser.format2(line, 'str') |
|
440 | 440 | if not err: |
|
441 | 441 | line = new_line |
|
442 | 442 | |
|
443 | 443 | bp = None |
|
444 | 444 | if lineno in self.get_file_breaks(filename): |
|
445 | 445 | bps = self.get_breaks(filename, lineno) |
|
446 | 446 | bp = bps[-1] |
|
447 | 447 | |
|
448 | 448 | if bp: |
|
449 | 449 | Colors = self.color_scheme_table.active_colors |
|
450 | 450 | bp_mark = str(bp.number) |
|
451 | 451 | bp_mark_color = Colors.breakpoint_enabled |
|
452 | 452 | if not bp.enabled: |
|
453 | 453 | bp_mark_color = Colors.breakpoint_disabled |
|
454 | 454 | |
|
455 | 455 | numbers_width = 7 |
|
456 | 456 | if arrow: |
|
457 | 457 | # This is the line with the error |
|
458 | 458 | pad = numbers_width - len(str(lineno)) - len(bp_mark) |
|
459 | 459 | num = '%s%s' % (make_arrow(pad), str(lineno)) |
|
460 | 460 | else: |
|
461 | 461 | num = '%*s' % (numbers_width - len(bp_mark), str(lineno)) |
|
462 | 462 | |
|
463 | 463 | return tpl_line % (bp_mark_color + bp_mark, num, line) |
|
464 | 464 | |
|
465 | 465 | |
|
466 | 466 | def print_list_lines(self, filename, first, last): |
|
467 | 467 | """The printing (as opposed to the parsing part of a 'list' |
|
468 | 468 | command.""" |
|
469 | 469 | try: |
|
470 | 470 | Colors = self.color_scheme_table.active_colors |
|
471 | 471 | ColorsNormal = Colors.Normal |
|
472 | 472 | tpl_line = '%%s%s%%s %s%%s' % (Colors.lineno, ColorsNormal) |
|
473 | 473 | tpl_line_em = '%%s%s%%s %s%%s%s' % (Colors.linenoEm, Colors.line, ColorsNormal) |
|
474 | 474 | src = [] |
|
475 | 475 | if filename == "<string>" and hasattr(self, "_exec_filename"): |
|
476 | 476 | filename = self._exec_filename |
|
477 | 477 | |
|
478 | 478 | for lineno in range(first, last+1): |
|
479 | 479 | line = linecache.getline(filename, lineno) |
|
480 | 480 | if not line: |
|
481 | 481 | break |
|
482 | 482 | |
|
483 | 483 | if lineno == self.curframe.f_lineno: |
|
484 | 484 | line = self.__format_line(tpl_line_em, filename, lineno, line, arrow = True) |
|
485 | 485 | else: |
|
486 | 486 | line = self.__format_line(tpl_line, filename, lineno, line, arrow = False) |
|
487 | 487 | |
|
488 | 488 | src.append(line) |
|
489 | 489 | self.lineno = lineno |
|
490 | 490 | |
|
491 | 491 | print(''.join(src)) |
|
492 | 492 | |
|
493 | 493 | except KeyboardInterrupt: |
|
494 | 494 | pass |
|
495 | 495 | |
|
496 | 496 | def do_list(self, arg): |
|
497 | 497 | """Print lines of code from the current stack frame |
|
498 | 498 | """ |
|
499 | 499 | self.lastcmd = 'list' |
|
500 | 500 | last = None |
|
501 | 501 | if arg: |
|
502 | 502 | try: |
|
503 | 503 | x = eval(arg, {}, {}) |
|
504 | 504 | if type(x) == type(()): |
|
505 | 505 | first, last = x |
|
506 | 506 | first = int(first) |
|
507 | 507 | last = int(last) |
|
508 | 508 | if last < first: |
|
509 | 509 | # Assume it's a count |
|
510 | 510 | last = first + last |
|
511 | 511 | else: |
|
512 | 512 | first = max(1, int(x) - 5) |
|
513 | 513 | except: |
|
514 | 514 | print('*** Error in argument:', repr(arg)) |
|
515 | 515 | return |
|
516 | 516 | elif self.lineno is None: |
|
517 | 517 | first = max(1, self.curframe.f_lineno - 5) |
|
518 | 518 | else: |
|
519 | 519 | first = self.lineno + 1 |
|
520 | 520 | if last is None: |
|
521 | 521 | last = first + 10 |
|
522 | 522 | self.print_list_lines(self.curframe.f_code.co_filename, first, last) |
|
523 | 523 | |
|
524 | 524 | # vds: >> |
|
525 | 525 | lineno = first |
|
526 | 526 | filename = self.curframe.f_code.co_filename |
|
527 | 527 | self.shell.hooks.synchronize_with_editor(filename, lineno, 0) |
|
528 | 528 | # vds: << |
|
529 | 529 | |
|
530 | 530 | do_l = do_list |
|
531 | 531 | |
|
532 | 532 | def getsourcelines(self, obj): |
|
533 | 533 | lines, lineno = inspect.findsource(obj) |
|
534 | 534 | if inspect.isframe(obj) and obj.f_globals is obj.f_locals: |
|
535 | 535 | # must be a module frame: do not try to cut a block out of it |
|
536 | 536 | return lines, 1 |
|
537 | 537 | elif inspect.ismodule(obj): |
|
538 | 538 | return lines, 1 |
|
539 | 539 | return inspect.getblock(lines[lineno:]), lineno+1 |
|
540 | 540 | |
|
541 | 541 | def do_longlist(self, arg): |
|
542 | 542 | """Print lines of code from the current stack frame. |
|
543 | 543 | |
|
544 | 544 | Shows more lines than 'list' does. |
|
545 | 545 | """ |
|
546 | 546 | self.lastcmd = 'longlist' |
|
547 | 547 | try: |
|
548 | 548 | lines, lineno = self.getsourcelines(self.curframe) |
|
549 | 549 | except OSError as err: |
|
550 | 550 | self.error(err) |
|
551 | 551 | return |
|
552 | 552 | last = lineno + len(lines) |
|
553 | 553 | self.print_list_lines(self.curframe.f_code.co_filename, lineno, last) |
|
554 | 554 | do_ll = do_longlist |
|
555 | 555 | |
|
556 | 556 | def do_debug(self, arg): |
|
557 | 557 | """debug code |
|
558 | 558 | Enter a recursive debugger that steps through the code |
|
559 | 559 | argument (which is an arbitrary expression or statement to be |
|
560 | 560 | executed in the current environment). |
|
561 | 561 | """ |
|
562 | 562 | sys.settrace(None) |
|
563 | 563 | globals = self.curframe.f_globals |
|
564 | 564 | locals = self.curframe_locals |
|
565 | 565 | p = self.__class__(completekey=self.completekey, |
|
566 | 566 | stdin=self.stdin, stdout=self.stdout) |
|
567 | 567 | p.use_rawinput = self.use_rawinput |
|
568 | 568 | p.prompt = "(%s) " % self.prompt.strip() |
|
569 | 569 | self.message("ENTERING RECURSIVE DEBUGGER") |
|
570 | 570 | sys.call_tracing(p.run, (arg, globals, locals)) |
|
571 | 571 | self.message("LEAVING RECURSIVE DEBUGGER") |
|
572 | 572 | sys.settrace(self.trace_dispatch) |
|
573 | 573 | self.lastcmd = p.lastcmd |
|
574 | 574 | |
|
575 | 575 | def do_pdef(self, arg): |
|
576 | 576 | """Print the call signature for any callable object. |
|
577 | 577 | |
|
578 | 578 | The debugger interface to %pdef""" |
|
579 | 579 | namespaces = [('Locals', self.curframe.f_locals), |
|
580 | 580 | ('Globals', self.curframe.f_globals)] |
|
581 | 581 | self.shell.find_line_magic('pdef')(arg, namespaces=namespaces) |
|
582 | 582 | |
|
583 | 583 | def do_pdoc(self, arg): |
|
584 | 584 | """Print the docstring for an object. |
|
585 | 585 | |
|
586 | 586 | The debugger interface to %pdoc.""" |
|
587 | 587 | namespaces = [('Locals', self.curframe.f_locals), |
|
588 | 588 | ('Globals', self.curframe.f_globals)] |
|
589 | 589 | self.shell.find_line_magic('pdoc')(arg, namespaces=namespaces) |
|
590 | 590 | |
|
591 | 591 | def do_pfile(self, arg): |
|
592 | 592 | """Print (or run through pager) the file where an object is defined. |
|
593 | 593 | |
|
594 | 594 | The debugger interface to %pfile. |
|
595 | 595 | """ |
|
596 | 596 | namespaces = [('Locals', self.curframe.f_locals), |
|
597 | 597 | ('Globals', self.curframe.f_globals)] |
|
598 | 598 | self.shell.find_line_magic('pfile')(arg, namespaces=namespaces) |
|
599 | 599 | |
|
600 | 600 | def do_pinfo(self, arg): |
|
601 | 601 | """Provide detailed information about an object. |
|
602 | 602 | |
|
603 | 603 | The debugger interface to %pinfo, i.e., obj?.""" |
|
604 | 604 | namespaces = [('Locals', self.curframe.f_locals), |
|
605 | 605 | ('Globals', self.curframe.f_globals)] |
|
606 | 606 | self.shell.find_line_magic('pinfo')(arg, namespaces=namespaces) |
|
607 | 607 | |
|
608 | 608 | def do_pinfo2(self, arg): |
|
609 | 609 | """Provide extra detailed information about an object. |
|
610 | 610 | |
|
611 | 611 | The debugger interface to %pinfo2, i.e., obj??.""" |
|
612 | 612 | namespaces = [('Locals', self.curframe.f_locals), |
|
613 | 613 | ('Globals', self.curframe.f_globals)] |
|
614 | 614 | self.shell.find_line_magic('pinfo2')(arg, namespaces=namespaces) |
|
615 | 615 | |
|
616 | 616 | def do_psource(self, arg): |
|
617 | 617 | """Print (or run through pager) the source code for an object.""" |
|
618 | 618 | namespaces = [('Locals', self.curframe.f_locals), |
|
619 | 619 | ('Globals', self.curframe.f_globals)] |
|
620 | 620 | self.shell.find_line_magic('psource')(arg, namespaces=namespaces) |
|
621 | 621 | |
|
622 | 622 | def do_where(self, arg): |
|
623 | 623 | """w(here) |
|
624 | 624 | Print a stack trace, with the most recent frame at the bottom. |
|
625 | 625 | An arrow indicates the "current frame", which determines the |
|
626 | 626 | context of most commands. 'bt' is an alias for this command. |
|
627 | 627 | |
|
628 | 628 | Take a number as argument as an (optional) number of context line to |
|
629 | 629 | print""" |
|
630 | 630 | if arg: |
|
631 | 631 | context = int(arg) |
|
632 | 632 | self.print_stack_trace(context) |
|
633 | 633 | else: |
|
634 | 634 | self.print_stack_trace() |
|
635 | 635 | |
|
636 | 636 | do_w = do_where |
|
637 | 637 | |
|
638 | 638 | |
|
639 | 639 | def set_trace(frame=None): |
|
640 | 640 | """ |
|
641 | 641 | Start debugging from `frame`. |
|
642 | 642 | |
|
643 | 643 | If frame is not specified, debugging starts from caller's frame. |
|
644 | 644 | """ |
|
645 | 645 | Pdb().set_trace(frame or sys._getframe().f_back) |
@@ -1,766 +1,766 b'' | |||
|
1 | 1 | """Input handling and transformation machinery. |
|
2 | 2 | |
|
3 | 3 | The first class in this module, :class:`InputSplitter`, is designed to tell when |
|
4 | 4 | input from a line-oriented frontend is complete and should be executed, and when |
|
5 | 5 | the user should be prompted for another line of code instead. The name 'input |
|
6 | 6 | splitter' is largely for historical reasons. |
|
7 | 7 | |
|
8 | 8 | A companion, :class:`IPythonInputSplitter`, provides the same functionality but |
|
9 | 9 | with full support for the extended IPython syntax (magics, system calls, etc). |
|
10 | 10 | The code to actually do these transformations is in :mod:`IPython.core.inputtransformer`. |
|
11 | 11 | :class:`IPythonInputSplitter` feeds the raw code to the transformers in order |
|
12 | 12 | and stores the results. |
|
13 | 13 | |
|
14 | 14 | For more details, see the class docstrings below. |
|
15 | 15 | """ |
|
16 | 16 | |
|
17 | 17 | # Copyright (c) IPython Development Team. |
|
18 | 18 | # Distributed under the terms of the Modified BSD License. |
|
19 | 19 | import ast |
|
20 | 20 | import codeop |
|
21 | 21 | import io |
|
22 | 22 | import re |
|
23 | 23 | import sys |
|
24 | 24 | import tokenize |
|
25 | 25 | import warnings |
|
26 | 26 | |
|
27 | 27 | from IPython.utils.py3compat import cast_unicode |
|
28 | 28 | from IPython.core.inputtransformer import (leading_indent, |
|
29 | 29 | classic_prompt, |
|
30 | 30 | ipy_prompt, |
|
31 | 31 | cellmagic, |
|
32 | 32 | assemble_logical_lines, |
|
33 | 33 | help_end, |
|
34 | 34 | escaped_commands, |
|
35 | 35 | assign_from_magic, |
|
36 | 36 | assign_from_system, |
|
37 | 37 | assemble_python_lines, |
|
38 | 38 | ) |
|
39 | 39 | |
|
40 | 40 | # These are available in this module for backwards compatibility. |
|
41 | 41 | from IPython.core.inputtransformer import (ESC_SHELL, ESC_SH_CAP, ESC_HELP, |
|
42 | 42 | ESC_HELP2, ESC_MAGIC, ESC_MAGIC2, |
|
43 | 43 | ESC_QUOTE, ESC_QUOTE2, ESC_PAREN, ESC_SEQUENCES) |
|
44 | 44 | |
|
45 | 45 | #----------------------------------------------------------------------------- |
|
46 | 46 | # Utilities |
|
47 | 47 | #----------------------------------------------------------------------------- |
|
48 | 48 | |
|
49 | 49 | # FIXME: These are general-purpose utilities that later can be moved to the |
|
50 | 50 | # general ward. Kept here for now because we're being very strict about test |
|
51 | 51 | # coverage with this code, and this lets us ensure that we keep 100% coverage |
|
52 | 52 | # while developing. |
|
53 | 53 | |
|
54 | 54 | # compiled regexps for autoindent management |
|
55 | 55 | dedent_re = re.compile('|'.join([ |
|
56 | 56 | r'^\s+raise(\s.*)?$', # raise statement (+ space + other stuff, maybe) |
|
57 | 57 | r'^\s+raise\([^\)]*\).*$', # wacky raise with immediate open paren |
|
58 | 58 | r'^\s+return(\s.*)?$', # normal return (+ space + other stuff, maybe) |
|
59 | 59 | r'^\s+return\([^\)]*\).*$', # wacky return with immediate open paren |
|
60 | 60 | r'^\s+pass\s*$', # pass (optionally followed by trailing spaces) |
|
61 | 61 | r'^\s+break\s*$', # break (optionally followed by trailing spaces) |
|
62 | 62 | r'^\s+continue\s*$', # continue (optionally followed by trailing spaces) |
|
63 | 63 | ])) |
|
64 | 64 | ini_spaces_re = re.compile(r'^([ \t\r\f\v]+)') |
|
65 | 65 | |
|
66 | 66 | # regexp to match pure comment lines so we don't accidentally insert 'if 1:' |
|
67 | 67 | # before pure comments |
|
68 | comment_line_re = re.compile('^\s*\#') | |
|
68 | comment_line_re = re.compile(r'^\s*\#') | |
|
69 | 69 | |
|
70 | 70 | |
|
71 | 71 | def num_ini_spaces(s): |
|
72 | 72 | """Return the number of initial spaces in a string. |
|
73 | 73 | |
|
74 | 74 | Note that tabs are counted as a single space. For now, we do *not* support |
|
75 | 75 | mixing of tabs and spaces in the user's input. |
|
76 | 76 | |
|
77 | 77 | Parameters |
|
78 | 78 | ---------- |
|
79 | 79 | s : string |
|
80 | 80 | |
|
81 | 81 | Returns |
|
82 | 82 | ------- |
|
83 | 83 | n : int |
|
84 | 84 | """ |
|
85 | 85 | |
|
86 | 86 | ini_spaces = ini_spaces_re.match(s) |
|
87 | 87 | if ini_spaces: |
|
88 | 88 | return ini_spaces.end() |
|
89 | 89 | else: |
|
90 | 90 | return 0 |
|
91 | 91 | |
|
92 | 92 | # Fake token types for partial_tokenize: |
|
93 | 93 | INCOMPLETE_STRING = tokenize.N_TOKENS |
|
94 | 94 | IN_MULTILINE_STATEMENT = tokenize.N_TOKENS + 1 |
|
95 | 95 | |
|
96 | 96 | # The 2 classes below have the same API as TokenInfo, but don't try to look up |
|
97 | 97 | # a token type name that they won't find. |
|
98 | 98 | class IncompleteString: |
|
99 | 99 | type = exact_type = INCOMPLETE_STRING |
|
100 | 100 | def __init__(self, s, start, end, line): |
|
101 | 101 | self.s = s |
|
102 | 102 | self.start = start |
|
103 | 103 | self.end = end |
|
104 | 104 | self.line = line |
|
105 | 105 | |
|
106 | 106 | class InMultilineStatement: |
|
107 | 107 | type = exact_type = IN_MULTILINE_STATEMENT |
|
108 | 108 | def __init__(self, pos, line): |
|
109 | 109 | self.s = '' |
|
110 | 110 | self.start = self.end = pos |
|
111 | 111 | self.line = line |
|
112 | 112 | |
|
113 | 113 | def partial_tokens(s): |
|
114 | 114 | """Iterate over tokens from a possibly-incomplete string of code. |
|
115 | 115 | |
|
116 | 116 | This adds two special token types: INCOMPLETE_STRING and |
|
117 | 117 | IN_MULTILINE_STATEMENT. These can only occur as the last token yielded, and |
|
118 | 118 | represent the two main ways for code to be incomplete. |
|
119 | 119 | """ |
|
120 | 120 | readline = io.StringIO(s).readline |
|
121 | 121 | token = tokenize.TokenInfo(tokenize.NEWLINE, '', (1, 0), (1, 0), '') |
|
122 | 122 | try: |
|
123 | 123 | for token in tokenize.generate_tokens(readline): |
|
124 | 124 | yield token |
|
125 | 125 | except tokenize.TokenError as e: |
|
126 | 126 | # catch EOF error |
|
127 | 127 | lines = s.splitlines(keepends=True) |
|
128 | 128 | end = len(lines), len(lines[-1]) |
|
129 | 129 | if 'multi-line string' in e.args[0]: |
|
130 | 130 | l, c = start = token.end |
|
131 | 131 | s = lines[l-1][c:] + ''.join(lines[l:]) |
|
132 | 132 | yield IncompleteString(s, start, end, lines[-1]) |
|
133 | 133 | elif 'multi-line statement' in e.args[0]: |
|
134 | 134 | yield InMultilineStatement(end, lines[-1]) |
|
135 | 135 | else: |
|
136 | 136 | raise |
|
137 | 137 | |
|
138 | 138 | def find_next_indent(code): |
|
139 | 139 | """Find the number of spaces for the next line of indentation""" |
|
140 | 140 | tokens = list(partial_tokens(code)) |
|
141 | 141 | if tokens[-1].type == tokenize.ENDMARKER: |
|
142 | 142 | tokens.pop() |
|
143 | 143 | if not tokens: |
|
144 | 144 | return 0 |
|
145 | 145 | while (tokens[-1].type in {tokenize.DEDENT, tokenize.NEWLINE, tokenize.COMMENT}): |
|
146 | 146 | tokens.pop() |
|
147 | 147 | |
|
148 | 148 | if tokens[-1].type == INCOMPLETE_STRING: |
|
149 | 149 | # Inside a multiline string |
|
150 | 150 | return 0 |
|
151 | 151 | |
|
152 | 152 | # Find the indents used before |
|
153 | 153 | prev_indents = [0] |
|
154 | 154 | def _add_indent(n): |
|
155 | 155 | if n != prev_indents[-1]: |
|
156 | 156 | prev_indents.append(n) |
|
157 | 157 | |
|
158 | 158 | tokiter = iter(tokens) |
|
159 | 159 | for tok in tokiter: |
|
160 | 160 | if tok.type in {tokenize.INDENT, tokenize.DEDENT}: |
|
161 | 161 | _add_indent(tok.end[1]) |
|
162 | 162 | elif (tok.type == tokenize.NL): |
|
163 | 163 | try: |
|
164 | 164 | _add_indent(next(tokiter).start[1]) |
|
165 | 165 | except StopIteration: |
|
166 | 166 | break |
|
167 | 167 | |
|
168 | 168 | last_indent = prev_indents.pop() |
|
169 | 169 | |
|
170 | 170 | # If we've just opened a multiline statement (e.g. 'a = ['), indent more |
|
171 | 171 | if tokens[-1].type == IN_MULTILINE_STATEMENT: |
|
172 | 172 | if tokens[-2].exact_type in {tokenize.LPAR, tokenize.LSQB, tokenize.LBRACE}: |
|
173 | 173 | return last_indent + 4 |
|
174 | 174 | return last_indent |
|
175 | 175 | |
|
176 | 176 | if tokens[-1].exact_type == tokenize.COLON: |
|
177 | 177 | # Line ends with colon - indent |
|
178 | 178 | return last_indent + 4 |
|
179 | 179 | |
|
180 | 180 | if last_indent: |
|
181 | 181 | # Examine the last line for dedent cues - statements like return or |
|
182 | 182 | # raise which normally end a block of code. |
|
183 | 183 | last_line_starts = 0 |
|
184 | 184 | for i, tok in enumerate(tokens): |
|
185 | 185 | if tok.type == tokenize.NEWLINE: |
|
186 | 186 | last_line_starts = i + 1 |
|
187 | 187 | |
|
188 | 188 | last_line_tokens = tokens[last_line_starts:] |
|
189 | 189 | names = [t.string for t in last_line_tokens if t.type == tokenize.NAME] |
|
190 | 190 | if names and names[0] in {'raise', 'return', 'pass', 'break', 'continue'}: |
|
191 | 191 | # Find the most recent indentation less than the current level |
|
192 | 192 | for indent in reversed(prev_indents): |
|
193 | 193 | if indent < last_indent: |
|
194 | 194 | return indent |
|
195 | 195 | |
|
196 | 196 | return last_indent |
|
197 | 197 | |
|
198 | 198 | |
|
199 | 199 | def last_blank(src): |
|
200 | 200 | """Determine if the input source ends in a blank. |
|
201 | 201 | |
|
202 | 202 | A blank is either a newline or a line consisting of whitespace. |
|
203 | 203 | |
|
204 | 204 | Parameters |
|
205 | 205 | ---------- |
|
206 | 206 | src : string |
|
207 | 207 | A single or multiline string. |
|
208 | 208 | """ |
|
209 | 209 | if not src: return False |
|
210 | 210 | ll = src.splitlines()[-1] |
|
211 | 211 | return (ll == '') or ll.isspace() |
|
212 | 212 | |
|
213 | 213 | |
|
214 | 214 | last_two_blanks_re = re.compile(r'\n\s*\n\s*$', re.MULTILINE) |
|
215 | 215 | last_two_blanks_re2 = re.compile(r'.+\n\s*\n\s+$', re.MULTILINE) |
|
216 | 216 | |
|
217 | 217 | def last_two_blanks(src): |
|
218 | 218 | """Determine if the input source ends in two blanks. |
|
219 | 219 | |
|
220 | 220 | A blank is either a newline or a line consisting of whitespace. |
|
221 | 221 | |
|
222 | 222 | Parameters |
|
223 | 223 | ---------- |
|
224 | 224 | src : string |
|
225 | 225 | A single or multiline string. |
|
226 | 226 | """ |
|
227 | 227 | if not src: return False |
|
228 | 228 | # The logic here is tricky: I couldn't get a regexp to work and pass all |
|
229 | 229 | # the tests, so I took a different approach: split the source by lines, |
|
230 | 230 | # grab the last two and prepend '###\n' as a stand-in for whatever was in |
|
231 | 231 | # the body before the last two lines. Then, with that structure, it's |
|
232 | 232 | # possible to analyze with two regexps. Not the most elegant solution, but |
|
233 | 233 | # it works. If anyone tries to change this logic, make sure to validate |
