Show More
@@ -1,773 +1,772 b'' | |||
|
1 | 1 | """DEPRECATED: Input handling and transformation machinery. |
|
2 | 2 | |
|
3 | 3 | This module was deprecated in IPython 7.0, in favour of inputtransformer2. |
|
4 | 4 | |
|
5 | 5 | The first class in this module, :class:`InputSplitter`, is designed to tell when |
|
6 | 6 | input from a line-oriented frontend is complete and should be executed, and when |
|
7 | 7 | the user should be prompted for another line of code instead. The name 'input |
|
8 | 8 | splitter' is largely for historical reasons. |
|
9 | 9 | |
|
10 | 10 | A companion, :class:`IPythonInputSplitter`, provides the same functionality but |
|
11 | 11 | with full support for the extended IPython syntax (magics, system calls, etc). |
|
12 | 12 | The code to actually do these transformations is in :mod:`IPython.core.inputtransformer`. |
|
13 | 13 | :class:`IPythonInputSplitter` feeds the raw code to the transformers in order |
|
14 | 14 | and stores the results. |
|
15 | 15 | |
|
16 | 16 | For more details, see the class docstrings below. |
|
17 | 17 | """ |
|
18 | 18 | |
|
19 | 19 | from warnings import warn |
|
20 | 20 | |
|
21 | 21 | warn('IPython.core.inputsplitter is deprecated since IPython 7 in favor of `IPython.core.inputtransformer2`', |
|
22 | 22 | DeprecationWarning) |
|
23 | 23 | |
|
24 | 24 | # Copyright (c) IPython Development Team. |
|
25 | 25 | # Distributed under the terms of the Modified BSD License. |
|
26 | 26 | import ast |
|
27 | 27 | import codeop |
|
28 | 28 | import io |
|
29 | 29 | import re |
|
30 | 30 | import sys |
|
31 | 31 | import tokenize |
|
32 | 32 | import warnings |
|
33 | 33 | |
|
34 | from IPython.utils.py3compat import cast_unicode | |
|
35 | 34 | from IPython.core.inputtransformer import (leading_indent, |
|
36 | 35 | classic_prompt, |
|
37 | 36 | ipy_prompt, |
|
38 | 37 | cellmagic, |
|
39 | 38 | assemble_logical_lines, |
|
40 | 39 | help_end, |
|
41 | 40 | escaped_commands, |
|
42 | 41 | assign_from_magic, |
|
43 | 42 | assign_from_system, |
|
44 | 43 | assemble_python_lines, |
|
45 | 44 | ) |
|
46 | 45 | |
|
47 | 46 | # These are available in this module for backwards compatibility. |
|
48 | 47 | from IPython.core.inputtransformer import (ESC_SHELL, ESC_SH_CAP, ESC_HELP, |
|
49 | 48 | ESC_HELP2, ESC_MAGIC, ESC_MAGIC2, |
|
50 | 49 | ESC_QUOTE, ESC_QUOTE2, ESC_PAREN, ESC_SEQUENCES) |
|
51 | 50 | |
|
52 | 51 | #----------------------------------------------------------------------------- |
|
53 | 52 | # Utilities |
|
54 | 53 | #----------------------------------------------------------------------------- |
|
55 | 54 | |
|
56 | 55 | # FIXME: These are general-purpose utilities that later can be moved to the |
|
57 | 56 | # general ward. Kept here for now because we're being very strict about test |
|
58 | 57 | # coverage with this code, and this lets us ensure that we keep 100% coverage |
|
59 | 58 | # while developing. |
|
60 | 59 | |
|
61 | 60 | # compiled regexps for autoindent management |
|
62 | 61 | dedent_re = re.compile('|'.join([ |
|
63 | 62 | r'^\s+raise(\s.*)?$', # raise statement (+ space + other stuff, maybe) |
|
64 | 63 | r'^\s+raise\([^\)]*\).*$', # wacky raise with immediate open paren |
|
65 | 64 | r'^\s+return(\s.*)?$', # normal return (+ space + other stuff, maybe) |
|
66 | 65 | r'^\s+return\([^\)]*\).*$', # wacky return with immediate open paren |
|
67 | 66 | r'^\s+pass\s*$', # pass (optionally followed by trailing spaces) |
|
68 | 67 | r'^\s+break\s*$', # break (optionally followed by trailing spaces) |
|