|
234 | 234 | # the whole test suite first! |
|
235 | 235 | new_src = '\n'.join(['###\n'] + src.splitlines()[-2:]) |
|
236 | 236 | return (bool(last_two_blanks_re.match(new_src)) or |
|
237 | 237 | bool(last_two_blanks_re2.match(new_src)) ) |
|
238 | 238 | |
|
239 | 239 | |
|
240 | 240 | def remove_comments(src): |
|
241 | 241 | """Remove all comments from input source. |
|
242 | 242 | |
|
243 | 243 | Note: comments are NOT recognized inside of strings! |
|
244 | 244 | |
|
245 | 245 | Parameters |
|
246 | 246 | ---------- |
|
247 | 247 | src : string |
|
248 | 248 | A single or multiline input string. |
|
249 | 249 | |
|
250 | 250 | Returns |
|
251 | 251 | ------- |
|
252 | 252 | String with all Python comments removed. |
|
253 | 253 | """ |
|
254 | 254 | |
|
255 | 255 | return re.sub('#.*', '', src) |
|
256 | 256 | |
|
257 | 257 | |
|
258 | 258 | def get_input_encoding(): |
|
259 | 259 | """Return the default standard input encoding. |
|
260 | 260 | |
|
261 | 261 | If sys.stdin has no encoding, 'ascii' is returned.""" |
|
262 | 262 | # There are strange environments for which sys.stdin.encoding is None. We |
|
263 | 263 | # ensure that a valid encoding is returned. |
|
264 | 264 | encoding = getattr(sys.stdin, 'encoding', None) |
|
265 | 265 | if encoding is None: |
|
266 | 266 | encoding = 'ascii' |
|
267 | 267 | return encoding |
|
268 | 268 | |
|
269 | 269 | #----------------------------------------------------------------------------- |
|
270 | 270 | # Classes and functions for normal Python syntax handling |
|
271 | 271 | #----------------------------------------------------------------------------- |
|
272 | 272 | |
|
273 | 273 | class InputSplitter(object): |
|
274 | 274 | r"""An object that can accumulate lines of Python source before execution. |
|
275 | 275 | |
|
276 | 276 | This object is designed to be fed python source line-by-line, using |
|
277 | 277 | :meth:`push`. It will return on each push whether the currently pushed |
|
278 | 278 | code could be executed already. In addition, it provides a method called |
|
279 | 279 | :meth:`push_accepts_more` that can be used to query whether more input |
|
280 | 280 | can be pushed into a single interactive block. |
|
281 | 281 | |
|
282 | 282 | This is a simple example of how an interactive terminal-based client can use |
|
283 | 283 | this tool:: |
|
284 | 284 | |
|
285 | 285 | isp = InputSplitter() |
|
286 | 286 | while isp.push_accepts_more(): |
|
287 | 287 | indent = ' '*isp.indent_spaces |
|
288 | 288 | prompt = '>>> ' + indent |
|
289 | 289 | line = indent + raw_input(prompt) |
|
290 | 290 | isp.push(line) |
|
291 | 291 | print 'Input source was:\n', isp.source_reset(), |
|
292 | 292 | """ |
|
293 | 293 | # A cache for storing the current indentation |
|
294 | 294 | # The first value stores the most recently processed source input |
|
295 | 295 | # The second value is the number of spaces for the current indentation |
|
296 | 296 | # If self.source matches the first value, the second value is a valid |
|
297 | 297 | # current indentation. Otherwise, the cache is invalid and the indentation |
|
298 | 298 | # must be recalculated. |
|
299 | 299 | _indent_spaces_cache = None, None |
|
300 | 300 | # String, indicating the default input encoding. It is computed by default |
|
301 | 301 | # at initialization time via get_input_encoding(), but it can be reset by a |
|
302 | 302 | # client with specific knowledge of the encoding. |
|
303 | 303 | encoding = '' |
|
304 | 304 | # String where the current full source input is stored, properly encoded. |
|
305 | 305 | # Reading this attribute is the normal way of querying the currently pushed |
|
306 | 306 | # source code, that has been properly encoded. |
|
307 | 307 | source = '' |
|
308 | 308 | # Code object corresponding to the current source. It is automatically |
|
309 | 309 | # synced to the source, so it can be queried at any time to obtain the code |
|
310 | 310 | # object; it will be None if the source doesn't compile to valid Python. |
|
311 | 311 | code = None |
|
312 | 312 | |
|
313 | 313 | # Private attributes |
|
314 | 314 | |
|
315 | 315 | # List with lines of input accumulated so far |
|
316 | 316 | _buffer = None |
|
317 | 317 | # Command compiler |
|
318 | 318 | _compile = None |
|
319 | 319 | # Boolean indicating whether the current block is complete |
|
320 | 320 | _is_complete = None |
|
321 | 321 | # Boolean indicating whether the current block has an unrecoverable syntax error |
|
322 | 322 | _is_invalid = False |
|
323 | 323 | |
|
324 | 324 | def __init__(self): |
|
325 | 325 | """Create a new InputSplitter instance. |
|
326 | 326 | """ |
|
327 | 327 | self._buffer = [] |
|
328 | 328 | self._compile = codeop.CommandCompiler() |
|
329 | 329 | self.encoding = get_input_encoding() |
|
330 | 330 | |
|
331 | 331 | def reset(self): |
|
332 | 332 | """Reset the input buffer and associated state.""" |
|
333 | 333 | self._buffer[:] = [] |
|
334 | 334 | self.source = '' |
|
335 | 335 | self.code = None |
|
336 | 336 | self._is_complete = False |
|
337 | 337 | self._is_invalid = False |
|
338 | 338 | |
|
339 | 339 | def source_reset(self): |
|
340 | 340 | """Return the input source and perform a full reset. |
|
341 | 341 | """ |
|
342 | 342 | out = self.source |
|
343 | 343 | self.reset() |
|
344 | 344 | return out |
|
345 | 345 | |
|
346 | 346 | def check_complete(self, source): |
|
347 | 347 | """Return whether a block of code is ready to execute, or should be continued |
|
348 | 348 | |
|
349 | 349 | This is a non-stateful API, and will reset the state of this InputSplitter. |
|
350 | 350 | |
|
351 | 351 | Parameters |
|
352 | 352 | ---------- |
|
353 | 353 | source : string |
|
354 | 354 | Python input code, which can be multiline. |
|
355 | 355 | |
|
356 | 356 | Returns |
|
357 | 357 | ------- |
|
358 | 358 | status : str |
|
359 | 359 | One of 'complete', 'incomplete', or 'invalid' if source is not a |
|
360 | 360 | prefix of valid code. |
|
361 | 361 | indent_spaces : int or None |
|
362 | 362 | The number of spaces by which to indent the next line of code. If |
|
363 | 363 | status is not 'incomplete', this is None. |
|
364 | 364 | """ |
|
365 | 365 | self.reset() |
|
366 | 366 | try: |
|
367 | 367 | self.push(source) |
|
368 | 368 | except SyntaxError: |
|
369 | 369 | # Transformers in IPythonInputSplitter can raise SyntaxError, |
|
370 | 370 | # which push() will not catch. |
|
371 | 371 | return 'invalid', None |
|
372 | 372 | else: |
|
373 | 373 | if self._is_invalid: |
|
374 | 374 | return 'invalid', None |
|
375 | 375 | elif self.push_accepts_more(): |
|
376 | 376 | return 'incomplete', self.get_indent_spaces() |
|
377 | 377 | else: |
|
378 | 378 | return 'complete', None |
|
379 | 379 | finally: |
|
380 | 380 | self.reset() |
|
381 | 381 | |
|
382 | 382 | def push(self, lines): |
|
383 | 383 | """Push one or more lines of input. |
|
384 | 384 | |
|
385 | 385 | This stores the given lines and returns a status code indicating |
|
386 | 386 | whether the code forms a complete Python block or not. |
|
387 | 387 | |
|
388 | 388 | Any exceptions generated in compilation are swallowed, but if an |
|
389 | 389 | exception was produced, the method returns True. |
|
390 | 390 | |
|
391 | 391 | Parameters |
|
392 | 392 | ---------- |
|
393 | 393 | lines : string |
|
394 | 394 | One or more lines of Python input. |
|
395 | 395 | |
|
396 | 396 | Returns |
|
397 | 397 | ------- |
|
398 | 398 | is_complete : boolean |
|
399 | 399 | True if the current input source (the result of the current input |
|
400 | 400 | plus prior inputs) forms a complete Python execution block. Note that |
|
401 | 401 | this value is also stored as a private attribute (``_is_complete``), so it |
|
402 | 402 | can be queried at any time. |
|
403 | 403 | """ |
|
404 | 404 | self._store(lines) |
|
405 | 405 | source = self.source |
|
406 | 406 | |
|
407 | 407 | # Before calling _compile(), reset the code object to None so that if an |
|
408 | 408 | # exception is raised in compilation, we don't mislead by having |
|
409 | 409 | # inconsistent code/source attributes. |
|
410 | 410 | self.code, self._is_complete = None, None |
|
411 | 411 | self._is_invalid = False |
|
412 | 412 | |
|
413 | 413 | # Honor termination lines properly |
|
414 | 414 | if source.endswith('\\\n'): |
|
415 | 415 | return False |
|
416 | 416 | |
|
417 | 417 | try: |
|
418 | 418 | with warnings.catch_warnings(): |
|
419 | 419 | warnings.simplefilter('error', SyntaxWarning) |
|
420 | 420 | self.code = self._compile(source, symbol="exec") |
|
421 | 421 | # Invalid syntax can produce any of a number of different errors from |
|
422 | 422 | # inside the compiler, so we have to catch them all. Syntax errors |
|
423 | 423 | # immediately produce a 'ready' block, so the invalid Python can be |
|
424 | 424 | # sent to the kernel for evaluation with possible ipython |
|
425 | 425 | # special-syntax conversion. |
|
426 | 426 | except (SyntaxError, OverflowError, ValueError, TypeError, |
|
427 | 427 | MemoryError, SyntaxWarning): |
|
428 | 428 | self._is_complete = True |
|
429 | 429 | self._is_invalid = True |
|
430 | 430 | else: |
|
431 | 431 | # Compilation didn't produce any exceptions (though it may not have |
|
432 | 432 | # given a complete code object) |
|
433 | 433 | self._is_complete = self.code is not None |
|
434 | 434 | |
|
435 | 435 | return self._is_complete |
|
436 | 436 | |
|
437 | 437 | def push_accepts_more(self): |
|
438 | 438 | """Return whether a block of interactive input can accept more input. |
|
439 | 439 | |
|
440 | 440 | This method is meant to be used by line-oriented frontends, who need to |
|
441 | 441 | guess whether a block is complete or not based solely on prior and |
|
442 | 442 | current input lines. The InputSplitter considers it has a complete |
|
443 | 443 | interactive block and will not accept more input when either: |
|
444 | 444 | |
|
445 | 445 | * A SyntaxError is raised |
|
446 | 446 | |
|
447 | 447 | * The code is complete and consists of a single line or a single |
|
448 | 448 | non-compound statement |
|
449 | 449 | |
|
450 | 450 | * The code is complete and has a blank line at the end |
|
451 | 451 | |
|
452 | 452 | If the current input produces a syntax error, this method immediately |
|
453 | 453 | returns False but does *not* raise the syntax error exception, as |
|
454 | 454 | typically clients will want to send invalid syntax to an execution |
|
455 | 455 | backend which might convert the invalid syntax into valid Python via |
|
456 | 456 | one of the dynamic IPython mechanisms. |
|
457 | 457 | """ |
|
458 | 458 | |
|
459 | 459 | # With incomplete input, unconditionally accept more |
|
460 | 460 | # A syntax error also sets _is_complete to True - see push() |
|
461 | 461 | if not self._is_complete: |
|
462 | 462 | #print("Not complete") # debug |
|
463 | 463 | return True |
|
464 | 464 | |
|
465 | 465 | # The user can make any (complete) input execute by leaving a blank line |
|
466 | 466 | last_line = self.source.splitlines()[-1] |
|
467 | 467 | if (not last_line) or last_line.isspace(): |
|
468 | 468 | #print("Blank line") # debug |
|
469 | 469 | return False |
|
470 | 470 | |
|
471 | 471 | # If there's just a single line or AST node, and we're flush left, as is |
|
472 | 472 | # the case after a simple statement such as 'a=1', we want to execute it |
|
473 | 473 | # straight away. |
|
474 | 474 | if self.get_indent_spaces() == 0: |
|
475 | 475 | if len(self.source.splitlines()) <= 1: |
|
476 | 476 | return False |
|
477 | 477 | |
|
478 | 478 | try: |
|
479 | 479 | code_ast = ast.parse(u''.join(self._buffer)) |
|
480 | 480 | except Exception: |
|
481 | 481 | #print("Can't parse AST") # debug |
|
482 | 482 | return False |
|
483 | 483 | else: |
|
484 | 484 | if len(code_ast.body) == 1 and \ |
|
485 | 485 | not hasattr(code_ast.body[0], 'body'): |
|
486 | 486 | #print("Simple statement") # debug |
|
487 | 487 | return False |
|
488 | 488 | |
|
489 | 489 | # General fallback - accept more code |
|
490 | 490 | return True |
|
491 | 491 | |
|
492 | 492 | def get_indent_spaces(self): |
|
493 | 493 | sourcefor, n = self._indent_spaces_cache |
|
494 | 494 | if sourcefor == self.source: |
|
495 | 495 | return n |
|
496 | 496 | |
|
497 | 497 | # self.source always has a trailing newline |
|
498 | 498 | n = find_next_indent(self.source[:-1]) |
|
499 | 499 | self._indent_spaces_cache = (self.source, n) |
|
500 | 500 | return n |
|
501 | 501 | |
|
502 | 502 | # Backwards compatibility. I think all code that used .indent_spaces was |
|
503 | 503 | # inside IPython, but we can leave this here until IPython 7 in case any |
|
504 | 504 | # other modules are using it. -TK, November 2017 |
|
505 | 505 | indent_spaces = property(get_indent_spaces) |
|
506 | 506 | |
|
507 | 507 | def _store(self, lines, buffer=None, store='source'): |
|
508 | 508 | """Store one or more lines of input. |
|
509 | 509 | |
|
510 | 510 | If input lines are not newline-terminated, a newline is automatically |
|
511 | 511 | appended.""" |
|
512 | 512 | |
|
513 | 513 | if buffer is None: |
|
514 | 514 | buffer = self._buffer |
|
515 | 515 | |
|
516 | 516 | if lines.endswith('\n'): |
|
517 | 517 | buffer.append(lines) |
|
518 | 518 | else: |
|
519 | 519 | buffer.append(lines+'\n') |
|
520 | 520 | setattr(self, store, self._set_source(buffer)) |
|
521 | 521 | |
|
522 | 522 | def _set_source(self, buffer): |
|
523 | 523 | return u''.join(buffer) |
|
524 | 524 | |
|
525 | 525 | |
|
526 | 526 | class IPythonInputSplitter(InputSplitter): |
|
527 | 527 | """An input splitter that recognizes all of IPython's special syntax.""" |
|
528 | 528 | |
|
529 | 529 | # String with raw, untransformed input. |
|
530 | 530 | source_raw = '' |
|
531 | 531 | |
|
532 | 532 | # Flag to track when a transformer has stored input that it hasn't given |
|
533 | 533 | # back yet. |
|
534 | 534 | transformer_accumulating = False |
|
535 | 535 | |
|
536 | 536 | # Flag to track when assemble_python_lines has stored input that it hasn't |
|
537 | 537 | # given back yet. |
|
538 | 538 | within_python_line = False |
|
539 | 539 | |
|
540 | 540 | # Private attributes |
|
541 | 541 | |
|
542 | 542 | # List with lines of raw input accumulated so far. |
|
543 | 543 | _buffer_raw = None |
|
544 | 544 | |
|
545 | 545 | def __init__(self, line_input_checker=True, physical_line_transforms=None, |
|
546 | 546 | logical_line_transforms=None, python_line_transforms=None): |
|
547 | 547 | super(IPythonInputSplitter, self).__init__() |
|
548 | 548 | self._buffer_raw = [] |
|
549 | 549 | self._validate = True |
|
550 | 550 | |
|
551 | 551 | if physical_line_transforms is not None: |
|
552 | 552 | self.physical_line_transforms = physical_line_transforms |
|
553 | 553 | else: |
|
554 | 554 | self.physical_line_transforms = [ |
|
555 | 555 | leading_indent(), |
|
556 | 556 | classic_prompt(), |
|
557 | 557 | ipy_prompt(), |
|
558 | 558 | cellmagic(end_on_blank_line=line_input_checker), |
|
559 | 559 | ] |
|
560 | 560 | |
|
561 | 561 | self.assemble_logical_lines = assemble_logical_lines() |
|
562 | 562 | if logical_line_transforms is not None: |
|
563 | 563 | self.logical_line_transforms = logical_line_transforms |
|
564 | 564 | else: |
|
565 | 565 | self.logical_line_transforms = [ |
|
566 | 566 | help_end(), |
|
567 | 567 | escaped_commands(), |
|
568 | 568 | assign_from_magic(), |
|
569 | 569 | assign_from_system(), |
|
570 | 570 | ] |
|
571 | 571 | |
|
572 | 572 | self.assemble_python_lines = assemble_python_lines() |
|
573 | 573 | if python_line_transforms is not None: |
|
574 | 574 | self.python_line_transforms = python_line_transforms |
|
575 | 575 | else: |
|
576 | 576 | # We don't use any of these at present |
|
577 | 577 | self.python_line_transforms = [] |
|
578 | 578 | |
|
579 | 579 | @property |
|
580 | 580 | def transforms(self): |
|
581 | 581 | "Quick access to all transformers." |
|
582 | 582 | return self.physical_line_transforms + \ |
|
583 | 583 | [self.assemble_logical_lines] + self.logical_line_transforms + \ |
|
584 | 584 | [self.assemble_python_lines] + self.python_line_transforms |
|
585 | 585 | |
|
586 | 586 | @property |
|
587 | 587 | def transforms_in_use(self): |
|
588 | 588 | """Transformers, excluding logical line transformers if we're in a |
|
589 | 589 | Python line.""" |
|
590 | 590 | t = self.physical_line_transforms[:] |
|
591 | 591 | if not self.within_python_line: |
|
592 | 592 | t += [self.assemble_logical_lines] + self.logical_line_transforms |
|
593 | 593 | return t + [self.assemble_python_lines] + self.python_line_transforms |
|
594 | 594 | |
|
595 | 595 | def reset(self): |
|
596 | 596 | """Reset the input buffer and associated state.""" |
|
597 | 597 | super(IPythonInputSplitter, self).reset() |
|
598 | 598 | self._buffer_raw[:] = [] |
|
599 | 599 | self.source_raw = '' |
|
600 | 600 | self.transformer_accumulating = False |
|
601 | 601 | self.within_python_line = False |
|
602 | 602 | |
|
603 | 603 | for t in self.transforms: |
|
604 | 604 | try: |
|
605 | 605 | t.reset() |
|
606 | 606 | except SyntaxError: |
|
607 | 607 | # Nothing that calls reset() expects to handle transformer |
|
608 | 608 | # errors |
|
609 | 609 | pass |
|
610 | 610 | |
|
611 | 611 | def flush_transformers(self): |
|
612 | 612 | def _flush(transform, outs): |
|
613 | 613 | """yield transformed lines |
|
614 | 614 | |
|
615 | 615 | always strings, never None |
|
616 | 616 | |
|
617 | 617 | transform: the current transform |
|
618 | 618 | outs: an iterable of previously transformed inputs. |
|
619 | 619 | Each may be multiline, which will be passed |
|
620 | 620 | one line at a time to transform. |
|
621 | 621 | """ |
|
622 | 622 | for out in outs: |
|
623 | 623 | for line in out.splitlines(): |
|
624 | 624 | # push one line at a time |
|
625 | 625 | tmp = transform.push(line) |
|
626 | 626 | if tmp is not None: |
|
627 | 627 | yield tmp |
|
628 | 628 | |
|
629 | 629 | # reset the transform |
|
630 | 630 | tmp = transform.reset() |
|
631 | 631 | if tmp is not None: |
|
632 | 632 | yield tmp |
|
633 | 633 | |
|
634 | 634 | out = [] |
|
635 | 635 | for t in self.transforms_in_use: |
|
636 | 636 | out = _flush(t, out) |
|
637 | 637 | |
|
638 | 638 | out = list(out) |
|
639 | 639 | if out: |
|
640 | 640 | self._store('\n'.join(out)) |
|
641 | 641 | |
|
642 | 642 | def raw_reset(self): |
|
643 | 643 | """Return raw input only and perform a full reset. |
|
644 | 644 | """ |
|
645 | 645 | out = self.source_raw |
|
646 | 646 | self.reset() |
|
647 | 647 | return out |
|
648 | 648 | |
|
649 | 649 | def source_reset(self): |
|
650 | 650 | try: |
|
651 | 651 | self.flush_transformers() |
|
652 | 652 | return self.source |
|
653 | 653 | finally: |
|
654 | 654 | self.reset() |
|
655 | 655 | |
|
656 | 656 | def push_accepts_more(self): |
|
657 | 657 | if self.transformer_accumulating: |
|
658 | 658 | return True |
|
659 | 659 | else: |
|
660 | 660 | return super(IPythonInputSplitter, self).push_accepts_more() |
|
661 | 661 | |
|
662 | 662 | def transform_cell(self, cell): |
|
663 | 663 | """Process and translate a cell of input. |
|
664 | 664 | """ |
|
665 | 665 | self.reset() |
|
666 | 666 | try: |
|
667 | 667 | self.push(cell) |
|
668 | 668 | self.flush_transformers() |
|
669 | 669 | return self.source |
|
670 | 670 | finally: |
|
671 | 671 | self.reset() |
|
672 | 672 | |
|
673 | 673 | def push(self, lines): |
|
674 | 674 | """Push one or more lines of IPython input. |
|
675 | 675 | |
|
676 | 676 | This stores the given lines and returns a status code indicating |
|
677 | 677 | whether the code forms a complete Python block or not, after processing |
|
678 | 678 | all input lines for special IPython syntax. |
|
679 | 679 | |
|
680 | 680 | Any exceptions generated in compilation are swallowed, but if an |
|
681 | 681 | exception was produced, the method returns True. |
|
682 | 682 | |
|
683 | 683 | Parameters |
|
684 | 684 | ---------- |
|
685 | 685 | lines : string |
|
686 | 686 | One or more lines of Python input. |
|
687 | 687 | |
|
688 | 688 | Returns |
|
689 | 689 | ------- |
|
690 | 690 | is_complete : boolean |
|
691 | 691 | True if the current input source (the result of the current input |
|
692 | 692 | plus prior inputs) forms a complete Python execution block. Note that |
|
693 | 693 | this value is also stored as a private attribute (_is_complete), so it |
|
694 | 694 | can be queried at any time. |
|
695 | 695 | """ |
|
696 | 696 | |
|
697 | 697 | # We must ensure all input is pure unicode |
|
698 | 698 | lines = cast_unicode(lines, self.encoding) |
|
699 | 699 | # ''.splitlines() --> [], but we need to push the empty line to transformers |
|
700 | 700 | lines_list = lines.splitlines() |
|
701 | 701 | if not lines_list: |
|
702 | 702 | lines_list = [''] |
|
703 | 703 | |
|
704 | 704 | # Store raw source before applying any transformations to it. Note |
|
705 | 705 | # that this must be done *after* the reset() call that would otherwise |
|
706 | 706 | # flush the buffer. |
|
707 | 707 | self._store(lines, self._buffer_raw, 'source_raw') |
|
708 | 708 | |
|
709 | 709 | transformed_lines_list = [] |
|
710 | 710 | for line in lines_list: |
|
711 | 711 | transformed = self._transform_line(line) |
|
712 | 712 | if transformed is not None: |
|
713 | 713 | transformed_lines_list.append(transformed) |
|
714 | 714 | |
|
715 | 715 | if transformed_lines_list: |
|
716 | 716 | transformed_lines = '\n'.join(transformed_lines_list) |
|
717 | 717 | return super(IPythonInputSplitter, self).push(transformed_lines) |
|
718 | 718 | else: |
|
719 | 719 | # Got nothing back from transformers - they must be waiting for |
|
720 | 720 | # more input. |
|
721 | 721 | return False |
|
722 | 722 | |
|
723 | 723 | def _transform_line(self, line): |
|
724 | 724 | """Push a line of input code through the various transformers. |
|
725 | 725 | |
|
726 | 726 | Returns any output from the transformers, or None if a transformer |
|
727 | 727 | is accumulating lines. |
|
728 | 728 | |
|
729 | 729 | Sets self.transformer_accumulating as a side effect. |
|
730 | 730 | """ |
|
731 | 731 | def _accumulating(dbg): |
|
732 | 732 | #print(dbg) |
|
733 | 733 | self.transformer_accumulating = True |
|
734 | 734 | return None |
|
735 | 735 | |
|
736 | 736 | for transformer in self.physical_line_transforms: |
|
737 | 737 | line = transformer.push(line) |
|
738 | 738 | if line is None: |
|
739 | 739 | return _accumulating(transformer) |
|
740 | 740 | |
|
741 | 741 | if not self.within_python_line: |
|
742 | 742 | line = self.assemble_logical_lines.push(line) |
|
743 | 743 | if line is None: |
|
744 | 744 | return _accumulating('acc logical line') |
|
745 | 745 | |
|
746 | 746 | for transformer in self.logical_line_transforms: |
|
747 | 747 | line = transformer.push(line) |
|
748 | 748 | if line is None: |
|
749 | 749 | return _accumulating(transformer) |
|
750 | 750 | |
|
751 | 751 | line = self.assemble_python_lines.push(line) |
|
752 | 752 | if line is None: |
|
753 | 753 | self.within_python_line = True |
|
754 | 754 | return _accumulating('acc python line') |
|
755 | 755 | else: |
|
756 | 756 | self.within_python_line = False |
|
757 | 757 | |
|
758 | 758 | for transformer in self.python_line_transforms: |
|
759 | 759 | line = transformer.push(line) |
|
760 | 760 | if line is None: |
|
761 | 761 | return _accumulating(transformer) |
|
762 | 762 | |
|
763 | 763 | #print("transformers clear") #debug |
|
764 | 764 | self.transformer_accumulating = False |
|
765 | 765 | return line |
|
766 | 766 |
@@ -1,534 +1,534 b'' | |||
|
1 | 1 | """Input transformer classes to support IPython special syntax. |
|
2 | 2 | |
|
3 | 3 | This includes the machinery to recognise and transform ``%magic`` commands, |
|
4 | 4 | ``!system`` commands, ``help?`` querying, prompt stripping, and so forth. |
|
5 | 5 | """ |
|
6 | 6 | import abc |
|
7 | 7 | import functools |
|
8 | 8 | import re |
|
9 | 9 | from io import StringIO |
|
10 | 10 | |
|
11 | 11 | from IPython.core.splitinput import LineInfo |
|
12 | 12 | from IPython.utils import tokenize2 |
|
13 | 13 | from IPython.utils.tokenize2 import generate_tokens, untokenize, TokenError |
|
14 | 14 | |
|
15 | 15 | #----------------------------------------------------------------------------- |
|
16 | 16 | # Globals |
|
17 | 17 | #----------------------------------------------------------------------------- |
|
18 | 18 | |
|
19 | 19 | # The escape sequences that define the syntax transformations IPython will |
|
20 | 20 | # apply to user input. These can NOT be just changed here: many regular |
|
21 | 21 | # expressions and other parts of the code may use their hardcoded values, and |
|
22 | 22 | # for all intents and purposes they constitute the 'IPython syntax', so they |
|
23 | 23 | # should be considered fixed. |
|
24 | 24 | |
|
25 | 25 | ESC_SHELL = '!' # Send line to underlying system shell |
|
26 | 26 | ESC_SH_CAP = '!!' # Send line to system shell and capture output |
|
27 | 27 | ESC_HELP = '?' # Find information about object |
|
28 | 28 | ESC_HELP2 = '??' # Find extra-detailed information about object |
|
29 | 29 | ESC_MAGIC = '%' # Call magic function |
|
30 | 30 | ESC_MAGIC2 = '%%' # Call cell-magic function |
|
31 | 31 | ESC_QUOTE = ',' # Split args on whitespace, quote each as string and call |
|
32 | 32 | ESC_QUOTE2 = ';' # Quote all args as a single string, call |
|
33 | 33 | ESC_PAREN = '/' # Call first argument with rest of line as arguments |
|
34 | 34 | |
|
35 | 35 | ESC_SEQUENCES = [ESC_SHELL, ESC_SH_CAP, ESC_HELP ,\ |
|
36 | 36 | ESC_HELP2, ESC_MAGIC, ESC_MAGIC2,\ |
|
37 | 37 | ESC_QUOTE, ESC_QUOTE2, ESC_PAREN ] |
|
38 | 38 | |
|
39 | 39 | |
|
40 | 40 | class InputTransformer(metaclass=abc.ABCMeta): |
|
41 | 41 | """Abstract base class for line-based input transformers.""" |
|
42 | 42 | |
|
43 | 43 | @abc.abstractmethod |
|
44 | 44 | def push(self, line): |
|
45 | 45 | """Send a line of input to the transformer, returning the transformed |
|
46 | 46 | input or None if the transformer is waiting for more input. |
|
47 | 47 | |
|
48 | 48 | Must be overridden by subclasses. |
|
49 | 49 | |
|
50 | 50 | Implementations may raise ``SyntaxError`` if the input is invalid. No |
|
51 | 51 | other exceptions may be raised. |
|
52 | 52 | """ |
|
53 | 53 | pass |
|
54 | 54 | |
|
55 | 55 | @abc.abstractmethod |
|
56 | 56 | def reset(self): |
|
57 | 57 | """Return, transformed any lines that the transformer has accumulated, |
|
58 | 58 | and reset its internal state. |
|
59 | 59 | |
|
60 | 60 | Must be overridden by subclasses. |
|
61 | 61 | """ |
|
62 | 62 | pass |
|
63 | 63 | |
|
64 | 64 | @classmethod |
|
65 | 65 | def wrap(cls, func): |
|
66 | 66 | """Can be used by subclasses as a decorator, to return a factory that |
|
67 | 67 | will allow instantiation with the decorated object. |
|
68 | 68 | """ |
|
69 | 69 | @functools.wraps(func) |
|
70 | 70 | def transformer_factory(**kwargs): |
|
71 | 71 | return cls(func, **kwargs) |
|
72 | 72 | |
|
73 | 73 | return transformer_factory |
|
74 | 74 | |
|
75 | 75 | class StatelessInputTransformer(InputTransformer): |
|
76 | 76 | """Wrapper for a stateless input transformer implemented as a function.""" |
|
77 | 77 | def __init__(self, func): |
|
78 | 78 | self.func = func |
|
79 | 79 | |
|
80 | 80 | def __repr__(self): |
|
81 | 81 | return "StatelessInputTransformer(func={0!r})".format(self.func) |
|
82 | 82 | |
|
83 | 83 | def push(self, line): |
|
84 | 84 | """Send a line of input to the transformer, returning the |
|
85 | 85 | transformed input.""" |
|
86 | 86 | return self.func(line) |
|
87 | 87 | |
|
88 | 88 | def reset(self): |
|
89 | 89 | """No-op - exists for compatibility.""" |
|
90 | 90 | pass |
|
91 | 91 | |
|
92 | 92 | class CoroutineInputTransformer(InputTransformer): |
|
93 | 93 | """Wrapper for an input transformer implemented as a coroutine.""" |
|
94 | 94 | def __init__(self, coro, **kwargs): |
|
95 | 95 | # Prime it |
|
96 | 96 | self.coro = coro(**kwargs) |
|
97 | 97 | next(self.coro) |
|
98 | 98 | |
|
99 | 99 | def __repr__(self): |
|
100 | 100 | return "CoroutineInputTransformer(coro={0!r})".format(self.coro) |
|
101 | 101 | |
|
102 | 102 | def push(self, line): |
|
103 | 103 | """Send a line of input to the transformer, returning the |
|
104 | 104 | transformed input or None if the transformer is waiting for more |
|
105 | 105 | input. |
|
106 | 106 | """ |
|
107 | 107 | return self.coro.send(line) |
|
108 | 108 | |
|
109 | 109 | def reset(self): |
|
110 | 110 | """Return, transformed any lines that the transformer has |
|
111 | 111 | accumulated, and reset its internal state. |
|
112 | 112 | """ |
|
113 | 113 | return self.coro.send(None) |
|
114 | 114 | |
|
115 | 115 | class TokenInputTransformer(InputTransformer): |
|
116 | 116 | """Wrapper for a token-based input transformer. |
|
117 | 117 | |
|
118 | 118 | func should accept a list of tokens (5-tuples, see tokenize docs), and |
|
119 | 119 | return an iterable which can be passed to tokenize.untokenize(). |
|
120 | 120 | """ |
|
121 | 121 | def __init__(self, func): |
|
122 | 122 | self.func = func |
|
123 | 123 | self.buf = [] |
|
124 | 124 | self.reset_tokenizer() |
|
125 | 125 | |
|
126 | 126 | def reset_tokenizer(self): |
|
127 | 127 | it = iter(self.buf) |
|
128 | 128 | self.tokenizer = generate_tokens(it.__next__) |
|
129 | 129 | |
|
130 | 130 | def push(self, line): |
|
131 | 131 | self.buf.append(line + '\n') |
|
132 | 132 | if all(l.isspace() for l in self.buf): |
|
133 | 133 | return self.reset() |
|
134 | 134 | |
|
135 | 135 | tokens = [] |
|
136 | 136 | stop_at_NL = False |
|
137 | 137 | try: |
|
138 | 138 | for intok in self.tokenizer: |
|
139 | 139 | tokens.append(intok) |
|
140 | 140 | t = intok[0] |
|
141 | 141 | if t == tokenize2.NEWLINE or (stop_at_NL and t == tokenize2.NL): |
|
142 | 142 | # Stop before we try to pull a line we don't have yet |
|
143 | 143 | break |
|
144 | 144 | elif t == tokenize2.ERRORTOKEN: |
|
145 | 145 | stop_at_NL = True |
|
146 | 146 | except TokenError: |
|
147 | 147 | # Multi-line statement - stop and try again with the next line |
|
148 | 148 | self.reset_tokenizer() |
|
149 | 149 | return None |
|
150 | 150 | |
|
151 | 151 | return self.output(tokens) |
|
152 | 152 | |
|
153 | 153 | def output(self, tokens): |
|
154 | 154 | self.buf.clear() |
|
155 | 155 | self.reset_tokenizer() |
|
156 | 156 | return untokenize(self.func(tokens)).rstrip('\n') |
|
157 | 157 | |
|
158 | 158 | def reset(self): |
|
159 | 159 | l = ''.join(self.buf) |
|
160 | 160 | self.buf.clear() |
|
161 | 161 | self.reset_tokenizer() |
|
162 | 162 | if l: |
|
163 | 163 | return l.rstrip('\n') |
|
164 | 164 | |
|
165 | 165 | class assemble_python_lines(TokenInputTransformer): |
|
166 | 166 | def __init__(self): |
|
167 | 167 | super(assemble_python_lines, self).__init__(None) |
|
168 | 168 | |
|
169 | 169 | def output(self, tokens): |
|
170 | 170 | return self.reset() |
|
171 | 171 | |
|
172 | 172 | @CoroutineInputTransformer.wrap |
|
173 | 173 | def assemble_logical_lines(): |
|
174 | """Join lines following explicit line continuations (\)""" | |
|
174 | r"""Join lines following explicit line continuations (\)""" | |
|
175 | 175 | line = '' |
|
176 | 176 | while True: |
|
177 | 177 | line = (yield line) |
|
178 | 178 | if not line or line.isspace(): |
|
179 | 179 | continue |
|
180 | 180 | |
|
181 | 181 | parts = [] |
|
182 | 182 | while line is not None: |
|
183 | 183 | if line.endswith('\\') and (not has_comment(line)): |
|
184 | 184 | parts.append(line[:-1]) |
|
185 | 185 | line = (yield None) # Get another line |
|
186 | 186 | else: |
|
187 | 187 | parts.append(line) |
|
188 | 188 | break |
|
189 | 189 | |
|
190 | 190 | # Output |
|
191 | 191 | line = ''.join(parts) |
|
192 | 192 | |
|
193 | 193 | # Utilities |
|
194 | 194 | def _make_help_call(target, esc, lspace, next_input=None): |
|
195 | 195 | """Prepares a pinfo(2)/psearch call from a target name and the escape |
|
196 | 196 | (i.e. ? or ??)""" |
|
197 | 197 | method = 'pinfo2' if esc == '??' \ |
|
198 | 198 | else 'psearch' if '*' in target \ |
|
199 | 199 | else 'pinfo' |
|
200 | 200 | arg = " ".join([method, target]) |
|
201 | 201 | #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args) |
|
202 | 202 | t_magic_name, _, t_magic_arg_s = arg.partition(' ') |
|
203 | 203 | t_magic_name = t_magic_name.lstrip(ESC_MAGIC) |
|
204 | 204 | if next_input is None: |
|
205 | 205 | return '%sget_ipython().run_line_magic(%r, %r)' % (lspace, t_magic_name, t_magic_arg_s) |
|
206 | 206 | else: |
|
207 | 207 | return '%sget_ipython().set_next_input(%r);get_ipython().run_line_magic(%r, %r)' % \ |
|
208 | 208 | (lspace, next_input, t_magic_name, t_magic_arg_s) |
|
209 | 209 | |
|
210 | 210 | # These define the transformations for the different escape characters. |
|
211 | 211 | def _tr_system(line_info): |
|
212 | 212 | "Translate lines escaped with: !" |
|
213 | 213 | cmd = line_info.line.lstrip().lstrip(ESC_SHELL) |
|
214 | 214 | return '%sget_ipython().system(%r)' % (line_info.pre, cmd) |
|
215 | 215 | |
|
216 | 216 | def _tr_system2(line_info): |
|
217 | 217 | "Translate lines escaped with: !!" |
|
218 | 218 | cmd = line_info.line.lstrip()[2:] |
|
219 | 219 | return '%sget_ipython().getoutput(%r)' % (line_info.pre, cmd) |
|
220 | 220 | |
|
221 | 221 | def _tr_help(line_info): |
|
222 | 222 | "Translate lines escaped with: ?/??" |
|
223 | 223 | # A naked help line should just fire the intro help screen |
|
224 | 224 | if not line_info.line[1:]: |
|
225 | 225 | return 'get_ipython().show_usage()' |
|
226 | 226 | |
|
227 | 227 | return _make_help_call(line_info.ifun, line_info.esc, line_info.pre) |
|
228 | 228 | |
|
229 | 229 | def _tr_magic(line_info): |
|
230 | 230 | "Translate lines escaped with: %" |
|
231 | 231 | tpl = '%sget_ipython().run_line_magic(%r, %r)' |
|
232 | 232 | if line_info.line.startswith(ESC_MAGIC2): |
|
233 | 233 | return line_info.line |
|
234 | 234 | cmd = ' '.join([line_info.ifun, line_info.the_rest]).strip() |
|
235 | 235 | #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args) |
|
236 | 236 | t_magic_name, _, t_magic_arg_s = cmd.partition(' ') |
|
237 | 237 | t_magic_name = t_magic_name.lstrip(ESC_MAGIC) |
|
238 | 238 | return tpl % (line_info.pre, t_magic_name, t_magic_arg_s) |
|
239 | 239 | |
|
240 | 240 | def _tr_quote(line_info): |
|
241 | 241 | "Translate lines escaped with: ," |
|
242 | 242 | return '%s%s("%s")' % (line_info.pre, line_info.ifun, |
|
243 | 243 | '", "'.join(line_info.the_rest.split()) ) |
|
244 | 244 | |
|
245 | 245 | def _tr_quote2(line_info): |
|
246 | 246 | "Translate lines escaped with: ;" |
|
247 | 247 | return '%s%s("%s")' % (line_info.pre, line_info.ifun, |
|
248 | 248 | line_info.the_rest) |
|
249 | 249 | |
|
250 | 250 | def _tr_paren(line_info): |
|
251 | 251 | "Translate lines escaped with: /" |
|
252 | 252 | return '%s%s(%s)' % (line_info.pre, line_info.ifun, |
|
253 | 253 | ", ".join(line_info.the_rest.split())) |
|
254 | 254 | |
|
255 | 255 | tr = { ESC_SHELL : _tr_system, |
|
256 | 256 | ESC_SH_CAP : _tr_system2, |
|
257 | 257 | ESC_HELP : _tr_help, |
|
258 | 258 | ESC_HELP2 : _tr_help, |
|
259 | 259 | ESC_MAGIC : _tr_magic, |
|
260 | 260 | ESC_QUOTE : _tr_quote, |
|
261 | 261 | ESC_QUOTE2 : _tr_quote2, |
|
262 | 262 | ESC_PAREN : _tr_paren } |
|
263 | 263 | |
|
264 | 264 | @StatelessInputTransformer.wrap |
|
265 | 265 | def escaped_commands(line): |
|
266 | 266 | """Transform escaped commands - %magic, !system, ?help + various autocalls. |
|
267 | 267 | """ |
|
268 | 268 | if not line or line.isspace(): |
|
269 | 269 | return line |
|
270 | 270 | lineinf = LineInfo(line) |
|
271 | 271 | if lineinf.esc not in tr: |
|
272 | 272 | return line |
|
273 | 273 | |
|
274 | 274 | return tr[lineinf.esc](lineinf) |
|
275 | 275 | |
|
276 | 276 | _initial_space_re = re.compile(r'\s*') |
|
277 | 277 | |
|
278 | 278 | _help_end_re = re.compile(r"""(%{0,2} |
|
279 | 279 | [a-zA-Z_*][\w*]* # Variable name |
|
280 | 280 | (\.[a-zA-Z_*][\w*]*)* # .etc.etc |
|
281 | 281 | ) |
|
282 | 282 | (\?\??)$ # ? or ?? |
|
283 | 283 | """, |
|
284 | 284 | re.VERBOSE) |
|
285 | 285 | |
|
286 | 286 | # Extra pseudotokens for multiline strings and data structures |
|
287 | 287 | _MULTILINE_STRING = object() |
|
288 | 288 | _MULTILINE_STRUCTURE = object() |
|
289 | 289 | |
|
290 | 290 | def _line_tokens(line): |
|
291 | 291 | """Helper for has_comment and ends_in_comment_or_string.""" |
|
292 | 292 | readline = StringIO(line).readline |
|
293 | 293 | toktypes = set() |
|
294 | 294 | try: |
|
295 | 295 | for t in generate_tokens(readline): |
|
296 | 296 | toktypes.add(t[0]) |
|
297 | 297 | except TokenError as e: |
|
298 | 298 | # There are only two cases where a TokenError is raised. |
|
299 | 299 | if 'multi-line string' in e.args[0]: |
|
300 | 300 | toktypes.add(_MULTILINE_STRING) |
|
301 | 301 | else: |
|
302 | 302 | toktypes.add(_MULTILINE_STRUCTURE) |
|
303 | 303 | return toktypes |
|
304 | 304 | |
|
305 | 305 | def has_comment(src): |
|
306 | 306 | """Indicate whether an input line has (i.e. ends in, or is) a comment. |
|
307 | 307 | |
|
308 | 308 | This uses tokenize, so it can distinguish comments from # inside strings. |
|
309 | 309 | |
|
310 | 310 | Parameters |
|
311 | 311 | ---------- |
|
312 | 312 | src : string |
|
313 | 313 | A single line input string. |
|
314 | 314 | |
|
315 | 315 | Returns |
|
316 | 316 | ------- |
|
317 | 317 | comment : bool |
|
318 | 318 | True if source has a comment. |
|
319 | 319 | """ |
|
320 | 320 | return (tokenize2.COMMENT in _line_tokens(src)) |
|
321 | 321 | |
|
322 | 322 | def ends_in_comment_or_string(src): |
|
323 | 323 | """Indicates whether or not an input line ends in a comment or within |
|
324 | 324 | a multiline string. |
|
325 | 325 | |
|
326 | 326 | Parameters |
|
327 | 327 | ---------- |
|
328 | 328 | src : string |
|
329 | 329 | A single line input string. |
|
330 | 330 | |
|
331 | 331 | Returns |
|
332 | 332 | ------- |
|
333 | 333 | comment : bool |
|
334 | 334 | True if source ends in a comment or multiline string. |
|
335 | 335 | """ |
|
336 | 336 | toktypes = _line_tokens(src) |
|
337 | 337 | return (tokenize2.COMMENT in toktypes) or (_MULTILINE_STRING in toktypes) |
|
338 | 338 | |
|
339 | 339 | |
|
340 | 340 | @StatelessInputTransformer.wrap |
|
341 | 341 | def help_end(line): |
|
342 | 342 | """Translate lines with ?/?? at the end""" |
|
343 | 343 | m = _help_end_re.search(line) |
|
344 | 344 | if m is None or ends_in_comment_or_string(line): |
|
345 | 345 | return line |
|
346 | 346 | target = m.group(1) |
|
347 | 347 | esc = m.group(3) |
|
348 | 348 | lspace = _initial_space_re.match(line).group(0) |
|
349 | 349 | |
|
350 | 350 | # If we're mid-command, put it back on the next prompt for the user. |
|
351 | 351 | next_input = line.rstrip('?') if line.strip() != m.group(0) else None |
|
352 | 352 | |
|
353 | 353 | return _make_help_call(target, esc, lspace, next_input) |
|
354 | 354 | |
|
355 | 355 | |
|
356 | 356 | @CoroutineInputTransformer.wrap |
|
357 | 357 | def cellmagic(end_on_blank_line=False): |
|
358 | 358 | """Captures & transforms cell magics. |
|
359 | 359 | |
|
360 | 360 | After a cell magic is started, this stores up any lines it gets until it is |
|
361 | 361 | reset (sent None). |
|
362 | 362 | """ |
|
363 | 363 | tpl = 'get_ipython().run_cell_magic(%r, %r, %r)' |
|
364 | cellmagic_help_re = re.compile('%%\w+\?') | |
|
364 | cellmagic_help_re = re.compile(r'%%\w+\?') | |
|
365 | 365 | line = '' |
|
366 | 366 | while True: |
|
367 | 367 | line = (yield line) |
|
368 | 368 | # consume leading empty lines |
|
369 | 369 | while not line: |
|
370 | 370 | line = (yield line) |
|
371 | 371 | |
|
372 | 372 | if not line.startswith(ESC_MAGIC2): |
|
373 | 373 | # This isn't a cell magic, idle waiting for reset then start over |
|
374 | 374 | while line is not None: |
|
375 | 375 | line = (yield line) |
|
376 | 376 | continue |
|
377 | 377 | |
|
378 | 378 | if cellmagic_help_re.match(line): |
|
379 | 379 | # This case will be handled by help_end |
|
380 | 380 | continue |
|
381 | 381 | |
|
382 | 382 | first = line |
|
383 | 383 | body = [] |
|
384 | 384 | line = (yield None) |
|
385 | 385 | while (line is not None) and \ |
|
386 | 386 | ((line.strip() != '') or not end_on_blank_line): |
|
387 | 387 | body.append(line) |
|
388 | 388 | line = (yield None) |
|
389 | 389 | |
|
390 | 390 | # Output |
|
391 | 391 | magic_name, _, first = first.partition(' ') |
|
392 | 392 | magic_name = magic_name.lstrip(ESC_MAGIC2) |
|
393 | 393 | line = tpl % (magic_name, first, u'\n'.join(body)) |
|
394 | 394 | |
|
395 | 395 | |
|
396 | 396 | def _strip_prompts(prompt_re, initial_re=None, turnoff_re=None): |
|
397 | 397 | """Remove matching input prompts from a block of input. |
|
398 | 398 | |
|
399 | 399 | Parameters |
|
400 | 400 | ---------- |
|
401 | 401 | prompt_re : regular expression |
|
402 | 402 | A regular expression matching any input prompt (including continuation) |
|
403 | 403 | initial_re : regular expression, optional |
|
404 | 404 | A regular expression matching only the initial prompt, but not continuation. |
|
405 | 405 | If no initial expression is given, prompt_re will be used everywhere. |
|
406 | 406 | Used mainly for plain Python prompts, where the continuation prompt |
|
407 | 407 | ``...`` is a valid Python expression in Python 3, so shouldn't be stripped. |
|
408 | 408 | |
|
409 | 409 | If initial_re and prompt_re differ, |
|
410 | 410 | only initial_re will be tested against the first line. |
|
411 | 411 | If any prompt is found on the first two lines, |
|
412 | 412 | prompts will be stripped from the rest of the block. |
|
413 | 413 | """ |
|
414 | 414 | if initial_re is None: |
|
415 | 415 | initial_re = prompt_re |
|
416 | 416 | line = '' |
|
417 | 417 | while True: |
|
418 | 418 | line = (yield line) |
|
419 | 419 | |
|
420 | 420 | # First line of cell |
|
421 | 421 | if line is None: |
|
422 | 422 | continue |
|
423 | 423 | out, n1 = initial_re.subn('', line, count=1) |
|
424 | 424 | if turnoff_re and not n1: |
|
425 | 425 | if turnoff_re.match(line): |
|
426 | 426 | # We're in e.g. a cell magic; disable this transformer for |
|
427 | 427 | # the rest of the cell. |
|
428 | 428 | while line is not None: |
|
429 | 429 | line = (yield line) |
|
430 | 430 | continue |
|
431 | 431 | |
|
432 | 432 | line = (yield out) |
|
433 | 433 | |
|
434 | 434 | if line is None: |
|
435 | 435 | continue |
|
436 | 436 | # check for any prompt on the second line of the cell, |
|
437 | 437 | # because people often copy from just after the first prompt, |
|
438 | 438 | # so we might not see it in the first line. |
|
439 | 439 | out, n2 = prompt_re.subn('', line, count=1) |
|
440 | 440 | line = (yield out) |
|
441 | 441 | |
|
442 | 442 | if n1 or n2: |
|
443 | 443 | # Found a prompt in the first two lines - check for it in |
|
444 | 444 | # the rest of the cell as well. |
|
445 | 445 | while line is not None: |
|
446 | 446 | line = (yield prompt_re.sub('', line, count=1)) |
|
447 | 447 | |
|
448 | 448 | else: |
|
449 | 449 | # Prompts not in input - wait for reset |
|
450 | 450 | while line is not None: |
|
451 | 451 | line = (yield line) |
|
452 | 452 | |
|
453 | 453 | @CoroutineInputTransformer.wrap |
|
454 | 454 | def classic_prompt(): |
|
455 | 455 | """Strip the >>>/... prompts of the Python interactive shell.""" |
|
456 | 456 | # FIXME: non-capturing version (?:...) usable? |
|
457 | 457 | prompt_re = re.compile(r'^(>>>|\.\.\.)( |$)') |
|
458 | 458 | initial_re = re.compile(r'^>>>( |$)') |
|
459 | 459 | # Any %magic/!system is IPython syntax, so we needn't look for >>> prompts |
|
460 | 460 | turnoff_re = re.compile(r'^[%!]') |
|
461 | 461 | return _strip_prompts(prompt_re, initial_re, turnoff_re) |
|
462 | 462 | |
|
463 | 463 | @CoroutineInputTransformer.wrap |
|
464 | 464 | def ipy_prompt(): |
|
465 | 465 | """Strip IPython's In [1]:/...: prompts.""" |
|
466 | 466 | # FIXME: non-capturing version (?:...) usable? |
|
467 | 467 | prompt_re = re.compile(r'^(In \[\d+\]: |\s*\.{3,}: ?)') |
|
468 | 468 | # Disable prompt stripping inside cell magics |
|
469 | 469 | turnoff_re = re.compile(r'^%%') |
|
470 | 470 | return _strip_prompts(prompt_re, turnoff_re=turnoff_re) |
|
471 | 471 | |
|
472 | 472 | |
|
473 | 473 | @CoroutineInputTransformer.wrap |
|
474 | 474 | def leading_indent(): |
|
475 | 475 | """Remove leading indentation. |
|
476 | 476 | |
|
477 | 477 | If the first line starts with a spaces or tabs, the same whitespace will be |
|
478 | 478 | removed from each following line until it is reset. |
|
479 | 479 | """ |
|
480 | 480 | space_re = re.compile(r'^[ \t]+') |
|
481 | 481 | line = '' |
|
482 | 482 | while True: |
|
483 | 483 | line = (yield line) |
|
484 | 484 | |
|
485 | 485 | if line is None: |
|
486 | 486 | continue |
|
487 | 487 | |
|
488 | 488 | m = space_re.match(line) |
|
489 | 489 | if m: |
|
490 | 490 | space = m.group(0) |
|
491 | 491 | while line is not None: |
|
492 | 492 | if line.startswith(space): |
|
493 | 493 | line = line[len(space):] |
|
494 | 494 | line = (yield line) |
|
495 | 495 | else: |
|
496 | 496 | # No leading spaces - wait for reset |
|
497 | 497 | while line is not None: |
|
498 | 498 | line = (yield line) |
|
499 | 499 | |
|
500 | 500 | |
|
501 | 501 | _assign_pat = \ |
|
502 | 502 | r'''(?P<lhs>(\s*) |
|
503 | 503 | ([\w\.]+) # Initial identifier |
|
504 | 504 | (\s*,\s* |
|
505 | 505 | \*?[\w\.]+)* # Further identifiers for unpacking |
|
506 | 506 | \s*?,? # Trailing comma |
|
507 | 507 | ) |
|
508 | 508 | \s*=\s* |
|
509 | 509 | ''' |
|
510 | 510 | |
|
511 | 511 | assign_system_re = re.compile(r'{}!\s*(?P<cmd>.*)'.format(_assign_pat), re.VERBOSE) |
|
512 | 512 | assign_system_template = '%s = get_ipython().getoutput(%r)' |
|
513 | 513 | @StatelessInputTransformer.wrap |
|
514 | 514 | def assign_from_system(line): |
|
515 | 515 | """Transform assignment from system commands (e.g. files = !ls)""" |
|
516 | 516 | m = assign_system_re.match(line) |
|
517 | 517 | if m is None: |
|
518 | 518 | return line |
|
519 | 519 | |
|
520 | 520 | return assign_system_template % m.group('lhs', 'cmd') |
|
521 | 521 | |
|
522 | 522 | assign_magic_re = re.compile(r'{}%\s*(?P<cmd>.*)'.format(_assign_pat), re.VERBOSE) |
|
523 | 523 | assign_magic_template = '%s = get_ipython().run_line_magic(%r, %r)' |
|
524 | 524 | @StatelessInputTransformer.wrap |
|
525 | 525 | def assign_from_magic(line): |
|
526 | 526 | """Transform assignment from magic commands (e.g. a = %who_ls)""" |
|
527 | 527 | m = assign_magic_re.match(line) |
|
528 | 528 | if m is None: |
|
529 | 529 | return line |
|
530 | 530 | #Prepare arguments for get_ipython().run_line_magic(magic_name, magic_args) |
|
531 | 531 | m_lhs, m_cmd = m.group('lhs', 'cmd') |
|
532 | 532 | t_magic_name, _, t_magic_arg_s = m_cmd.partition(' ') |
|
533 | 533 | t_magic_name = t_magic_name.lstrip(ESC_MAGIC) |
|
534 | 534 | return assign_magic_template % (m_lhs, t_magic_name, t_magic_arg_s) |
@@ -1,158 +1,158 b'' | |||
|
1 | 1 | """Implementation of configuration-related magic functions. |
|
2 | 2 | """ |
|
3 | 3 | #----------------------------------------------------------------------------- |
|
4 | 4 | # Copyright (c) 2012 The IPython Development Team. |
|
5 | 5 | # |
|
6 | 6 | # Distributed under the terms of the Modified BSD License. |
|
7 | 7 | # |
|
8 | 8 | # The full license is in the file COPYING.txt, distributed with this software. |
|
9 | 9 | #----------------------------------------------------------------------------- |
|
10 | 10 | |
|
11 | 11 | #----------------------------------------------------------------------------- |
|
12 | 12 | # Imports |
|
13 | 13 | #----------------------------------------------------------------------------- |
|
14 | 14 | |
|
15 | 15 | # Stdlib |
|
16 | 16 | import re |
|
17 | 17 | |
|
18 | 18 | # Our own packages |
|
19 | 19 | from IPython.core.error import UsageError |
|
20 | 20 | from IPython.core.magic import Magics, magics_class, line_magic |
|
21 | 21 | from logging import error |
|
22 | 22 | |
|
23 | 23 | #----------------------------------------------------------------------------- |
|
24 | 24 | # Magic implementation classes |
|
25 | 25 | #----------------------------------------------------------------------------- |
|
26 | 26 | |
|
27 | reg = re.compile('^\w+\.\w+$') | |
|
27 | reg = re.compile(r'^\w+\.\w+$') | |
|
28 | 28 | @magics_class |
|
29 | 29 | class ConfigMagics(Magics): |
|
30 | 30 | |
|
31 | 31 | def __init__(self, shell): |
|
32 | 32 | super(ConfigMagics, self).__init__(shell) |
|
33 | 33 | self.configurables = [] |
|
34 | 34 | |
|
35 | 35 | @line_magic |
|
36 | 36 | def config(self, s): |
|
37 | 37 | """configure IPython |
|
38 | 38 | |
|
39 | 39 | %config Class[.trait=value] |
|
40 | 40 | |
|
41 | 41 | This magic exposes most of the IPython config system. Any |
|
42 | 42 | Configurable class should be able to be configured with the simple |
|
43 | 43 | line:: |
|
44 | 44 | |
|
45 | 45 | %config Class.trait=value |
|
46 | 46 | |
|
47 | 47 | Where `value` will be resolved in the user's namespace, if it is an |
|
48 | 48 | expression or variable name. |
|
49 | 49 | |
|
50 | 50 | Examples |
|
51 | 51 | -------- |
|
52 | 52 | |
|
53 | 53 | To see what classes are available for config, pass no arguments:: |
|
54 | 54 | |
|
55 | 55 | In [1]: %config |
|
56 | 56 | Available objects for config: |
|
57 | 57 | TerminalInteractiveShell |
|
58 | 58 | HistoryManager |
|
59 | 59 | PrefilterManager |
|
60 | 60 | AliasManager |
|
61 | 61 | IPCompleter |
|
62 | 62 | DisplayFormatter |
|
63 | 63 | |
|
64 | 64 | To view what is configurable on a given class, just pass the class |
|
65 | 65 | name:: |
|
66 | 66 | |
|
67 | 67 | In [2]: %config IPCompleter |
|
68 | 68 | IPCompleter options |
|
69 | 69 | ----------------- |
|
70 | 70 | IPCompleter.omit__names=<Enum> |
|
71 | 71 | Current: 2 |
|
72 | 72 | Choices: (0, 1, 2) |
|
73 | 73 | Instruct the completer to omit private method names |
|
74 | 74 | Specifically, when completing on ``object.<tab>``. |
|
75 | 75 | When 2 [default]: all names that start with '_' will be excluded. |
|
76 | 76 | When 1: all 'magic' names (``__foo__``) will be excluded. |
|
77 | 77 | When 0: nothing will be excluded. |
|
78 | 78 | IPCompleter.merge_completions=<CBool> |
|
79 | 79 | Current: True |
|
80 | 80 | Whether to merge completion results into a single list |
|
81 | 81 | If False, only the completion results from the first non-empty |
|
82 | 82 | completer will be returned. |
|
83 | 83 | IPCompleter.limit_to__all__=<CBool> |
|
84 | 84 | Current: False |
|
85 | 85 | Instruct the completer to use __all__ for the completion |
|
86 | 86 | Specifically, when completing on ``object.<tab>``. |
|
87 | 87 | When True: only those names in obj.__all__ will be included. |
|
88 | 88 | When False [default]: the __all__ attribute is ignored |
|
89 | 89 | IPCompleter.greedy=<CBool> |
|
90 | 90 | Current: False |
|
91 | 91 | Activate greedy completion |
|
92 | 92 | This will enable completion on elements of lists, results of |
|
93 | 93 | function calls, etc., but can be unsafe because the code is |
|
94 | 94 | actually evaluated on TAB. |
|
95 | 95 | |
|
96 | 96 | but the real use is in setting values:: |
|
97 | 97 | |
|
98 | 98 | In [3]: %config IPCompleter.greedy = True |
|
99 | 99 | |
|
100 | 100 | and these values are read from the user_ns if they are variables:: |
|
101 | 101 | |
|
102 | 102 | In [4]: feeling_greedy=False |
|
103 | 103 | |
|
104 | 104 | In [5]: %config IPCompleter.greedy = feeling_greedy |
|
105 | 105 | |
|
106 | 106 | """ |
|
107 | 107 | from traitlets.config.loader import Config |
|
108 | 108 | # some IPython objects are Configurable, but do not yet have |
|
109 | 109 | # any configurable traits. Exclude them from the effects of |
|
110 | 110 | # this magic, as their presence is just noise: |
|
111 | 111 | configurables = sorted(set([ c for c in self.shell.configurables |
|
112 | 112 | if c.__class__.class_traits(config=True) |
|
113 | 113 | ]), key=lambda x: x.__class__.__name__) |
|
114 | 114 | classnames = [ c.__class__.__name__ for c in configurables ] |
|
115 | 115 | |
|
116 | 116 | line = s.strip() |
|
117 | 117 | if not line: |
|
118 | 118 | # print available configurable names |
|
119 | 119 | print("Available objects for config:") |
|
120 | 120 | for name in classnames: |
|
121 | 121 | print(" ", name) |
|
122 | 122 | return |
|
123 | 123 | elif line in classnames: |
|
124 | 124 | # `%config TerminalInteractiveShell` will print trait info for |
|
125 | 125 | # TerminalInteractiveShell |
|
126 | 126 | c = configurables[classnames.index(line)] |
|
127 | 127 | cls = c.__class__ |
|
128 | 128 | help = cls.class_get_help(c) |
|
129 | 129 | # strip leading '--' from cl-args: |
|
130 | 130 | help = re.sub(re.compile(r'^--', re.MULTILINE), '', help) |
|
131 | 131 | print(help) |
|
132 | 132 | return |
|
133 | 133 | elif reg.match(line): |
|
134 | 134 | cls, attr = line.split('.') |
|
135 | 135 | return getattr(configurables[classnames.index(cls)],attr) |
|
136 | 136 | elif '=' not in line: |
|
137 | 137 | msg = "Invalid config statement: %r, "\ |
|
138 | 138 | "should be `Class.trait = value`." |
|
139 | 139 | |
|
140 | 140 | ll = line.lower() |
|
141 | 141 | for classname in classnames: |
|
142 | 142 | if ll == classname.lower(): |
|
143 | 143 | msg = msg + '\nDid you mean %s (note the case)?' % classname |
|
144 | 144 | break |
|
145 | 145 | |
|
146 | 146 | raise UsageError( msg % line) |
|
147 | 147 | |
|
148 | 148 | # otherwise, assume we are setting configurables. |
|
149 | 149 | # leave quotes on args when splitting, because we want |
|
150 | 150 | # unquoted args to eval in user_ns |
|
151 | 151 | cfg = Config() |
|
152 | 152 | exec("cfg."+line, locals(), self.shell.user_ns) |
|
153 | 153 | |
|
154 | 154 | for configurable in configurables: |
|
155 | 155 | try: |
|
156 | 156 | configurable.update_config(cfg) |
|
157 | 157 | except Exception as e: |
|
158 | 158 | error(e) |
@@ -1,137 +1,137 b'' | |||
|
1 | 1 | # encoding: utf-8 |
|
2 | 2 | """ |
|
3 | 3 | Simple utility for splitting user input. This is used by both inputsplitter and |
|
4 | 4 | prefilter. |
|
5 | 5 | |
|
6 | 6 | Authors: |
|
7 | 7 | |
|
8 | 8 | * Brian Granger |
|
9 | 9 | * Fernando Perez |
|
10 | 10 | """ |
|
11 | 11 | |
|
12 | 12 | #----------------------------------------------------------------------------- |
|
13 | 13 | # Copyright (C) 2008-2011 The IPython Development Team |
|
14 | 14 | # |
|
15 | 15 | # Distributed under the terms of the BSD License. The full license is in |
|
16 | 16 | # the file COPYING, distributed as part of this software. |
|
17 | 17 | #----------------------------------------------------------------------------- |
|
18 | 18 | |
|
19 | 19 | #----------------------------------------------------------------------------- |
|
20 | 20 | # Imports |
|
21 | 21 | #----------------------------------------------------------------------------- |
|
22 | 22 | |
|
23 | 23 | import re |
|
24 | 24 | import sys |
|
25 | 25 | |
|
26 | 26 | from IPython.utils import py3compat |
|
27 | 27 | from IPython.utils.encoding import get_stream_enc |
|
28 | 28 | |
|
29 | 29 | #----------------------------------------------------------------------------- |
|
30 | 30 | # Main function |
|
31 | 31 | #----------------------------------------------------------------------------- |
|
32 | 32 | |
|
33 | 33 | # RegExp for splitting line contents into pre-char//first word-method//rest. |
|
34 | 34 | # For clarity, each group in on one line. |
|
35 | 35 | |
|
36 | 36 | # WARNING: update the regexp if the escapes in interactiveshell are changed, as |
|
37 | 37 | # they are hardwired in. |
|
38 | 38 | |
|
39 | 39 | # Although it's not solely driven by the regex, note that: |
|
40 | 40 | # ,;/% only trigger if they are the first character on the line |
|
41 | 41 | # ! and !! trigger if they are first char(s) *or* follow an indent |
|
42 | 42 | # ? triggers as first or last char. |
|
43 | 43 | |
|
44 | line_split = re.compile(""" | |
|
44 | line_split = re.compile(r""" | |
|
45 | 45 | ^(\s*) # any leading space |
|
46 | 46 | ([,;/%]|!!?|\?\??)? # escape character or characters |
|
47 | 47 | \s*(%{0,2}[\w\.\*]*) # function/method, possibly with leading % |
|
48 | 48 | # to correctly treat things like '?%magic' |
|
49 | 49 | (.*?$|$) # rest of line |
|
50 | 50 | """, re.VERBOSE) |
|
51 | 51 | |
|
52 | 52 | |
|
53 | 53 | def split_user_input(line, pattern=None): |
|
54 | 54 | """Split user input into initial whitespace, escape character, function part |
|
55 | 55 | and the rest. |
|
56 | 56 | """ |
|
57 | 57 | # We need to ensure that the rest of this routine deals only with unicode |
|
58 | 58 | encoding = get_stream_enc(sys.stdin, 'utf-8') |
|
59 | 59 | line = py3compat.cast_unicode(line, encoding) |
|
60 | 60 | |
|
61 | 61 | if pattern is None: |
|
62 | 62 | pattern = line_split |
|
63 | 63 | match = pattern.match(line) |
|
64 | 64 | if not match: |
|
65 | 65 | # print "match failed for line '%s'" % line |
|
66 | 66 | try: |
|
67 | 67 | ifun, the_rest = line.split(None,1) |
|
68 | 68 | except ValueError: |
|
69 | 69 | # print "split failed for line '%s'" % line |
|
70 | 70 | ifun, the_rest = line, u'' |
|
71 | pre = re.match('^(\s*)(.*)',line).groups()[0] | |
|
71 | pre = re.match(r'^(\s*)(.*)',line).groups()[0] | |
|
72 | 72 | esc = "" |
|
73 | 73 | else: |
|
74 | 74 | pre, esc, ifun, the_rest = match.groups() |
|
75 | 75 | |
|
76 | 76 | #print 'line:<%s>' % line # dbg |
|
77 | 77 | #print 'pre <%s> ifun <%s> rest <%s>' % (pre,ifun.strip(),the_rest) # dbg |
|
78 | 78 | return pre, esc or '', ifun.strip(), the_rest.lstrip() |
|
79 | 79 | |
|
80 | 80 | |
|
81 | 81 | class LineInfo(object): |
|
82 | 82 | """A single line of input and associated info. |
|
83 | 83 | |
|
84 | 84 | Includes the following as properties: |
|
85 | 85 | |
|
86 | 86 | line |
|
87 | 87 | The original, raw line |
|
88 | 88 | |
|
89 | 89 | continue_prompt |
|
90 | 90 | Is this line a continuation in a sequence of multiline input? |
|
91 | 91 | |
|
92 | 92 | pre |
|
93 | 93 | Any leading whitespace. |
|
94 | 94 | |
|
95 | 95 | esc |
|
96 | 96 | The escape character(s) in pre or the empty string if there isn't one. |
|
97 | 97 | Note that '!!' and '??' are possible values for esc. Otherwise it will |
|
98 | 98 | always be a single character. |
|
99 | 99 | |
|
100 | 100 | ifun |
|
101 | 101 | The 'function part', which is basically the maximal initial sequence |
|
102 | 102 | of valid python identifiers and the '.' character. This is what is |
|
103 | 103 | checked for alias and magic transformations, used for auto-calling, |
|
104 | 104 | etc. In contrast to Python identifiers, it may start with "%" and contain |
|
105 | 105 | "*". |
|
106 | 106 | |
|
107 | 107 | the_rest |
|
108 | 108 | Everything else on the line. |
|
109 | 109 | """ |
|
110 | 110 | def __init__(self, line, continue_prompt=False): |
|
111 | 111 | self.line = line |
|
112 | 112 | self.continue_prompt = continue_prompt |
|
113 | 113 | self.pre, self.esc, self.ifun, self.the_rest = split_user_input(line) |
|
114 | 114 | |
|
115 | 115 | self.pre_char = self.pre.strip() |
|
116 | 116 | if self.pre_char: |
|
117 | 117 | self.pre_whitespace = '' # No whitespace allowed before esc chars |
|
118 | 118 | else: |
|
119 | 119 | self.pre_whitespace = self.pre |
|
120 | 120 | |
|
121 | 121 | def ofind(self, ip): |
|
122 | 122 | """Do a full, attribute-walking lookup of the ifun in the various |
|
123 | 123 | namespaces for the given IPython InteractiveShell instance. |
|
124 | 124 | |
|
125 | 125 | Return a dict with keys: {found, obj, ospace, ismagic} |
|
126 | 126 | |
|
127 | 127 | Note: can cause state changes because of calling getattr, but should |
|
128 | 128 | only be run if autocall is on and if the line hasn't matched any |
|
129 | 129 | other, less dangerous handlers. |
|
130 | 130 | |
|
131 | 131 | Does cache the results of the call, so can be called multiple times |
|
132 | 132 | without worrying about *further* damaging state. |
|
133 | 133 | """ |
|
134 | 134 | return ip._ofind(self.ifun) |
|
135 | 135 | |
|
136 | 136 | def __str__(self): |
|
137 | 137 | return "LineInfo [%s|%s|%s|%s]" %(self.pre, self.esc, self.ifun, self.the_rest) |
@@ -1,438 +1,438 b'' | |||
|
1 | 1 | # encoding: utf-8 |
|
2 | 2 | """ |
|
3 | 3 | Utilities for path handling. |
|
4 | 4 | """ |
|
5 | 5 | |
|
6 | 6 | # Copyright (c) IPython Development Team. |
|
7 | 7 | # Distributed under the terms of the Modified BSD License. |
|
8 | 8 | |
|
9 | 9 | import os |
|
10 | 10 | import sys |
|
11 | 11 | import errno |
|
12 | 12 | import shutil |
|
13 | 13 | import random |
|
14 | 14 | import glob |
|
15 | 15 | from warnings import warn |
|
16 | 16 | |
|
17 | 17 | from IPython.utils.process import system |
|
18 | 18 | from IPython.utils import py3compat |
|
19 | 19 | from IPython.utils.decorators import undoc |
|
20 | 20 | |
|
21 | 21 | #----------------------------------------------------------------------------- |
|
22 | 22 | # Code |
|
23 | 23 | #----------------------------------------------------------------------------- |
|
24 | 24 | |
|
25 | 25 | fs_encoding = sys.getfilesystemencoding() |
|
26 | 26 | |
|
27 | 27 | def _writable_dir(path): |
|
28 | 28 | """Whether `path` is a directory, to which the user has write access.""" |
|
29 | 29 | return os.path.isdir(path) and os.access(path, os.W_OK) |
|
30 | 30 | |
|
31 | 31 | if sys.platform == 'win32': |
|
32 | 32 | def _get_long_path_name(path): |
|
33 | 33 | """Get a long path name (expand ~) on Windows using ctypes. |
|
34 | 34 | |
|
35 | 35 | Examples |
|
36 | 36 | -------- |
|
37 | 37 | |
|
38 | 38 | >>> get_long_path_name('c:\\docume~1') |
|
39 | 39 | 'c:\\\\Documents and Settings' |
|
40 | 40 | |
|
41 | 41 | """ |
|
42 | 42 | try: |
|
43 | 43 | import ctypes |
|
44 | 44 | except ImportError: |
|
45 | 45 | raise ImportError('you need to have ctypes installed for this to work') |
|
46 | 46 | _GetLongPathName = ctypes.windll.kernel32.GetLongPathNameW |
|
47 | 47 | _GetLongPathName.argtypes = [ctypes.c_wchar_p, ctypes.c_wchar_p, |
|
48 | 48 | ctypes.c_uint ] |
|
49 | 49 | |
|
50 | 50 | buf = ctypes.create_unicode_buffer(260) |
|
51 | 51 | rv = _GetLongPathName(path, buf, 260) |
|
52 | 52 | if rv == 0 or rv > 260: |
|
53 | 53 | return path |
|
54 | 54 | else: |
|
55 | 55 | return buf.value |
|
56 | 56 | else: |
|
57 | 57 | def _get_long_path_name(path): |
|
58 | 58 | """Dummy no-op.""" |
|
59 | 59 | return path |
|
60 | 60 | |
|
61 | 61 | |
|
62 | 62 | |
|
63 | 63 | def get_long_path_name(path): |
|
64 | 64 | """Expand a path into its long form. |
|
65 | 65 | |
|
66 | 66 | On Windows this expands any ~ in the paths. On other platforms, it is |
|
67 | 67 | a null operation. |
|
68 | 68 | """ |
|
69 | 69 | return _get_long_path_name(path) |
|
70 | 70 | |
|
71 | 71 | |
|
72 | 72 | def unquote_filename(name, win32=(sys.platform=='win32')): |
|
73 | 73 | """ On Windows, remove leading and trailing quotes from filenames. |
|
74 | 74 | |
|
75 | 75 | This function has been deprecated and should not be used any more: |
|
76 | 76 | unquoting is now taken care of by :func:`IPython.utils.process.arg_split`. |
|
77 | 77 | """ |
|
78 | 78 | warn("'unquote_filename' is deprecated since IPython 5.0 and should not " |
|
79 | 79 | "be used anymore", DeprecationWarning, stacklevel=2) |
|
80 | 80 | if win32: |
|
81 | 81 | if name.startswith(("'", '"')) and name.endswith(("'", '"')): |
|
82 | 82 | name = name[1:-1] |
|
83 | 83 | return name |
|
84 | 84 | |
|
85 | 85 | |
|
86 | 86 | def compress_user(path): |
|
87 | 87 | """Reverse of :func:`os.path.expanduser` |
|
88 | 88 | """ |
|
89 | 89 | home = os.path.expanduser('~') |
|
90 | 90 | if path.startswith(home): |
|
91 | 91 | path = "~" + path[len(home):] |
|
92 | 92 | return path |
|
93 | 93 | |
|
94 | 94 | def get_py_filename(name, force_win32=None): |
|
95 | 95 | """Return a valid python filename in the current directory. |
|
96 | 96 | |
|
97 | 97 | If the given name is not a file, it adds '.py' and searches again. |
|
98 | 98 | Raises IOError with an informative message if the file isn't found. |
|
99 | 99 | """ |
|
100 | 100 | |
|
101 | 101 | name = os.path.expanduser(name) |
|
102 | 102 | if force_win32 is not None: |
|
103 | 103 | warn("The 'force_win32' argument to 'get_py_filename' is deprecated " |
|
104 | 104 | "since IPython 5.0 and should not be used anymore", |
|
105 | 105 | DeprecationWarning, stacklevel=2) |
|
106 | 106 | if not os.path.isfile(name) and not name.endswith('.py'): |
|
107 | 107 | name += '.py' |
|
108 | 108 | if os.path.isfile(name): |
|
109 | 109 | return name |
|
110 | 110 | else: |
|
111 | 111 | raise IOError('File `%r` not found.' % name) |
|
112 | 112 | |
|
113 | 113 | |
|
114 | 114 | def filefind(filename, path_dirs=None): |
|
115 | 115 | """Find a file by looking through a sequence of paths. |
|
116 | 116 | |
|
117 | 117 | This iterates through a sequence of paths looking for a file and returns |
|
118 | 118 | the full, absolute path of the first occurrence of the file. If no set of |
|
119 | 119 | path dirs is given, the filename is tested as is, after running through |
|
120 | 120 | :func:`expandvars` and :func:`expanduser`. Thus a simple call:: |
|
121 | 121 | |
|
122 | 122 | filefind('myfile.txt') |
|
123 | 123 | |
|
124 | 124 | will find the file in the current working dir, but:: |
|
125 | 125 | |
|
126 | 126 | filefind('~/myfile.txt') |
|
127 | 127 | |
|
128 | 128 | Will find the file in the users home directory. This function does not |
|
129 | 129 | automatically try any paths, such as the cwd or the user's home directory. |
|
130 | 130 | |
|
131 | 131 | Parameters |
|
132 | 132 | ---------- |
|
133 | 133 | filename : str |
|
134 | 134 | The filename to look for. |
|
135 | 135 | path_dirs : str, None or sequence of str |
|
136 | 136 | The sequence of paths to look for the file in. If None, the filename |
|
137 | 137 | need to be absolute or be in the cwd. If a string, the string is |
|
138 | 138 | put into a sequence and the searched. If a sequence, walk through |
|
139 | 139 | each element and join with ``filename``, calling :func:`expandvars` |
|
140 | 140 | and :func:`expanduser` before testing for existence. |
|
141 | 141 | |
|
142 | 142 | Returns |
|
143 | 143 | ------- |
|
144 | 144 | Raises :exc:`IOError` or returns absolute path to file. |
|
145 | 145 | """ |
|
146 | 146 | |
|
147 | 147 | # If paths are quoted, abspath gets confused, strip them... |
|
148 | 148 | filename = filename.strip('"').strip("'") |
|
149 | 149 | # If the input is an absolute path, just check it exists |
|
150 | 150 | if os.path.isabs(filename) and os.path.isfile(filename): |
|
151 | 151 | return filename |
|
152 | 152 | |
|
153 | 153 | if path_dirs is None: |
|
154 | 154 | path_dirs = ("",) |
|
155 | 155 | elif isinstance(path_dirs, str): |
|
156 | 156 | path_dirs = (path_dirs,) |
|
157 | 157 | |
|
158 | 158 | for path in path_dirs: |
|
159 | 159 | if path == '.': path = os.getcwd() |
|
160 | 160 | testname = expand_path(os.path.join(path, filename)) |
|
161 | 161 | if os.path.isfile(testname): |
|
162 | 162 | return os.path.abspath(testname) |
|
163 | 163 | |
|
164 | 164 | raise IOError("File %r does not exist in any of the search paths: %r" % |
|
165 | 165 | (filename, path_dirs) ) |
|
166 | 166 | |
|
167 | 167 | |
|
168 | 168 | class HomeDirError(Exception): |
|
169 | 169 | pass |
|
170 | 170 | |
|
171 | 171 | |
|
172 | 172 | def get_home_dir(require_writable=False): |
|
173 | 173 | """Return the 'home' directory, as a unicode string. |
|
174 | 174 | |
|
175 | 175 | Uses os.path.expanduser('~'), and checks for writability. |
|
176 | 176 | |
|
177 | 177 | See stdlib docs for how this is determined. |
|
178 | 178 | $HOME is first priority on *ALL* platforms. |
|
179 | 179 | |
|
180 | 180 | Parameters |
|
181 | 181 | ---------- |
|
182 | 182 | |
|
183 | 183 | require_writable : bool [default: False] |
|
184 | 184 | if True: |
|
185 | 185 | guarantees the return value is a writable directory, otherwise |
|
186 | 186 | raises HomeDirError |
|
187 | 187 | if False: |
|
188 | 188 | The path is resolved, but it is not guaranteed to exist or be writable. |
|
189 | 189 | """ |
|
190 | 190 | |
|
191 | 191 | homedir = os.path.expanduser('~') |
|
192 | 192 | # Next line will make things work even when /home/ is a symlink to |
|
193 | 193 | # /usr/home as it is on FreeBSD, for example |
|
194 | 194 | homedir = os.path.realpath(homedir) |
|
195 | 195 | |
|
196 | 196 | if not _writable_dir(homedir) and os.name == 'nt': |
|
197 | 197 | # expanduser failed, use the registry to get the 'My Documents' folder. |
|
198 | 198 | try: |
|
199 | 199 | try: |
|
200 | 200 | import winreg as wreg # Py 3 |
|
201 | 201 | except ImportError: |
|
202 | 202 | import _winreg as wreg # Py 2 |
|
203 | 203 | key = wreg.OpenKey( |
|
204 | 204 | wreg.HKEY_CURRENT_USER, |
|
205 | "Software\Microsoft\Windows\CurrentVersion\Explorer\Shell Folders" | |
|
205 | r"Software\Microsoft\Windows\CurrentVersion\Explorer\Shell Folders" | |
|
206 | 206 | ) |
|
207 | 207 | homedir = wreg.QueryValueEx(key,'Personal')[0] |
|
208 | 208 | key.Close() |
|
209 | 209 | except: |
|
210 | 210 | pass |
|
211 | 211 | |
|
212 | 212 | if (not require_writable) or _writable_dir(homedir): |
|
213 | 213 | return py3compat.cast_unicode(homedir, fs_encoding) |
|
214 | 214 | else: |
|
215 | 215 | raise HomeDirError('%s is not a writable dir, ' |
|
216 | 216 | 'set $HOME environment variable to override' % homedir) |
|
217 | 217 | |
|
218 | 218 | def get_xdg_dir(): |
|
219 | 219 | """Return the XDG_CONFIG_HOME, if it is defined and exists, else None. |
|
220 | 220 | |
|
221 | 221 | This is only for non-OS X posix (Linux,Unix,etc.) systems. |
|
222 | 222 | """ |
|
223 | 223 | |
|
224 | 224 | env = os.environ |
|
225 | 225 | |
|
226 | 226 | if os.name == 'posix' and sys.platform != 'darwin': |
|
227 | 227 | # Linux, Unix, AIX, etc. |
|
228 | 228 | # use ~/.config if empty OR not set |
|
229 | 229 | xdg = env.get("XDG_CONFIG_HOME", None) or os.path.join(get_home_dir(), '.config') |
|
230 | 230 | if xdg and _writable_dir(xdg): |
|
231 | 231 | return py3compat.cast_unicode(xdg, fs_encoding) |
|
232 | 232 | |
|
233 | 233 | return None |
|
234 | 234 | |
|
235 | 235 | |
|
236 | 236 | def get_xdg_cache_dir(): |
|
237 | 237 | """Return the XDG_CACHE_HOME, if it is defined and exists, else None. |
|
238 | 238 | |
|
239 | 239 | This is only for non-OS X posix (Linux,Unix,etc.) systems. |
|
240 | 240 | """ |
|
241 | 241 | |
|
242 | 242 | env = os.environ |
|
243 | 243 | |
|
244 | 244 | if os.name == 'posix' and sys.platform != 'darwin': |
|
245 | 245 | # Linux, Unix, AIX, etc. |
|
246 | 246 | # use ~/.cache if empty OR not set |
|
247 | 247 | xdg = env.get("XDG_CACHE_HOME", None) or os.path.join(get_home_dir(), '.cache') |
|
248 | 248 | if xdg and _writable_dir(xdg): |
|
249 | 249 | return py3compat.cast_unicode(xdg, fs_encoding) |
|
250 | 250 | |
|
251 | 251 | return None |
|
252 | 252 | |
|
253 | 253 | |
|
254 | 254 | @undoc |
|
255 | 255 | def get_ipython_dir(): |
|
256 | 256 | warn("get_ipython_dir has moved to the IPython.paths module since IPython 4.0.", stacklevel=2) |
|
257 | 257 | from IPython.paths import get_ipython_dir |
|
258 | 258 | return get_ipython_dir() |
|
259 | 259 | |
|
260 | 260 | @undoc |
|
261 | 261 | def get_ipython_cache_dir(): |
|
262 | 262 | warn("get_ipython_cache_dir has moved to the IPython.paths module since IPython 4.0.", stacklevel=2) |
|
263 | 263 | from IPython.paths import get_ipython_cache_dir |
|
264 | 264 | return get_ipython_cache_dir() |
|
265 | 265 | |
|
266 | 266 | @undoc |
|
267 | 267 | def get_ipython_package_dir(): |
|
268 | 268 | warn("get_ipython_package_dir has moved to the IPython.paths module since IPython 4.0.", stacklevel=2) |
|
269 | 269 | from IPython.paths import get_ipython_package_dir |
|
270 | 270 | return get_ipython_package_dir() |
|
271 | 271 | |
|
272 | 272 | @undoc |
|
273 | 273 | def get_ipython_module_path(module_str): |
|
274 | 274 | warn("get_ipython_module_path has moved to the IPython.paths module since IPython 4.0.", stacklevel=2) |
|
275 | 275 | from IPython.paths import get_ipython_module_path |
|
276 | 276 | return get_ipython_module_path(module_str) |
|
277 | 277 | |
|
278 | 278 | @undoc |
|
279 | 279 | def locate_profile(profile='default'): |
|
280 | 280 | warn("locate_profile has moved to the IPython.paths module since IPython 4.0.", stacklevel=2) |
|
281 | 281 | from IPython.paths import locate_profile |
|
282 | 282 | return locate_profile(profile=profile) |
|
283 | 283 | |
|
284 | 284 | def expand_path(s): |
|
285 | 285 | """Expand $VARS and ~names in a string, like a shell |
|
286 | 286 | |
|
287 | 287 | :Examples: |
|
288 | 288 | |
|
289 | 289 | In [2]: os.environ['FOO']='test' |
|
290 | 290 | |
|
291 | 291 | In [3]: expand_path('variable FOO is $FOO') |
|
292 | 292 | Out[3]: 'variable FOO is test' |
|
293 | 293 | """ |
|
294 | 294 | # This is a pretty subtle hack. When expand user is given a UNC path |
|
295 | 295 | # on Windows (\\server\share$\%username%), os.path.expandvars, removes |
|
296 | 296 | # the $ to get (\\server\share\%username%). I think it considered $ |
|
297 | 297 | # alone an empty var. But, we need the $ to remains there (it indicates |
|
298 | 298 | # a hidden share). |
|
299 | 299 | if os.name=='nt': |
|
300 | 300 | s = s.replace('$\\', 'IPYTHON_TEMP') |
|
301 | 301 | s = os.path.expandvars(os.path.expanduser(s)) |
|
302 | 302 | if os.name=='nt': |
|
303 | 303 | s = s.replace('IPYTHON_TEMP', '$\\') |
|
304 | 304 | return s |
|
305 | 305 | |
|
306 | 306 | |
|
307 | 307 | def unescape_glob(string): |
|
308 | 308 | """Unescape glob pattern in `string`.""" |
|
309 | 309 | def unescape(s): |
|
310 | 310 | for pattern in '*[]!?': |
|
311 | 311 | s = s.replace(r'\{0}'.format(pattern), pattern) |
|
312 | 312 | return s |
|
313 | 313 | return '\\'.join(map(unescape, string.split('\\\\'))) |
|
314 | 314 | |
|
315 | 315 | |
|
316 | 316 | def shellglob(args): |
|
317 | 317 | """ |
|
318 | 318 | Do glob expansion for each element in `args` and return a flattened list. |
|
319 | 319 | |
|
320 | 320 | Unmatched glob pattern will remain as-is in the returned list. |
|
321 | 321 | |
|
322 | 322 | """ |
|
323 | 323 | expanded = [] |
|
324 | 324 | # Do not unescape backslash in Windows as it is interpreted as |
|
325 | 325 | # path separator: |
|
326 | 326 | unescape = unescape_glob if sys.platform != 'win32' else lambda x: x |
|
327 | 327 | for a in args: |
|
328 | 328 | expanded.extend(glob.glob(a) or [unescape(a)]) |
|
329 | 329 | return expanded |
|
330 | 330 | |
|
331 | 331 | |
|
332 | 332 | def target_outdated(target,deps): |
|
333 | 333 | """Determine whether a target is out of date. |
|
334 | 334 | |
|
335 | 335 | target_outdated(target,deps) -> 1/0 |
|
336 | 336 | |
|
337 | 337 | deps: list of filenames which MUST exist. |
|
338 | 338 | target: single filename which may or may not exist. |
|
339 | 339 | |
|
340 | 340 | If target doesn't exist or is older than any file listed in deps, return |
|
341 | 341 | true, otherwise return false. |
|
342 | 342 | """ |
|
343 | 343 | try: |
|
344 | 344 | target_time = os.path.getmtime(target) |
|
345 | 345 | except os.error: |
|
346 | 346 | return 1 |
|
347 | 347 | for dep in deps: |
|
348 | 348 | dep_time = os.path.getmtime(dep) |
|
349 | 349 | if dep_time > target_time: |
|
350 | 350 | #print "For target",target,"Dep failed:",dep # dbg |
|
351 | 351 | #print "times (dep,tar):",dep_time,target_time # dbg |
|
352 | 352 | return 1 |
|
353 | 353 | return 0 |
|
354 | 354 | |
|
355 | 355 | |
|
356 | 356 | def target_update(target,deps,cmd): |
|
357 | 357 | """Update a target with a given command given a list of dependencies. |
|
358 | 358 | |
|
359 | 359 | target_update(target,deps,cmd) -> runs cmd if target is outdated. |
|
360 | 360 | |
|
361 | 361 | This is just a wrapper around target_outdated() which calls the given |
|
362 | 362 | command if target is outdated.""" |
|
363 | 363 | |
|
364 | 364 | if target_outdated(target,deps): |
|
365 | 365 | system(cmd) |
|
366 | 366 | |
|
367 | 367 | |
|
368 | 368 | ENOLINK = 1998 |
|
369 | 369 | |
|
370 | 370 | def link(src, dst): |
|
371 | 371 | """Hard links ``src`` to ``dst``, returning 0 or errno. |
|
372 | 372 | |
|
373 | 373 | Note that the special errno ``ENOLINK`` will be returned if ``os.link`` isn't |
|
374 | 374 | supported by the operating system. |
|
375 | 375 | """ |
|
376 | 376 | |
|
377 | 377 | if not hasattr(os, "link"): |
|
378 | 378 | return ENOLINK |
|
379 | 379 | link_errno = 0 |
|
380 | 380 | try: |
|
381 | 381 | os.link(src, dst) |
|
382 | 382 | except OSError as e: |
|
383 | 383 | link_errno = e.errno |
|
384 | 384 | return link_errno |
|
385 | 385 | |
|
386 | 386 | |
|
387 | 387 | def link_or_copy(src, dst): |
|
388 | 388 | """Attempts to hardlink ``src`` to ``dst``, copying if the link fails. |
|
389 | 389 | |
|
390 | 390 | Attempts to maintain the semantics of ``shutil.copy``. |
|
391 | 391 | |
|
392 | 392 | Because ``os.link`` does not overwrite files, a unique temporary file |
|
393 | 393 | will be used if the target already exists, then that file will be moved |
|
394 | 394 | into place. |
|
395 | 395 | """ |
|
396 | 396 | |
|
397 | 397 | if os.path.isdir(dst): |
|
398 | 398 | dst = os.path.join(dst, os.path.basename(src)) |
|
399 | 399 | |
|
400 | 400 | link_errno = link(src, dst) |
|
401 | 401 | if link_errno == errno.EEXIST: |
|
402 | 402 | if os.stat(src).st_ino == os.stat(dst).st_ino: |
|
403 | 403 | # dst is already a hard link to the correct file, so we don't need |
|
404 | 404 | # to do anything else. If we try to link and rename the file |
|
405 | 405 | # anyway, we get duplicate files - see http://bugs.python.org/issue21876 |
|
406 | 406 | return |
|
407 | 407 | |
|
408 | 408 | new_dst = dst + "-temp-%04X" %(random.randint(1, 16**4), ) |
|
409 | 409 | try: |
|
410 | 410 | link_or_copy(src, new_dst) |
|
411 | 411 | except: |
|
412 | 412 | try: |
|
413 | 413 | os.remove(new_dst) |
|
414 | 414 | except OSError: |
|
415 | 415 | pass |
|
416 | 416 | raise |
|
417 | 417 | os.rename(new_dst, dst) |
|
418 | 418 | elif link_errno != 0: |
|
419 | 419 | # Either link isn't supported, or the filesystem doesn't support |
|
420 | 420 | # linking, or 'src' and 'dst' are on different filesystems. |
|
421 | 421 | shutil.copy(src, dst) |
|
422 | 422 | |
|
423 | 423 | def ensure_dir_exists(path, mode=0o755): |
|
424 | 424 | """ensure that a directory exists |
|
425 | 425 | |
|
426 | 426 | If it doesn't exist, try to create it and protect against a race condition |
|
427 | 427 | if another process is doing the same. |
|
428 | 428 | |
|
429 | 429 | The default permissions are 755, which differ from os.makedirs default of 777. |
|
430 | 430 | """ |
|
431 | 431 | if not os.path.exists(path): |
|
432 | 432 | try: |
|
433 | 433 | os.makedirs(path, mode=mode) |
|
434 | 434 | except OSError as e: |
|
435 | 435 | if e.errno != errno.EEXIST: |
|
436 | 436 | raise |
|
437 | 437 | elif not os.path.isdir(path): |
|
438 | 438 | raise IOError("%r exists but is not a directory" % path) |
@@ -1,772 +1,772 b'' | |||
|
1 | 1 | # encoding: utf-8 |
|
2 | 2 | """ |
|
3 | 3 | Utilities for working with strings and text. |
|
4 | 4 | |
|
5 | 5 | Inheritance diagram: |
|
6 | 6 | |
|
7 | 7 | .. inheritance-diagram:: IPython.utils.text |
|
8 | 8 | :parts: 3 |
|
9 | 9 | """ |
|
10 | 10 | |
|
11 | 11 | import os |
|
12 | 12 | import re |
|
13 | 13 | import sys |
|
14 | 14 | import textwrap |
|
15 | 15 | from string import Formatter |
|
16 | 16 | from pathlib import Path |
|
17 | 17 | |
|
18 | 18 | from IPython.utils import py3compat |
|
19 | 19 | |
|
20 | 20 | # datetime.strftime date format for ipython |
|
21 | 21 | if sys.platform == 'win32': |
|
22 | 22 | date_format = "%B %d, %Y" |
|
23 | 23 | else: |
|
24 | 24 | date_format = "%B %-d, %Y" |
|
25 | 25 | |
|
26 | 26 | class LSString(str): |
|
27 | 27 | """String derivative with a special access attributes. |
|
28 | 28 | |
|
29 | 29 | These are normal strings, but with the special attributes: |
|
30 | 30 | |
|
31 | 31 | .l (or .list) : value as list (split on newlines). |
|
32 | 32 | .n (or .nlstr): original value (the string itself). |
|
33 | 33 | .s (or .spstr): value as whitespace-separated string. |
|
34 | 34 | .p (or .paths): list of path objects (requires path.py package) |
|
35 | 35 | |
|
36 | 36 | Any values which require transformations are computed only once and |
|
37 | 37 | cached. |
|
38 | 38 | |
|
39 | 39 | Such strings are very useful to efficiently interact with the shell, which |
|
40 | 40 | typically only understands whitespace-separated options for commands.""" |
|
41 | 41 | |
|
42 | 42 | def get_list(self): |
|
43 | 43 | try: |
|
44 | 44 | return self.__list |
|
45 | 45 | except AttributeError: |
|
46 | 46 | self.__list = self.split('\n') |
|
47 | 47 | return self.__list |
|
48 | 48 | |
|
49 | 49 | l = list = property(get_list) |
|
50 | 50 | |
|
51 | 51 | def get_spstr(self): |
|
52 | 52 | try: |
|
53 | 53 | return self.__spstr |
|
54 | 54 | except AttributeError: |
|
55 | 55 | self.__spstr = self.replace('\n',' ') |
|
56 | 56 | return self.__spstr |
|
57 | 57 | |
|
58 | 58 | s = spstr = property(get_spstr) |
|
59 | 59 | |
|
60 | 60 | def get_nlstr(self): |
|
61 | 61 | return self |
|
62 | 62 | |
|
63 | 63 | n = nlstr = property(get_nlstr) |
|
64 | 64 | |
|
65 | 65 | def get_paths(self): |
|
66 | 66 | try: |
|
67 | 67 | return self.__paths |
|
68 | 68 | except AttributeError: |
|
69 | 69 | self.__paths = [Path(p) for p in self.split('\n') if os.path.exists(p)] |
|
70 | 70 | return self.__paths |
|
71 | 71 | |
|
72 | 72 | p = paths = property(get_paths) |
|
73 | 73 | |
|
74 | 74 | # FIXME: We need to reimplement type specific displayhook and then add this |
|
75 | 75 | # back as a custom printer. This should also be moved outside utils into the |
|
76 | 76 | # core. |
|
77 | 77 | |
|
78 | 78 | # def print_lsstring(arg): |
|
79 | 79 | # """ Prettier (non-repr-like) and more informative printer for LSString """ |
|
80 | 80 | # print "LSString (.p, .n, .l, .s available). Value:" |
|
81 | 81 | # print arg |
|
82 | 82 | # |
|
83 | 83 | # |
|
84 | 84 | # print_lsstring = result_display.when_type(LSString)(print_lsstring) |
|
85 | 85 | |
|
86 | 86 | |
|
87 | 87 | class SList(list): |
|
88 | 88 | """List derivative with a special access attributes. |
|
89 | 89 | |
|
90 | 90 | These are normal lists, but with the special attributes: |
|
91 | 91 | |
|
92 | 92 | * .l (or .list) : value as list (the list itself). |
|
93 | 93 | * .n (or .nlstr): value as a string, joined on newlines. |
|
94 | 94 | * .s (or .spstr): value as a string, joined on spaces. |
|
95 | 95 | * .p (or .paths): list of path objects (requires path.py package) |
|
96 | 96 | |
|
97 | 97 | Any values which require transformations are computed only once and |
|
98 | 98 | cached.""" |
|
99 | 99 | |
|
100 | 100 | def get_list(self): |
|
101 | 101 | return self |
|
102 | 102 | |
|
103 | 103 | l = list = property(get_list) |
|
104 | 104 | |
|
105 | 105 | def get_spstr(self): |
|
106 | 106 | try: |
|
107 | 107 | return self.__spstr |
|
108 | 108 | except AttributeError: |
|
109 | 109 | self.__spstr = ' '.join(self) |
|
110 | 110 | return self.__spstr |
|
111 | 111 | |
|
112 | 112 | s = spstr = property(get_spstr) |
|
113 | 113 | |
|
114 | 114 | def get_nlstr(self): |
|
115 | 115 | try: |
|
116 | 116 | return self.__nlstr |
|
117 | 117 | except AttributeError: |
|
118 | 118 | self.__nlstr = '\n'.join(self) |
|
119 | 119 | return self.__nlstr |
|
120 | 120 | |
|
121 | 121 | n = nlstr = property(get_nlstr) |
|
122 | 122 | |
|
123 | 123 | def get_paths(self): |
|
124 | 124 | try: |
|
125 | 125 | return self.__paths |
|
126 | 126 | except AttributeError: |
|
127 | 127 | self.__paths = [Path(p) for p in self if os.path.exists(p)] |
|
128 | 128 | return self.__paths |
|
129 | 129 | |
|
130 | 130 | p = paths = property(get_paths) |
|
131 | 131 | |
|
132 | 132 | def grep(self, pattern, prune = False, field = None): |
|
133 | 133 | """ Return all strings matching 'pattern' (a regex or callable) |
|
134 | 134 | |
|
135 | 135 | This is case-insensitive. If prune is true, return all items |
|
136 | 136 | NOT matching the pattern. |
|
137 | 137 | |
|
138 | 138 | If field is specified, the match must occur in the specified |
|
139 | 139 | whitespace-separated field. |
|
140 | 140 | |
|
141 | 141 | Examples:: |
|
142 | 142 | |
|
143 | 143 | a.grep( lambda x: x.startswith('C') ) |
|
144 | 144 | a.grep('Cha.*log', prune=1) |
|
145 | 145 | a.grep('chm', field=-1) |
|
146 | 146 | """ |
|
147 | 147 | |
|
148 | 148 | def match_target(s): |
|
149 | 149 | if field is None: |
|
150 | 150 | return s |
|
151 | 151 | parts = s.split() |
|
152 | 152 | try: |
|
153 | 153 | tgt = parts[field] |
|
154 | 154 | return tgt |
|
155 | 155 | except IndexError: |
|
156 | 156 | return "" |
|
157 | 157 | |
|
158 | 158 | if isinstance(pattern, str): |
|
159 | 159 | pred = lambda x : re.search(pattern, x, re.IGNORECASE) |
|
160 | 160 | else: |
|
161 | 161 | pred = pattern |
|
162 | 162 | if not prune: |
|
163 | 163 | return SList([el for el in self if pred(match_target(el))]) |
|
164 | 164 | else: |
|
165 | 165 | return SList([el for el in self if not pred(match_target(el))]) |
|
166 | 166 | |
|
167 | 167 | def fields(self, *fields): |
|
168 | 168 | """ Collect whitespace-separated fields from string list |
|
169 | 169 | |
|
170 | 170 | Allows quick awk-like usage of string lists. |
|
171 | 171 | |
|
172 | 172 | Example data (in var a, created by 'a = !ls -l'):: |
|
173 | 173 | |
|
174 | 174 | -rwxrwxrwx 1 ville None 18 Dec 14 2006 ChangeLog |
|
175 | 175 | drwxrwxrwx+ 6 ville None 0 Oct 24 18:05 IPython |
|
176 | 176 | |
|
177 | 177 | * ``a.fields(0)`` is ``['-rwxrwxrwx', 'drwxrwxrwx+']`` |
|
178 | 178 | * ``a.fields(1,0)`` is ``['1 -rwxrwxrwx', '6 drwxrwxrwx+']`` |
|
179 | 179 | (note the joining by space). |
|
180 | 180 | * ``a.fields(-1)`` is ``['ChangeLog', 'IPython']`` |
|
181 | 181 | |
|
182 | 182 | IndexErrors are ignored. |
|
183 | 183 | |
|
184 | 184 | Without args, fields() just split()'s the strings. |
|
185 | 185 | """ |
|
186 | 186 | if len(fields) == 0: |
|
187 | 187 | return [el.split() for el in self] |
|
188 | 188 | |
|
189 | 189 | res = SList() |
|
190 | 190 | for el in [f.split() for f in self]: |
|
191 | 191 | lineparts = [] |
|
192 | 192 | |
|
193 | 193 | for fd in fields: |
|
194 | 194 | try: |
|
195 | 195 | lineparts.append(el[fd]) |
|
196 | 196 | except IndexError: |
|
197 | 197 | pass |
|
198 | 198 | if lineparts: |
|
199 | 199 | res.append(" ".join(lineparts)) |
|
200 | 200 | |
|
201 | 201 | return res |
|
202 | 202 | |
|
203 | 203 | def sort(self,field= None, nums = False): |
|
204 | 204 | """ sort by specified fields (see fields()) |
|
205 | 205 | |
|
206 | 206 | Example:: |
|
207 | 207 | |
|
208 | 208 | a.sort(1, nums = True) |
|
209 | 209 | |
|
210 | 210 | Sorts a by second field, in numerical order (so that 21 > 3) |
|
211 | 211 | |
|
212 | 212 | """ |
|
213 | 213 | |
|
214 | 214 | #decorate, sort, undecorate |
|
215 | 215 | if field is not None: |
|
216 | 216 | dsu = [[SList([line]).fields(field), line] for line in self] |
|
217 | 217 | else: |
|
218 | 218 | dsu = [[line, line] for line in self] |
|
219 | 219 | if nums: |
|
220 | 220 | for i in range(len(dsu)): |
|
221 | 221 | numstr = "".join([ch for ch in dsu[i][0] if ch.isdigit()]) |
|
222 | 222 | try: |
|
223 | 223 | n = int(numstr) |
|
224 | 224 | except ValueError: |
|
225 | 225 | n = 0 |
|
226 | 226 | dsu[i][0] = n |
|
227 | 227 | |
|
228 | 228 | |
|
229 | 229 | dsu.sort() |
|
230 | 230 | return SList([t[1] for t in dsu]) |
|
231 | 231 | |
|
232 | 232 | |
|
233 | 233 | # FIXME: We need to reimplement type specific displayhook and then add this |
|
234 | 234 | # back as a custom printer. This should also be moved outside utils into the |
|
235 | 235 | # core. |
|
236 | 236 | |
|
237 | 237 | # def print_slist(arg): |
|
238 | 238 | # """ Prettier (non-repr-like) and more informative printer for SList """ |
|
239 | 239 | # print "SList (.p, .n, .l, .s, .grep(), .fields(), sort() available):" |
|
240 | 240 | # if hasattr(arg, 'hideonce') and arg.hideonce: |
|
241 | 241 | # arg.hideonce = False |
|
242 | 242 | # return |
|
243 | 243 | # |
|
244 | 244 | # nlprint(arg) # This was a nested list printer, now removed. |
|
245 | 245 | # |
|
246 | 246 | # print_slist = result_display.when_type(SList)(print_slist) |
|
247 | 247 | |
|
248 | 248 | |
|
249 | 249 | def indent(instr,nspaces=4, ntabs=0, flatten=False): |
|
250 | 250 | """Indent a string a given number of spaces or tabstops. |
|
251 | 251 | |
|
252 | 252 | indent(str,nspaces=4,ntabs=0) -> indent str by ntabs+nspaces. |
|
253 | 253 | |
|
254 | 254 | Parameters |
|
255 | 255 | ---------- |
|
256 | 256 | |
|
257 | 257 | instr : basestring |
|
258 | 258 | The string to be indented. |
|
259 | 259 | nspaces : int (default: 4) |
|
260 | 260 | The number of spaces to be indented. |
|
261 | 261 | ntabs : int (default: 0) |
|
262 | 262 | The number of tabs to be indented. |
|
263 | 263 | flatten : bool (default: False) |
|
264 | 264 | Whether to scrub existing indentation. If True, all lines will be |
|
265 | 265 | aligned to the same indentation. If False, existing indentation will |
|
266 | 266 | be strictly increased. |
|
267 | 267 | |
|
268 | 268 | Returns |
|
269 | 269 | ------- |
|
270 | 270 | |
|
271 | 271 | str|unicode : string indented by ntabs and nspaces. |
|
272 | 272 | |
|
273 | 273 | """ |
|
274 | 274 | if instr is None: |
|
275 | 275 | return |
|
276 | 276 | ind = '\t'*ntabs+' '*nspaces |
|
277 | 277 | if flatten: |
|
278 | 278 | pat = re.compile(r'^\s*', re.MULTILINE) |
|
279 | 279 | else: |
|
280 | 280 | pat = re.compile(r'^', re.MULTILINE) |
|
281 | 281 | outstr = re.sub(pat, ind, instr) |
|
282 | 282 | if outstr.endswith(os.linesep+ind): |
|
283 | 283 | return outstr[:-len(ind)] |
|
284 | 284 | else: |
|
285 | 285 | return outstr |
|
286 | 286 | |
|
287 | 287 | |
|
288 | 288 | def list_strings(arg): |
|
289 | 289 | """Always return a list of strings, given a string or list of strings |
|
290 | 290 | as input. |
|
291 | 291 | |
|
292 | 292 | Examples |
|
293 | 293 | -------- |
|
294 | 294 | :: |
|
295 | 295 | |
|
296 | 296 | In [7]: list_strings('A single string') |
|
297 | 297 | Out[7]: ['A single string'] |
|
298 | 298 | |
|
299 | 299 | In [8]: list_strings(['A single string in a list']) |
|
300 | 300 | Out[8]: ['A single string in a list'] |
|
301 | 301 | |
|
302 | 302 | In [9]: list_strings(['A','list','of','strings']) |
|
303 | 303 | Out[9]: ['A', 'list', 'of', 'strings'] |
|
304 | 304 | """ |
|
305 | 305 | |
|
306 | 306 | if isinstance(arg, str): |
|
307 | 307 | return [arg] |
|
308 | 308 | else: |
|
309 | 309 | return arg |
|
310 | 310 | |
|
311 | 311 | |
|
312 | 312 | def marquee(txt='',width=78,mark='*'): |
|
313 | 313 | """Return the input string centered in a 'marquee'. |
|
314 | 314 | |
|
315 | 315 | Examples |
|
316 | 316 | -------- |
|
317 | 317 | :: |
|
318 | 318 | |
|
319 | 319 | In [16]: marquee('A test',40) |
|
320 | 320 | Out[16]: '**************** A test ****************' |
|
321 | 321 | |
|
322 | 322 | In [17]: marquee('A test',40,'-') |
|
323 | 323 | Out[17]: '---------------- A test ----------------' |
|
324 | 324 | |
|
325 | 325 | In [18]: marquee('A test',40,' ') |
|
326 | 326 | Out[18]: ' A test ' |
|
327 | 327 | |
|
328 | 328 | """ |
|
329 | 329 | if not txt: |
|
330 | 330 | return (mark*width)[:width] |
|
331 | 331 | nmark = (width-len(txt)-2)//len(mark)//2 |
|
332 | 332 | if nmark < 0: nmark =0 |
|
333 | 333 | marks = mark*nmark |
|
334 | 334 | return '%s %s %s' % (marks,txt,marks) |
|
335 | 335 | |
|
336 | 336 | |
|
337 | 337 | ini_spaces_re = re.compile(r'^(\s+)') |
|
338 | 338 | |
|
339 | 339 | def num_ini_spaces(strng): |
|
340 | 340 | """Return the number of initial spaces in a string""" |
|
341 | 341 | |
|
342 | 342 | ini_spaces = ini_spaces_re.match(strng) |
|
343 | 343 | if ini_spaces: |
|
344 | 344 | return ini_spaces.end() |
|
345 | 345 | else: |
|
346 | 346 | return 0 |
|
347 | 347 | |
|
348 | 348 | |
|
349 | 349 | def format_screen(strng): |
|
350 | 350 | """Format a string for screen printing. |
|
351 | 351 | |
|
352 | 352 | This removes some latex-type format codes.""" |
|
353 | 353 | # Paragraph continue |
|
354 | 354 | par_re = re.compile(r'\\$',re.MULTILINE) |
|
355 | 355 | strng = par_re.sub('',strng) |
|
356 | 356 | return strng |
|
357 | 357 | |
|
358 | 358 | |
|
359 | 359 | def dedent(text): |
|
360 | 360 | """Equivalent of textwrap.dedent that ignores unindented first line. |
|
361 | 361 | |
|
362 | 362 | This means it will still dedent strings like: |
|
363 | 363 | '''foo |
|
364 | 364 | is a bar |
|
365 | 365 | ''' |
|
366 | 366 | |
|
367 | 367 | For use in wrap_paragraphs. |
|
368 | 368 | """ |
|
369 | 369 | |
|
370 | 370 | if text.startswith('\n'): |
|
371 | 371 | # text starts with blank line, don't ignore the first line |
|
372 | 372 | return textwrap.dedent(text) |
|
373 | 373 | |
|
374 | 374 | # split first line |
|
375 | 375 | splits = text.split('\n',1) |
|
376 | 376 | if len(splits) == 1: |
|
377 | 377 | # only one line |
|
378 | 378 | return textwrap.dedent(text) |
|
379 | 379 | |
|
380 | 380 | first, rest = splits |
|
381 | 381 | # dedent everything but the first line |
|
382 | 382 | rest = textwrap.dedent(rest) |
|
383 | 383 | return '\n'.join([first, rest]) |
|
384 | 384 | |
|
385 | 385 | |
|
386 | 386 | def wrap_paragraphs(text, ncols=80): |
|
387 | 387 | """Wrap multiple paragraphs to fit a specified width. |
|
388 | 388 | |
|
389 | 389 | This is equivalent to textwrap.wrap, but with support for multiple |
|
390 | 390 | paragraphs, as separated by empty lines. |
|
391 | 391 | |
|
392 | 392 | Returns |
|
393 | 393 | ------- |
|
394 | 394 | |
|
395 | 395 | list of complete paragraphs, wrapped to fill `ncols` columns. |
|
396 | 396 | """ |
|
397 | 397 | paragraph_re = re.compile(r'\n(\s*\n)+', re.MULTILINE) |
|
398 | 398 | text = dedent(text).strip() |
|
399 | 399 | paragraphs = paragraph_re.split(text)[::2] # every other entry is space |
|
400 | 400 | out_ps = [] |
|
401 | 401 | indent_re = re.compile(r'\n\s+', re.MULTILINE) |
|
402 | 402 | for p in paragraphs: |
|
403 | 403 | # presume indentation that survives dedent is meaningful formatting, |
|
404 | 404 | # so don't fill unless text is flush. |
|
405 | 405 | if indent_re.search(p) is None: |
|
406 | 406 | # wrap paragraph |
|
407 | 407 | p = textwrap.fill(p, ncols) |
|
408 | 408 | out_ps.append(p) |
|
409 | 409 | return out_ps |
|
410 | 410 | |
|
411 | 411 | |
|
412 | 412 | def long_substr(data): |
|
413 | 413 | """Return the longest common substring in a list of strings. |
|
414 | 414 | |
|
415 | 415 | Credit: http://stackoverflow.com/questions/2892931/longest-common-substring-from-more-than-two-strings-python |
|
416 | 416 | """ |
|
417 | 417 | substr = '' |
|
418 | 418 | if len(data) > 1 and len(data[0]) > 0: |
|
419 | 419 | for i in range(len(data[0])): |
|
420 | 420 | for j in range(len(data[0])-i+1): |
|
421 | 421 | if j > len(substr) and all(data[0][i:i+j] in x for x in data): |
|
422 | 422 | substr = data[0][i:i+j] |
|
423 | 423 | elif len(data) == 1: |
|
424 | 424 | substr = data[0] |
|
425 | 425 | return substr |
|
426 | 426 | |
|
427 | 427 | |
|
428 | 428 | def strip_email_quotes(text): |
|
429 | 429 | """Strip leading email quotation characters ('>'). |
|
430 | 430 | |
|
431 | 431 | Removes any combination of leading '>' interspersed with whitespace that |
|
432 | 432 | appears *identically* in all lines of the input text. |
|
433 | 433 | |
|
434 | 434 | Parameters |
|
435 | 435 | ---------- |
|
436 | 436 | text : str |
|
437 | 437 | |
|
438 | 438 | Examples |
|
439 | 439 | -------- |
|
440 | 440 | |
|
441 | 441 | Simple uses:: |
|
442 | 442 | |
|
443 | 443 | In [2]: strip_email_quotes('> > text') |
|
444 | 444 | Out[2]: 'text' |
|
445 | 445 | |
|
446 | 446 | In [3]: strip_email_quotes('> > text\\n> > more') |
|
447 | 447 | Out[3]: 'text\\nmore' |
|
448 | 448 | |
|
449 | 449 | Note how only the common prefix that appears in all lines is stripped:: |
|
450 | 450 | |
|
451 | 451 | In [4]: strip_email_quotes('> > text\\n> > more\\n> more...') |
|
452 | 452 | Out[4]: '> text\\n> more\\nmore...' |
|
453 | 453 | |
|
454 | 454 | So if any line has no quote marks ('>') , then none are stripped from any |
|
455 | 455 | of them :: |
|
456 | 456 | |
|
457 | 457 | In [5]: strip_email_quotes('> > text\\n> > more\\nlast different') |
|
458 | 458 | Out[5]: '> > text\\n> > more\\nlast different' |
|
459 | 459 | """ |
|
460 | 460 | lines = text.splitlines() |
|
461 | 461 | matches = set() |
|
462 | 462 | for line in lines: |
|
463 | 463 | prefix = re.match(r'^(\s*>[ >]*)', line) |
|
464 | 464 | if prefix: |
|
465 | 465 | matches.add(prefix.group(1)) |
|
466 | 466 | else: |
|
467 | 467 | break |
|
468 | 468 | else: |
|
469 | 469 | prefix = long_substr(list(matches)) |
|
470 | 470 | if prefix: |
|
471 | 471 | strip = len(prefix) |
|
472 | 472 | text = '\n'.join([ ln[strip:] for ln in lines]) |
|
473 | 473 | return text |
|
474 | 474 | |
|
475 | 475 | def strip_ansi(source): |
|
476 | 476 | """ |
|
477 | 477 | Remove ansi escape codes from text. |
|
478 | 478 | |
|
479 | 479 | Parameters |
|
480 | 480 | ---------- |
|
481 | 481 | source : str |
|
482 | 482 | Source to remove the ansi from |
|
483 | 483 | """ |
|
484 | 484 | return re.sub(r'\033\[(\d|;)+?m', '', source) |
|
485 | 485 | |
|
486 | 486 | |
|
487 | 487 | class EvalFormatter(Formatter): |
|
488 | 488 | """A String Formatter that allows evaluation of simple expressions. |
|
489 | 489 | |
|
490 | 490 | Note that this version interprets a : as specifying a format string (as per |
|
491 | 491 | standard string formatting), so if slicing is required, you must explicitly |
|
492 | 492 | create a slice. |
|
493 | 493 | |
|
494 | 494 | This is to be used in templating cases, such as the parallel batch |
|
495 | 495 | script templates, where simple arithmetic on arguments is useful. |
|
496 | 496 | |
|
497 | 497 | Examples |
|
498 | 498 | -------- |
|
499 | 499 | :: |
|
500 | 500 | |
|
501 | 501 | In [1]: f = EvalFormatter() |
|
502 | 502 | In [2]: f.format('{n//4}', n=8) |
|
503 | 503 | Out[2]: '2' |
|
504 | 504 | |
|
505 | 505 | In [3]: f.format("{greeting[slice(2,4)]}", greeting="Hello") |
|
506 | 506 | Out[3]: 'll' |
|
507 | 507 | """ |
|
508 | 508 | def get_field(self, name, args, kwargs): |
|
509 | 509 | v = eval(name, kwargs) |
|
510 | 510 | return v, name |
|
511 | 511 | |
|
512 | 512 | #XXX: As of Python 3.4, the format string parsing no longer splits on a colon |
|
513 | 513 | # inside [], so EvalFormatter can handle slicing. Once we only support 3.4 and |
|
514 | 514 | # above, it should be possible to remove FullEvalFormatter. |
|
515 | 515 | |
|
516 | 516 | class FullEvalFormatter(Formatter): |
|
517 | 517 | """A String Formatter that allows evaluation of simple expressions. |
|
518 | 518 | |
|
519 | 519 | Any time a format key is not found in the kwargs, |
|
520 | 520 | it will be tried as an expression in the kwargs namespace. |
|
521 | 521 | |
|
522 | 522 | Note that this version allows slicing using [1:2], so you cannot specify |
|
523 | 523 | a format string. Use :class:`EvalFormatter` to permit format strings. |
|
524 | 524 | |
|
525 | 525 | Examples |
|
526 | 526 | -------- |
|
527 | 527 | :: |
|
528 | 528 | |
|
529 | 529 | In [1]: f = FullEvalFormatter() |
|
530 | 530 | In [2]: f.format('{n//4}', n=8) |
|
531 | 531 | Out[2]: '2' |
|
532 | 532 | |
|
533 | 533 | In [3]: f.format('{list(range(5))[2:4]}') |
|
534 | 534 | Out[3]: '[2, 3]' |
|
535 | 535 | |
|
536 | 536 | In [4]: f.format('{3*2}') |
|
537 | 537 | Out[4]: '6' |
|
538 | 538 | """ |
|
539 | 539 | # copied from Formatter._vformat with minor changes to allow eval |
|
540 | 540 | # and replace the format_spec code with slicing |
|
541 | 541 | def vformat(self, format_string, args, kwargs): |
|
542 | 542 | result = [] |
|
543 | 543 | for literal_text, field_name, format_spec, conversion in \ |
|
544 | 544 | self.parse(format_string): |
|
545 | 545 | |
|
546 | 546 | # output the literal text |
|
547 | 547 | if literal_text: |
|
548 | 548 | result.append(literal_text) |
|
549 | 549 | |
|
550 | 550 | # if there's a field, output it |
|
551 | 551 | if field_name is not None: |
|
552 | 552 | # this is some markup, find the object and do |
|
553 | 553 | # the formatting |
|
554 | 554 | |
|
555 | 555 | if format_spec: |
|
556 | 556 | # override format spec, to allow slicing: |
|
557 | 557 | field_name = ':'.join([field_name, format_spec]) |
|
558 | 558 | |
|
559 | 559 | # eval the contents of the field for the object |
|
560 | 560 | # to be formatted |
|
561 | 561 | obj = eval(field_name, kwargs) |
|
562 | 562 | |
|
563 | 563 | # do any conversion on the resulting object |
|
564 | 564 | obj = self.convert_field(obj, conversion) |
|
565 | 565 | |
|
566 | 566 | # format the object and append to the result |
|
567 | 567 | result.append(self.format_field(obj, '')) |
|
568 | 568 | |
|
569 | 569 | return ''.join(py3compat.cast_unicode(s) for s in result) |
|
570 | 570 | |
|
571 | 571 | |
|
572 | 572 | class DollarFormatter(FullEvalFormatter): |
|
573 | 573 | """Formatter allowing Itpl style $foo replacement, for names and attribute |
|
574 | 574 | access only. Standard {foo} replacement also works, and allows full |
|
575 | 575 | evaluation of its arguments. |
|
576 | 576 | |
|
577 | 577 | Examples |
|
578 | 578 | -------- |
|
579 | 579 | :: |
|
580 | 580 | |
|
581 | 581 | In [1]: f = DollarFormatter() |
|
582 | 582 | In [2]: f.format('{n//4}', n=8) |
|
583 | 583 | Out[2]: '2' |
|
584 | 584 | |
|
585 | 585 | In [3]: f.format('23 * 76 is $result', result=23*76) |
|
586 | 586 | Out[3]: '23 * 76 is 1748' |
|
587 | 587 | |
|
588 | 588 | In [4]: f.format('$a or {b}', a=1, b=2) |
|
589 | 589 | Out[4]: '1 or 2' |
|
590 | 590 | """ |
|
591 | _dollar_pattern_ignore_single_quote = re.compile("(.*?)\$(\$?[\w\.]+)(?=([^']*'[^']*')*[^']*$)") | |
|
591 | _dollar_pattern_ignore_single_quote = re.compile(r"(.*?)\$(\$?[\w\.]+)(?=([^']*'[^']*')*[^']*$)") | |
|
592 | 592 | def parse(self, fmt_string): |
|
593 | 593 | for literal_txt, field_name, format_spec, conversion \ |
|
594 | 594 | in Formatter.parse(self, fmt_string): |
|
595 | 595 | |
|
596 | 596 | # Find $foo patterns in the literal text. |
|
597 | 597 | continue_from = 0 |
|
598 | 598 | txt = "" |
|
599 | 599 | for m in self._dollar_pattern_ignore_single_quote.finditer(literal_txt): |
|
600 | 600 | new_txt, new_field = m.group(1,2) |
|
601 | 601 | # $$foo --> $foo |
|
602 | 602 | if new_field.startswith("$"): |
|
603 | 603 | txt += new_txt + new_field |
|
604 | 604 | else: |
|
605 | 605 | yield (txt + new_txt, new_field, "", None) |
|
606 | 606 | txt = "" |
|
607 | 607 | continue_from = m.end() |
|
608 | 608 | |
|
609 | 609 | # Re-yield the {foo} style pattern |
|
610 | 610 | yield (txt + literal_txt[continue_from:], field_name, format_spec, conversion) |
|
611 | 611 | |
|
612 | 612 | #----------------------------------------------------------------------------- |
|
613 | 613 | # Utils to columnize a list of string |
|
614 | 614 | #----------------------------------------------------------------------------- |
|
615 | 615 | |
|
616 | 616 | def _col_chunks(l, max_rows, row_first=False): |
|
617 | 617 | """Yield successive max_rows-sized column chunks from l.""" |
|
618 | 618 | if row_first: |
|
619 | 619 | ncols = (len(l) // max_rows) + (len(l) % max_rows > 0) |
|
620 | 620 | for i in range(ncols): |
|
621 | 621 | yield [l[j] for j in range(i, len(l), ncols)] |
|
622 | 622 | else: |
|
623 | 623 | for i in range(0, len(l), max_rows): |
|
624 | 624 | yield l[i:(i + max_rows)] |
|
625 | 625 | |
|
626 | 626 | |
|
627 | 627 | def _find_optimal(rlist, row_first=False, separator_size=2, displaywidth=80): |
|
628 | 628 | """Calculate optimal info to columnize a list of string""" |
|
629 | 629 | for max_rows in range(1, len(rlist) + 1): |
|
630 | 630 | col_widths = list(map(max, _col_chunks(rlist, max_rows, row_first))) |
|
631 | 631 | sumlength = sum(col_widths) |
|
632 | 632 | ncols = len(col_widths) |
|
633 | 633 | if sumlength + separator_size * (ncols - 1) <= displaywidth: |
|
634 | 634 | break |
|
635 | 635 | return {'num_columns': ncols, |
|
636 | 636 | 'optimal_separator_width': (displaywidth - sumlength) // (ncols - 1) if (ncols - 1) else 0, |
|
637 | 637 | 'max_rows': max_rows, |
|
638 | 638 | 'column_widths': col_widths |
|
639 | 639 | } |
|
640 | 640 | |
|
641 | 641 | |
|
642 | 642 | def _get_or_default(mylist, i, default=None): |
|
643 | 643 | """return list item number, or default if don't exist""" |
|
644 | 644 | if i >= len(mylist): |
|
645 | 645 | return default |
|
646 | 646 | else : |
|
647 | 647 | return mylist[i] |
|
648 | 648 | |
|
649 | 649 | |
|
650 | 650 | def compute_item_matrix(items, row_first=False, empty=None, *args, **kwargs) : |
|
651 | 651 | """Returns a nested list, and info to columnize items |
|
652 | 652 | |
|
653 | 653 | Parameters |
|
654 | 654 | ---------- |
|
655 | 655 | |
|
656 | 656 | items |
|
657 | 657 | list of strings to columize |
|
658 | 658 | row_first : (default False) |
|
659 | 659 | Whether to compute columns for a row-first matrix instead of |
|
660 | 660 | column-first (default). |
|
661 | 661 | empty : (default None) |
|
662 | 662 | default value to fill list if needed |
|
663 | 663 | separator_size : int (default=2) |
|
664 | 664 | How much characters will be used as a separation between each columns. |
|
665 | 665 | displaywidth : int (default=80) |
|
666 | 666 | The width of the area onto which the columns should enter |
|
667 | 667 | |
|
668 | 668 | Returns |
|
669 | 669 | ------- |
|
670 | 670 | |
|
671 | 671 | strings_matrix |
|
672 | 672 | |
|
673 | 673 | nested list of string, the outer most list contains as many list as |
|
674 | 674 | rows, the innermost lists have each as many element as columns. If the |
|
675 | 675 | total number of elements in `items` does not equal the product of |
|
676 | 676 | rows*columns, the last element of some lists are filled with `None`. |
|
677 | 677 | |
|
678 | 678 | dict_info |
|
679 | 679 | some info to make columnize easier: |
|
680 | 680 | |
|
681 | 681 | num_columns |
|
682 | 682 | number of columns |
|
683 | 683 | max_rows |
|
684 | 684 | maximum number of rows (final number may be less) |
|
685 | 685 | column_widths |
|
686 | 686 | list of with of each columns |
|
687 | 687 | optimal_separator_width |
|
688 | 688 | best separator width between columns |
|
689 | 689 | |
|
690 | 690 | Examples |
|
691 | 691 | -------- |
|
692 | 692 | :: |
|
693 | 693 | |
|
694 | 694 | In [1]: l = ['aaa','b','cc','d','eeeee','f','g','h','i','j','k','l'] |
|
695 | 695 | In [2]: list, info = compute_item_matrix(l, displaywidth=12) |
|
696 | 696 | In [3]: list |
|
697 | 697 | Out[3]: [['aaa', 'f', 'k'], ['b', 'g', 'l'], ['cc', 'h', None], ['d', 'i', None], ['eeeee', 'j', None]] |
|
698 | 698 | In [4]: ideal = {'num_columns': 3, 'column_widths': [5, 1, 1], 'optimal_separator_width': 2, 'max_rows': 5} |
|
699 | 699 | In [5]: all((info[k] == ideal[k] for k in ideal.keys())) |
|
700 | 700 | Out[5]: True |
|
701 | 701 | """ |
|
702 | 702 | info = _find_optimal(list(map(len, items)), row_first, *args, **kwargs) |
|
703 | 703 | nrow, ncol = info['max_rows'], info['num_columns'] |
|
704 | 704 | if row_first: |
|
705 | 705 | return ([[_get_or_default(items, r * ncol + c, default=empty) for c in range(ncol)] for r in range(nrow)], info) |
|
706 | 706 | else: |
|
707 | 707 | return ([[_get_or_default(items, c * nrow + r, default=empty) for c in range(ncol)] for r in range(nrow)], info) |
|
708 | 708 | |
|
709 | 709 | |
|
710 | 710 | def columnize(items, row_first=False, separator=' ', displaywidth=80, spread=False): |
|
711 | 711 | """ Transform a list of strings into a single string with columns. |
|
712 | 712 | |
|
713 | 713 | Parameters |
|
714 | 714 | ---------- |
|
715 | 715 | items : sequence of strings |
|
716 | 716 | The strings to process. |
|
717 | 717 | |
|
718 | 718 | row_first : (default False) |
|
719 | 719 | Whether to compute columns for a row-first matrix instead of |
|
720 | 720 | column-first (default). |
|
721 | 721 | |
|
722 | 722 | separator : str, optional [default is two spaces] |
|
723 | 723 | The string that separates columns. |
|
724 | 724 | |
|
725 | 725 | displaywidth : int, optional [default is 80] |
|
726 | 726 | Width of the display in number of characters. |
|
727 | 727 | |
|
728 | 728 | Returns |
|
729 | 729 | ------- |
|
730 | 730 | The formatted string. |
|
731 | 731 | """ |
|
732 | 732 | if not items: |
|
733 | 733 | return '\n' |
|
734 | 734 | matrix, info = compute_item_matrix(items, row_first=row_first, separator_size=len(separator), displaywidth=displaywidth) |
|
735 | 735 | if spread: |
|
736 | 736 | separator = separator.ljust(int(info['optimal_separator_width'])) |
|
737 | 737 | fmatrix = [filter(None, x) for x in matrix] |
|
738 | 738 | sjoin = lambda x : separator.join([ y.ljust(w, ' ') for y, w in zip(x, info['column_widths'])]) |
|
739 | 739 | return '\n'.join(map(sjoin, fmatrix))+'\n' |
|
740 | 740 | |
|
741 | 741 | |
|
742 | 742 | def get_text_list(list_, last_sep=' and ', sep=", ", wrap_item_with=""): |
|
743 | 743 | """ |
|
744 | 744 | Return a string with a natural enumeration of items |
|
745 | 745 | |
|
746 | 746 | >>> get_text_list(['a', 'b', 'c', 'd']) |
|
747 | 747 | 'a, b, c and d' |
|
748 | 748 | >>> get_text_list(['a', 'b', 'c'], ' or ') |
|
749 | 749 | 'a, b or c' |
|
750 | 750 | >>> get_text_list(['a', 'b', 'c'], ', ') |
|
751 | 751 | 'a, b, c' |
|
752 | 752 | >>> get_text_list(['a', 'b'], ' or ') |
|
753 | 753 | 'a or b' |
|
754 | 754 | >>> get_text_list(['a']) |
|
755 | 755 | 'a' |
|
756 | 756 | >>> get_text_list([]) |
|
757 | 757 | '' |
|
758 | 758 | >>> get_text_list(['a', 'b'], wrap_item_with="`") |
|
759 | 759 | '`a` and `b`' |
|
760 | 760 | >>> get_text_list(['a', 'b', 'c', 'd'], " = ", sep=" + ") |
|
761 | 761 | 'a + b + c = d' |
|
762 | 762 | """ |
|
763 | 763 | if len(list_) == 0: |
|
764 | 764 | return '' |
|
765 | 765 | if wrap_item_with: |
|
766 | 766 | list_ = ['%s%s%s' % (wrap_item_with, item, wrap_item_with) for |
|
767 | 767 | item in list_] |
|
768 | 768 | if len(list_) == 1: |
|
769 | 769 | return list_[0] |
|
770 | 770 | return '%s%s%s' % ( |
|
771 | 771 | sep.join(i for i in list_[:-1]), |
|
772 | 772 | last_sep, list_[-1]) |
@@ -1,590 +1,590 b'' | |||
|
1 | 1 | """Patched version of standard library tokenize, to deal with various bugs. |
|
2 | 2 | |
|
3 | 3 | Based on Python 3.2 code. |
|
4 | 4 | |
|
5 | 5 | Patches: |
|
6 | 6 | |
|
7 | 7 | - Gareth Rees' patch for Python issue #12691 (untokenizing) |
|
8 | 8 | - Except we don't encode the output of untokenize |
|
9 | 9 | - Python 2 compatible syntax, so that it can be byte-compiled at installation |
|
10 | 10 | - Newlines in comments and blank lines should be either NL or NEWLINE, depending |
|
11 | 11 | on whether they are in a multi-line statement. Filed as Python issue #17061. |
|
12 | 12 | - Export generate_tokens & TokenError |
|
13 | 13 | - u and rb literals are allowed under Python 3.3 and above. |
|
14 | 14 | |
|
15 | 15 | ------------------------------------------------------------------------------ |
|
16 | 16 | |
|
17 | 17 | Tokenization help for Python programs. |
|
18 | 18 | |
|
19 | 19 | tokenize(readline) is a generator that breaks a stream of bytes into |
|
20 | 20 | Python tokens. It decodes the bytes according to PEP-0263 for |
|
21 | 21 | determining source file encoding. |
|
22 | 22 | |
|
23 | 23 | It accepts a readline-like method which is called repeatedly to get the |
|
24 | 24 | next line of input (or b"" for EOF). It generates 5-tuples with these |
|
25 | 25 | members: |
|
26 | 26 | |
|
27 | 27 | the token type (see token.py) |
|
28 | 28 | the token (a string) |
|
29 | 29 | the starting (row, column) indices of the token (a 2-tuple of ints) |
|
30 | 30 | the ending (row, column) indices of the token (a 2-tuple of ints) |
|
31 | 31 | the original line (string) |
|
32 | 32 | |
|
33 | 33 | It is designed to match the working of the Python tokenizer exactly, except |
|
34 | 34 | that it produces COMMENT tokens for comments and gives type OP for all |
|
35 | 35 | operators. Additionally, all token lists start with an ENCODING token |
|
36 | 36 | which tells you which encoding was used to decode the bytes stream. |
|
37 | 37 | """ |
|
38 | 38 | |
|
39 | 39 | __author__ = 'Ka-Ping Yee <ping@lfw.org>' |
|
40 | 40 | __credits__ = ('GvR, ESR, Tim Peters, Thomas Wouters, Fred Drake, ' |
|
41 | 41 | 'Skip Montanaro, Raymond Hettinger, Trent Nelson, ' |
|
42 | 42 | 'Michael Foord') |
|
43 | 43 | import builtins |
|
44 | 44 | import re |
|
45 | 45 | import sys |
|
46 | 46 | from token import * |
|
47 | 47 | from codecs import lookup, BOM_UTF8 |
|
48 | 48 | import collections |
|
49 | 49 | from io import TextIOWrapper |
|
50 | cookie_re = re.compile("coding[:=]\s*([-\w.]+)") | |
|
50 | cookie_re = re.compile(r"coding[:=]\s*([-\w.]+)") | |
|
51 | 51 | |
|
52 | 52 | import token |
|
53 | 53 | __all__ = token.__all__ + ["COMMENT", "tokenize", "detect_encoding", |
|
54 | 54 | "NL", "untokenize", "ENCODING", "TokenInfo"] |
|
55 | 55 | del token |
|
56 | 56 | |
|
57 | 57 | __all__ += ["generate_tokens", "TokenError"] |
|
58 | 58 | |
|
59 | 59 | COMMENT = N_TOKENS |
|
60 | 60 | tok_name[COMMENT] = 'COMMENT' |
|
61 | 61 | NL = N_TOKENS + 1 |
|
62 | 62 | tok_name[NL] = 'NL' |
|
63 | 63 | ENCODING = N_TOKENS + 2 |
|
64 | 64 | tok_name[ENCODING] = 'ENCODING' |
|
65 | 65 | N_TOKENS += 3 |
|
66 | 66 | |
|
67 | 67 | class TokenInfo(collections.namedtuple('TokenInfo', 'type string start end line')): |
|
68 | 68 | def __repr__(self): |
|
69 | 69 | annotated_type = '%d (%s)' % (self.type, tok_name[self.type]) |
|
70 | 70 | return ('TokenInfo(type=%s, string=%r, start=%r, end=%r, line=%r)' % |
|
71 | 71 | self._replace(type=annotated_type)) |
|
72 | 72 | |
|
73 | 73 | def group(*choices): return '(' + '|'.join(choices) + ')' |
|
74 | 74 | def any(*choices): return group(*choices) + '*' |
|
75 | 75 | def maybe(*choices): return group(*choices) + '?' |
|
76 | 76 | |
|
77 | 77 | # Note: we use unicode matching for names ("\w") but ascii matching for |
|
78 | 78 | # number literals. |
|
79 | 79 | Whitespace = r'[ \f\t]*' |
|
80 | 80 | Comment = r'#[^\r\n]*' |
|
81 | 81 | Ignore = Whitespace + any(r'\\\r?\n' + Whitespace) + maybe(Comment) |
|
82 | 82 | Name = r'\w+' |
|
83 | 83 | |
|
84 | 84 | Hexnumber = r'0[xX][0-9a-fA-F]+' |
|
85 | 85 | Binnumber = r'0[bB][01]+' |
|
86 | 86 | Octnumber = r'0[oO][0-7]+' |
|
87 | 87 | Decnumber = r'(?:0+|[1-9][0-9]*)' |
|
88 | 88 | Intnumber = group(Hexnumber, Binnumber, Octnumber, Decnumber) |
|
89 | 89 | Exponent = r'[eE][-+]?[0-9]+' |
|
90 | 90 | Pointfloat = group(r'[0-9]+\.[0-9]*', r'\.[0-9]+') + maybe(Exponent) |
|
91 | 91 | Expfloat = r'[0-9]+' + Exponent |
|
92 | 92 | Floatnumber = group(Pointfloat, Expfloat) |
|
93 | 93 | Imagnumber = group(r'[0-9]+[jJ]', Floatnumber + r'[jJ]') |
|
94 | 94 | Number = group(Imagnumber, Floatnumber, Intnumber) |
|
95 | 95 | StringPrefix = r'(?:[bB][rR]?|[rR][bB]?|[uU])?' |
|
96 | 96 | |
|
97 | 97 | # Tail end of ' string. |
|
98 | 98 | Single = r"[^'\\]*(?:\\.[^'\\]*)*'" |
|
99 | 99 | # Tail end of " string. |
|
100 | 100 | Double = r'[^"\\]*(?:\\.[^"\\]*)*"' |
|
101 | 101 | # Tail end of ''' string. |
|
102 | 102 | Single3 = r"[^'\\]*(?:(?:\\.|'(?!''))[^'\\]*)*'''" |
|
103 | 103 | # Tail end of """ string. |
|
104 | 104 | Double3 = r'[^"\\]*(?:(?:\\.|"(?!""))[^"\\]*)*"""' |
|
105 | 105 | Triple = group(StringPrefix + "'''", StringPrefix + '"""') |
|
106 | 106 | # Single-line ' or " string. |
|
107 | 107 | String = group(StringPrefix + r"'[^\n'\\]*(?:\\.[^\n'\\]*)*'", |
|
108 | 108 | StringPrefix + r'"[^\n"\\]*(?:\\.[^\n"\\]*)*"') |
|
109 | 109 | |
|
110 | 110 | # Because of leftmost-then-longest match semantics, be sure to put the |
|
111 | 111 | # longest operators first (e.g., if = came before ==, == would get |
|
112 | 112 | # recognized as two instances of =). |
|
113 | 113 | Operator = group(r"\*\*=?", r">>=?", r"<<=?", r"!=", |
|
114 | 114 | r"//=?", r"->", |
|
115 | 115 | r"[+\-*/%&|^=<>]=?", |
|
116 | 116 | r"~") |
|
117 | 117 | |
|
118 | 118 | Bracket = '[][(){}]' |
|
119 | 119 | Special = group(r'\r?\n', r'\.\.\.', r'[:;.,@]') |
|
120 | 120 | Funny = group(Operator, Bracket, Special) |
|
121 | 121 | |
|
122 | 122 | PlainToken = group(Number, Funny, String, Name) |
|
123 | 123 | Token = Ignore + PlainToken |
|
124 | 124 | |
|
125 | 125 | # First (or only) line of ' or " string. |
|
126 | 126 | ContStr = group(StringPrefix + r"'[^\n'\\]*(?:\\.[^\n'\\]*)*" + |
|
127 | 127 | group("'", r'\\\r?\n'), |
|
128 | 128 | StringPrefix + r'"[^\n"\\]*(?:\\.[^\n"\\]*)*' + |
|
129 | 129 | group('"', r'\\\r?\n')) |
|
130 | 130 | PseudoExtras = group(r'\\\r?\n', Comment, Triple) |
|
131 | 131 | PseudoToken = Whitespace + group(PseudoExtras, Number, Funny, ContStr, Name) |
|
132 | 132 | |
|
133 | 133 | def _compile(expr): |
|
134 | 134 | return re.compile(expr, re.UNICODE) |
|
135 | 135 | |
|
136 | 136 | tokenprog, pseudoprog, single3prog, double3prog = map( |
|
137 | 137 | _compile, (Token, PseudoToken, Single3, Double3)) |
|
138 | 138 | endprogs = {"'": _compile(Single), '"': _compile(Double), |
|
139 | 139 | "'''": single3prog, '"""': double3prog, |
|
140 | 140 | "r'''": single3prog, 'r"""': double3prog, |
|
141 | 141 | "b'''": single3prog, 'b"""': double3prog, |
|
142 | 142 | "R'''": single3prog, 'R"""': double3prog, |
|
143 | 143 | "B'''": single3prog, 'B"""': double3prog, |
|
144 | 144 | "br'''": single3prog, 'br"""': double3prog, |
|
145 | 145 | "bR'''": single3prog, 'bR"""': double3prog, |
|
146 | 146 | "Br'''": single3prog, 'Br"""': double3prog, |
|
147 | 147 | "BR'''": single3prog, 'BR"""': double3prog, |
|
148 | 148 | 'r': None, 'R': None, 'b': None, 'B': None} |
|
149 | 149 | |
|
150 | 150 | triple_quoted = {} |
|
151 | 151 | for t in ("'''", '"""', |
|
152 | 152 | "r'''", 'r"""', "R'''", 'R"""', |
|
153 | 153 | "b'''", 'b"""', "B'''", 'B"""', |
|
154 | 154 | "br'''", 'br"""', "Br'''", 'Br"""', |
|
155 | 155 | "bR'''", 'bR"""', "BR'''", 'BR"""'): |
|
156 | 156 | triple_quoted[t] = t |
|
157 | 157 | single_quoted = {} |
|
158 | 158 | for t in ("'", '"', |
|
159 | 159 | "r'", 'r"', "R'", 'R"', |
|
160 | 160 | "b'", 'b"', "B'", 'B"', |
|
161 | 161 | "br'", 'br"', "Br'", 'Br"', |
|
162 | 162 | "bR'", 'bR"', "BR'", 'BR"' ): |
|
163 | 163 | single_quoted[t] = t |
|
164 | 164 | |
|
165 | 165 | for _prefix in ['rb', 'rB', 'Rb', 'RB', 'u', 'U']: |
|
166 | 166 | _t2 = _prefix+'"""' |
|
167 | 167 | endprogs[_t2] = double3prog |
|
168 | 168 | triple_quoted[_t2] = _t2 |
|
169 | 169 | _t1 = _prefix + "'''" |
|
170 | 170 | endprogs[_t1] = single3prog |
|
171 | 171 | triple_quoted[_t1] = _t1 |
|
172 | 172 | single_quoted[_prefix+'"'] = _prefix+'"' |
|
173 | 173 | single_quoted[_prefix+"'"] = _prefix+"'" |
|
174 | 174 | del _prefix, _t2, _t1 |
|
175 | 175 | endprogs['u'] = None |
|
176 | 176 | endprogs['U'] = None |
|
177 | 177 | |
|
178 | 178 | del _compile |
|
179 | 179 | |
|
180 | 180 | tabsize = 8 |
|
181 | 181 | |
|
182 | 182 | class TokenError(Exception): pass |
|
183 | 183 | |
|
184 | 184 | class StopTokenizing(Exception): pass |
|
185 | 185 | |
|
186 | 186 | |
|
187 | 187 | class Untokenizer: |
|
188 | 188 | |
|
189 | 189 | def __init__(self): |
|
190 | 190 | self.tokens = [] |
|
191 | 191 | self.prev_row = 1 |
|
192 | 192 | self.prev_col = 0 |
|
193 | 193 | self.encoding = 'utf-8' |
|
194 | 194 | |
|
195 | 195 | def add_whitespace(self, tok_type, start): |
|
196 | 196 | row, col = start |
|
197 | 197 | assert row >= self.prev_row |
|
198 | 198 | col_offset = col - self.prev_col |
|
199 | 199 | if col_offset > 0: |
|
200 | 200 | self.tokens.append(" " * col_offset) |
|
201 | 201 | elif row > self.prev_row and tok_type not in (NEWLINE, NL, ENDMARKER): |
|
202 | 202 | # Line was backslash-continued. |
|
203 | 203 | self.tokens.append(" ") |
|
204 | 204 | |
|
205 | 205 | def untokenize(self, tokens): |
|
206 | 206 | iterable = iter(tokens) |
|
207 | 207 | for t in iterable: |
|
208 | 208 | if len(t) == 2: |
|
209 | 209 | self.compat(t, iterable) |
|
210 | 210 | break |
|
211 | 211 | tok_type, token, start, end = t[:4] |
|
212 | 212 | if tok_type == ENCODING: |
|
213 | 213 | self.encoding = token |
|
214 | 214 | continue |
|
215 | 215 | self.add_whitespace(tok_type, start) |
|
216 | 216 | self.tokens.append(token) |
|
217 | 217 | self.prev_row, self.prev_col = end |
|
218 | 218 | if tok_type in (NEWLINE, NL): |
|
219 | 219 | self.prev_row += 1 |
|
220 | 220 | self.prev_col = 0 |
|
221 | 221 | return "".join(self.tokens) |
|
222 | 222 | |
|
223 | 223 | def compat(self, token, iterable): |
|
224 | 224 | # This import is here to avoid problems when the itertools |
|
225 | 225 | # module is not built yet and tokenize is imported. |
|
226 | 226 | from itertools import chain |
|
227 | 227 | startline = False |
|
228 | 228 | prevstring = False |
|
229 | 229 | indents = [] |
|
230 | 230 | toks_append = self.tokens.append |
|
231 | 231 | |
|
232 | 232 | for tok in chain([token], iterable): |
|
233 | 233 | toknum, tokval = tok[:2] |
|
234 | 234 | if toknum == ENCODING: |
|
235 | 235 | self.encoding = tokval |
|
236 | 236 | continue |
|
237 | 237 | |
|
238 | 238 | if toknum in (NAME, NUMBER): |
|
239 | 239 | tokval += ' ' |
|
240 | 240 | |
|
241 | 241 | # Insert a space between two consecutive strings |
|
242 | 242 | if toknum == STRING: |
|
243 | 243 | if prevstring: |
|
244 | 244 | tokval = ' ' + tokval |
|
245 | 245 | prevstring = True |
|
246 | 246 | else: |
|
247 | 247 | prevstring = False |
|
248 | 248 | |
|
249 | 249 | if toknum == INDENT: |
|
250 | 250 | indents.append(tokval) |
|
251 | 251 | continue |
|
252 | 252 | elif toknum == DEDENT: |
|
253 | 253 | indents.pop() |
|
254 | 254 | continue |
|
255 | 255 | elif toknum in (NEWLINE, NL): |
|
256 | 256 | startline = True |
|
257 | 257 | elif startline and indents: |
|
258 | 258 | toks_append(indents[-1]) |
|
259 | 259 | startline = False |
|
260 | 260 | toks_append(tokval) |
|
261 | 261 | |
|
262 | 262 | |
|
263 | 263 | def untokenize(tokens): |
|
264 | 264 | """ |
|
265 | 265 | Convert ``tokens`` (an iterable) back into Python source code. Return |
|
266 | 266 | a bytes object, encoded using the encoding specified by the last |
|
267 | 267 | ENCODING token in ``tokens``, or UTF-8 if no ENCODING token is found. |
|
268 | 268 | |
|
269 | 269 | The result is guaranteed to tokenize back to match the input so that |
|
270 | 270 | the conversion is lossless and round-trips are assured. The |
|
271 | 271 | guarantee applies only to the token type and token string as the |
|
272 | 272 | spacing between tokens (column positions) may change. |
|
273 | 273 | |
|
274 | 274 | :func:`untokenize` has two modes. If the input tokens are sequences |
|
275 | 275 | of length 2 (``type``, ``string``) then spaces are added as necessary to |
|
276 | 276 | preserve the round-trip property. |
|
277 | 277 | |
|
278 | 278 | If the input tokens are sequences of length 4 or more (``type``, |
|
279 | 279 | ``string``, ``start``, ``end``), as returned by :func:`tokenize`, then |
|
280 | 280 | spaces are added so that each token appears in the result at the |
|
281 | 281 | position indicated by ``start`` and ``end``, if possible. |
|
282 | 282 | """ |
|
283 | 283 | return Untokenizer().untokenize(tokens) |
|
284 | 284 | |
|
285 | 285 | |
|
286 | 286 | def _get_normal_name(orig_enc): |
|
287 | 287 | """Imitates get_normal_name in tokenizer.c.""" |
|
288 | 288 | # Only care about the first 12 characters. |
|
289 | 289 | enc = orig_enc[:12].lower().replace("_", "-") |
|
290 | 290 | if enc == "utf-8" or enc.startswith("utf-8-"): |
|
291 | 291 | return "utf-8" |
|
292 | 292 | if enc in ("latin-1", "iso-8859-1", "iso-latin-1") or \ |
|
293 | 293 | enc.startswith(("latin-1-", "iso-8859-1-", "iso-latin-1-")): |
|
294 | 294 | return "iso-8859-1" |
|
295 | 295 | return orig_enc |
|
296 | 296 | |
|
297 | 297 | def detect_encoding(readline): |
|
298 | 298 | """ |
|
299 | 299 | The detect_encoding() function is used to detect the encoding that should |
|
300 | 300 | be used to decode a Python source file. It requires one argument, readline, |
|
301 | 301 | in the same way as the tokenize() generator. |
|
302 | 302 | |
|
303 | 303 | It will call readline a maximum of twice, and return the encoding used |
|
304 | 304 | (as a string) and a list of any lines (left as bytes) it has read in. |
|
305 | 305 | |
|
306 | 306 | It detects the encoding from the presence of a utf-8 bom or an encoding |
|
307 | 307 | cookie as specified in pep-0263. If both a bom and a cookie are present, |
|
308 | 308 | but disagree, a SyntaxError will be raised. If the encoding cookie is an |
|
309 | 309 | invalid charset, raise a SyntaxError. Note that if a utf-8 bom is found, |
|
310 | 310 | 'utf-8-sig' is returned. |
|
311 | 311 | |
|
312 | 312 | If no encoding is specified, then the default of 'utf-8' will be returned. |
|
313 | 313 | """ |
|
314 | 314 | bom_found = False |
|
315 | 315 | encoding = None |
|
316 | 316 | default = 'utf-8' |
|
317 | 317 | def read_or_stop(): |
|
318 | 318 | try: |
|
319 | 319 | return readline() |
|
320 | 320 | except StopIteration: |
|
321 | 321 | return b'' |
|
322 | 322 | |
|
323 | 323 | def find_cookie(line): |
|
324 | 324 | try: |
|
325 | 325 | # Decode as UTF-8. Either the line is an encoding declaration, |
|
326 | 326 | # in which case it should be pure ASCII, or it must be UTF-8 |
|
327 | 327 | # per default encoding. |
|
328 | 328 | line_string = line.decode('utf-8') |
|
329 | 329 | except UnicodeDecodeError: |
|
330 | 330 | raise SyntaxError("invalid or missing encoding declaration") |
|
331 | 331 | |
|
332 | 332 | matches = cookie_re.findall(line_string) |
|
333 | 333 | if not matches: |
|
334 | 334 | return None |
|
335 | 335 | encoding = _get_normal_name(matches[0]) |
|
336 | 336 | try: |
|
337 | 337 | codec = lookup(encoding) |
|
338 | 338 | except LookupError: |
|
339 | 339 | # This behaviour mimics the Python interpreter |
|
340 | 340 | raise SyntaxError("unknown encoding: " + encoding) |
|
341 | 341 | |
|
342 | 342 | if bom_found: |
|
343 | 343 | if encoding != 'utf-8': |
|
344 | 344 | # This behaviour mimics the Python interpreter |
|
345 | 345 | raise SyntaxError('encoding problem: utf-8') |
|
346 | 346 | encoding += '-sig' |
|
347 | 347 | return encoding |
|
348 | 348 | |
|
349 | 349 | first = read_or_stop() |
|
350 | 350 | if first.startswith(BOM_UTF8): |
|
351 | 351 | bom_found = True |
|
352 | 352 | first = first[3:] |
|
353 | 353 | default = 'utf-8-sig' |
|
354 | 354 | if not first: |
|
355 | 355 | return default, [] |
|
356 | 356 | |
|
357 | 357 | encoding = find_cookie(first) |
|
358 | 358 | if encoding: |
|
359 | 359 | return encoding, [first] |
|
360 | 360 | |
|
361 | 361 | second = read_or_stop() |
|
362 | 362 | if not second: |
|
363 | 363 | return default, [first] |
|
364 | 364 | |
|
365 | 365 | encoding = find_cookie(second) |
|
366 | 366 | if encoding: |
|
367 | 367 | return encoding, [first, second] |
|
368 | 368 | |
|
369 | 369 | return default, [first, second] |
|
370 | 370 | |
|
371 | 371 | |
|
372 | 372 | def open(filename): |
|
373 | 373 | """Open a file in read only mode using the encoding detected by |
|
374 | 374 | detect_encoding(). |
|
375 | 375 | """ |
|
376 | 376 | buffer = builtins.open(filename, 'rb') |
|
377 | 377 | encoding, lines = detect_encoding(buffer.readline) |
|
378 | 378 | buffer.seek(0) |
|
379 | 379 | text = TextIOWrapper(buffer, encoding, line_buffering=True) |
|
380 | 380 | text.mode = 'r' |
|
381 | 381 | return text |
|
382 | 382 | |
|
383 | 383 | |
|
384 | 384 | def tokenize(readline): |
|
385 | 385 | """ |
|
386 | 386 | The tokenize() generator requires one argument, readline, which |
|
387 | 387 | must be a callable object which provides the same interface as the |
|
388 | 388 | readline() method of built-in file objects. Each call to the function |
|
389 | 389 | should return one line of input as bytes. Alternately, readline |
|
390 | 390 | can be a callable function terminating with :class:`StopIteration`:: |
|
391 | 391 | |
|
392 | 392 | readline = open(myfile, 'rb').__next__ # Example of alternate readline |
|
393 | 393 | |
|
394 | 394 | The generator produces 5-tuples with these members: the token type; the |
|
395 | 395 | token string; a 2-tuple (srow, scol) of ints specifying the row and |
|
396 | 396 | column where the token begins in the source; a 2-tuple (erow, ecol) of |
|
397 | 397 | ints specifying the row and column where the token ends in the source; |
|
398 | 398 | and the line on which the token was found. The line passed is the |
|
399 | 399 | logical line; continuation lines are included. |
|
400 | 400 | |
|
401 | 401 | The first token sequence will always be an ENCODING token |
|
402 | 402 | which tells you which encoding was used to decode the bytes stream. |
|
403 | 403 | """ |
|
404 | 404 | # This import is here to avoid problems when the itertools module is not |
|
405 | 405 | # built yet and tokenize is imported. |
|
406 | 406 | from itertools import chain, repeat |
|
407 | 407 | encoding, consumed = detect_encoding(readline) |
|
408 | 408 | rl_gen = iter(readline, b"") |
|
409 | 409 | empty = repeat(b"") |
|
410 | 410 | return _tokenize(chain(consumed, rl_gen, empty).__next__, encoding) |
|
411 | 411 | |
|
412 | 412 | |
|
413 | 413 | def _tokenize(readline, encoding): |
|
414 | 414 | lnum = parenlev = continued = 0 |
|
415 | 415 | numchars = '0123456789' |
|
416 | 416 | contstr, needcont = '', 0 |
|
417 | 417 | contline = None |
|
418 | 418 | indents = [0] |
|
419 | 419 | |
|
420 | 420 | if encoding is not None: |
|
421 | 421 | if encoding == "utf-8-sig": |
|
422 | 422 | # BOM will already have been stripped. |
|
423 | 423 | encoding = "utf-8" |
|
424 | 424 | yield TokenInfo(ENCODING, encoding, (0, 0), (0, 0), '') |
|
425 | 425 | while True: # loop over lines in stream |
|
426 | 426 | try: |
|
427 | 427 | line = readline() |
|
428 | 428 | except StopIteration: |
|
429 | 429 | line = b'' |
|
430 | 430 | |
|
431 | 431 | if encoding is not None: |
|
432 | 432 | line = line.decode(encoding) |
|
433 | 433 | lnum += 1 |
|
434 | 434 | pos, max = 0, len(line) |
|
435 | 435 | |
|
436 | 436 | if contstr: # continued string |
|
437 | 437 | if not line: |
|
438 | 438 | raise TokenError("EOF in multi-line string", strstart) |
|
439 | 439 | endmatch = endprog.match(line) |
|
440 | 440 | if endmatch: |
|
441 | 441 | pos = end = endmatch.end(0) |
|
442 | 442 | yield TokenInfo(STRING, contstr + line[:end], |
|
443 | 443 | strstart, (lnum, end), contline + line) |
|
444 | 444 | contstr, needcont = '', 0 |
|
445 | 445 | contline = None |
|
446 | 446 | elif needcont and line[-2:] != '\\\n' and line[-3:] != '\\\r\n': |
|
447 | 447 | yield TokenInfo(ERRORTOKEN, contstr + line, |
|
448 | 448 | strstart, (lnum, len(line)), contline) |
|
449 | 449 | contstr = '' |
|
450 | 450 | contline = None |
|
451 | 451 | continue |
|
452 | 452 | else: |
|
453 | 453 | contstr = contstr + line |
|
454 | 454 | contline = contline + line |
|
455 | 455 | continue |
|
456 | 456 | |
|
457 | 457 | elif parenlev == 0 and not continued: # new statement |
|
458 | 458 | if not line: break |
|
459 | 459 | column = 0 |
|
460 | 460 | while pos < max: # measure leading whitespace |
|
461 | 461 | if line[pos] == ' ': |
|
462 | 462 | column += 1 |
|
463 | 463 | elif line[pos] == '\t': |
|
464 | 464 | column = (column//tabsize + 1)*tabsize |
|
465 | 465 | elif line[pos] == '\f': |
|
466 | 466 | column = 0 |
|
467 | 467 | else: |
|
468 | 468 | break |
|
469 | 469 | pos += 1 |
|
470 | 470 | if pos == max: |
|
471 | 471 | break |
|
472 | 472 | |
|
473 | 473 | if line[pos] in '#\r\n': # skip comments or blank lines |
|
474 | 474 | if line[pos] == '#': |
|
475 | 475 | comment_token = line[pos:].rstrip('\r\n') |
|
476 | 476 | nl_pos = pos + len(comment_token) |
|
477 | 477 | yield TokenInfo(COMMENT, comment_token, |
|
478 | 478 | (lnum, pos), (lnum, pos + len(comment_token)), line) |
|
479 | 479 | yield TokenInfo(NEWLINE, line[nl_pos:], |
|
480 | 480 | (lnum, nl_pos), (lnum, len(line)), line) |
|
481 | 481 | else: |
|
482 | 482 | yield TokenInfo(NEWLINE, line[pos:], |
|
483 | 483 | (lnum, pos), (lnum, len(line)), line) |
|
484 | 484 | continue |
|
485 | 485 | |
|
486 | 486 | if column > indents[-1]: # count indents or dedents |
|
487 | 487 | indents.append(column) |
|
488 | 488 | yield TokenInfo(INDENT, line[:pos], (lnum, 0), (lnum, pos), line) |
|
489 | 489 | while column < indents[-1]: |
|
490 | 490 | if column not in indents: |
|
491 | 491 | raise IndentationError( |
|
492 | 492 | "unindent does not match any outer indentation level", |
|
493 | 493 | ("<tokenize>", lnum, pos, line)) |
|
494 | 494 | indents = indents[:-1] |
|
495 | 495 | yield TokenInfo(DEDENT, '', (lnum, pos), (lnum, pos), line) |
|
496 | 496 | |
|
497 | 497 | else: # continued statement |
|
498 | 498 | if not line: |
|
499 | 499 | raise TokenError("EOF in multi-line statement", (lnum, 0)) |
|
500 | 500 | continued = 0 |
|
501 | 501 | |
|
502 | 502 | while pos < max: |
|
503 | 503 | pseudomatch = pseudoprog.match(line, pos) |
|
504 | 504 | if pseudomatch: # scan for tokens |
|
505 | 505 | start, end = pseudomatch.span(1) |
|
506 | 506 | spos, epos, pos = (lnum, start), (lnum, end), end |
|
507 | 507 | token, initial = line[start:end], line[start] |
|
508 | 508 | |
|
509 | 509 | if (initial in numchars or # ordinary number |
|
510 | 510 | (initial == '.' and token != '.' and token != '...')): |
|
511 | 511 | yield TokenInfo(NUMBER, token, spos, epos, line) |
|
512 | 512 | elif initial in '\r\n': |
|
513 | 513 | yield TokenInfo(NL if parenlev > 0 else NEWLINE, |
|
514 | 514 | token, spos, epos, line) |
|
515 | 515 | elif initial == '#': |
|
516 | 516 | assert not token.endswith("\n") |
|
517 | 517 | yield TokenInfo(COMMENT, token, spos, epos, line) |
|
518 | 518 | elif token in triple_quoted: |
|
519 | 519 | endprog = endprogs[token] |
|
520 | 520 | endmatch = endprog.match(line, pos) |
|
521 | 521 | if endmatch: # all on one line |
|
522 | 522 | pos = endmatch.end(0) |
|
523 | 523 | token = line[start:pos] |
|
524 | 524 | yield TokenInfo(STRING, token, spos, (lnum, pos), line) |
|
525 | 525 | else: |
|
526 | 526 | strstart = (lnum, start) # multiple lines |
|
527 | 527 | contstr = line[start:] |
|
528 | 528 | contline = line |
|
529 | 529 | break |
|
530 | 530 | elif initial in single_quoted or \ |
|
531 | 531 | token[:2] in single_quoted or \ |
|
532 | 532 | token[:3] in single_quoted: |
|
533 | 533 | if token[-1] == '\n': # continued string |
|
534 | 534 | strstart = (lnum, start) |
|
535 | 535 | endprog = (endprogs[initial] or endprogs[token[1]] or |
|
536 | 536 | endprogs[token[2]]) |
|
537 | 537 | contstr, needcont = line[start:], 1 |
|
538 | 538 | contline = line |
|
539 | 539 | break |
|
540 | 540 | else: # ordinary string |
|
541 | 541 | yield TokenInfo(STRING, token, spos, epos, line) |
|
542 | 542 | elif initial.isidentifier(): # ordinary name |
|
543 | 543 | yield TokenInfo(NAME, token, spos, epos, line) |
|
544 | 544 | elif initial == '\\': # continued stmt |
|
545 | 545 | continued = 1 |
|
546 | 546 | else: |
|
547 | 547 | if initial in '([{': |
|
548 | 548 | parenlev += 1 |
|
549 | 549 | elif initial in ')]}': |
|
550 | 550 | parenlev -= 1 |
|
551 | 551 | yield TokenInfo(OP, token, spos, epos, line) |
|
552 | 552 | else: |
|
553 | 553 | yield TokenInfo(ERRORTOKEN, line[pos], |
|
554 | 554 | (lnum, pos), (lnum, pos+1), line) |
|
555 | 555 | pos += 1 |
|
556 | 556 | |
|
557 | 557 | for indent in indents[1:]: # pop remaining indent levels |
|
558 | 558 | yield TokenInfo(DEDENT, '', (lnum, 0), (lnum, 0), '') |
|
559 | 559 | yield TokenInfo(ENDMARKER, '', (lnum, 0), (lnum, 0), '') |
|
560 | 560 | |
|
561 | 561 | |
|
562 | 562 | # An undocumented, backwards compatible, API for all the places in the standard |
|
563 | 563 | # library that expect to be able to use tokenize with strings |
|
564 | 564 | def generate_tokens(readline): |
|
565 | 565 | return _tokenize(readline, None) |
|
566 | 566 | |
|
567 | 567 | if __name__ == "__main__": |
|
568 | 568 | # Quick sanity check |
|
569 | 569 | s = b'''def parseline(self, line): |
|
570 | 570 | """Parse the line into a command name and a string containing |
|
571 | 571 | the arguments. Returns a tuple containing (command, args, line). |
|
572 | 572 | 'command' and 'args' may be None if the line couldn't be parsed. |
|
573 | 573 | """ |
|
574 | 574 | line = line.strip() |
|
575 | 575 | if not line: |
|
576 | 576 | return None, None, line |
|
577 | 577 | elif line[0] == '?': |
|
578 | 578 | line = 'help ' + line[1:] |
|
579 | 579 | elif line[0] == '!': |
|
580 | 580 | if hasattr(self, 'do_shell'): |
|
581 | 581 | line = 'shell ' + line[1:] |
|
582 | 582 | else: |
|
583 | 583 | return None, None, line |
|
584 | 584 | i, n = 0, len(line) |
|
585 | 585 | while i < n and line[i] in self.identchars: i = i+1 |
|
586 | 586 | cmd, arg = line[:i], line[i:].strip() |
|
587 | 587 | return cmd, arg, line |
|
588 | 588 | ''' |
|
589 | 589 | for tok in tokenize(iter(s.splitlines()).__next__): |
|
590 | 590 | print(tok) |
General Comments 0
You need to be logged in to leave comments.
Login now