69 | 68 | r'^\s+continue\s*$', # continue (optionally followed by trailing spaces) |
|
70 | 69 | ])) |
|
71 | 70 | ini_spaces_re = re.compile(r'^([ \t\r\f\v]+)') |
|
72 | 71 | |
|
73 | 72 | # regexp to match pure comment lines so we don't accidentally insert 'if 1:' |
|
74 | 73 | # before pure comments |
|
75 | 74 | comment_line_re = re.compile(r'^\s*\#') |
|
76 | 75 | |
|
77 | 76 | |
|
78 | 77 | def num_ini_spaces(s): |
|
79 | 78 | """Return the number of initial spaces in a string. |
|
80 | 79 | |
|
81 | 80 | Note that tabs are counted as a single space. For now, we do *not* support |
|
82 | 81 | mixing of tabs and spaces in the user's input. |
|
83 | 82 | |
|
84 | 83 | Parameters |
|
85 | 84 | ---------- |
|
86 | 85 | s : string |
|
87 | 86 | |
|
88 | 87 | Returns |
|
89 | 88 | ------- |
|
90 | 89 | n : int |
|
91 | 90 | """ |
|
92 | 91 | |
|
93 | 92 | ini_spaces = ini_spaces_re.match(s) |
|
94 | 93 | if ini_spaces: |
|
95 | 94 | return ini_spaces.end() |
|
96 | 95 | else: |
|
97 | 96 | return 0 |
|
98 | 97 | |
|
99 | 98 | # Fake token types for partial_tokenize: |
|
100 | 99 | INCOMPLETE_STRING = tokenize.N_TOKENS |
|
101 | 100 | IN_MULTILINE_STATEMENT = tokenize.N_TOKENS + 1 |
|
102 | 101 | |
|
103 | 102 | # The 2 classes below have the same API as TokenInfo, but don't try to look up |
|
104 | 103 | # a token type name that they won't find. |
|
105 | 104 | class IncompleteString: |
|
106 | 105 | type = exact_type = INCOMPLETE_STRING |
|
107 | 106 | def __init__(self, s, start, end, line): |
|
108 | 107 | self.s = s |
|
109 | 108 | self.start = start |
|
110 | 109 | self.end = end |
|
111 | 110 | self.line = line |
|
112 | 111 | |
|
113 | 112 | class InMultilineStatement: |
|
114 | 113 | type = exact_type = IN_MULTILINE_STATEMENT |
|
115 | 114 | def __init__(self, pos, line): |
|
116 | 115 | self.s = '' |
|
117 | 116 | self.start = self.end = pos |
|
118 | 117 | self.line = line |
|
119 | 118 | |
|
120 | 119 | def partial_tokens(s): |
|
121 | 120 | """Iterate over tokens from a possibly-incomplete string of code. |
|
122 | 121 | |
|
123 | 122 | This adds two special token types: INCOMPLETE_STRING and |
|
124 | 123 | IN_MULTILINE_STATEMENT. These can only occur as the last token yielded, and |
|
125 | 124 | represent the two main ways for code to be incomplete. |
|
126 | 125 | """ |
|
127 | 126 | readline = io.StringIO(s).readline |
|
128 | 127 | token = tokenize.TokenInfo(tokenize.NEWLINE, '', (1, 0), (1, 0), '') |
|
129 | 128 | try: |
|
130 | 129 | for token in tokenize.generate_tokens(readline): |
|
131 | 130 | yield token |
|
132 | 131 | except tokenize.TokenError as e: |
|
133 | 132 | # catch EOF error |
|
134 | 133 | lines = s.splitlines(keepends=True) |
|
135 | 134 | end = len(lines), len(lines[-1]) |
|
136 | 135 | if 'multi-line string' in e.args[0]: |
|
137 | 136 | l, c = start = token.end |
|
138 | 137 | s = lines[l-1][c:] + ''.join(lines[l:]) |
|
139 | 138 | yield IncompleteString(s, start, end, lines[-1]) |
|
140 | 139 | elif 'multi-line statement' in e.args[0]: |
|
141 | 140 | yield InMultilineStatement(end, lines[-1]) |
|
142 | 141 | else: |
|
143 | 142 | raise |
|
144 | 143 | |
|
145 | 144 | def find_next_indent(code): |
|
146 | 145 | """Find the number of spaces for the next line of indentation""" |
|
147 | 146 | tokens = list(partial_tokens(code)) |
|
148 | 147 | if tokens[-1].type == tokenize.ENDMARKER: |
|
149 | 148 | tokens.pop() |
|
150 | 149 | if not tokens: |
|
151 | 150 | return 0 |
|
152 | 151 | while (tokens[-1].type in {tokenize.DEDENT, tokenize.NEWLINE, tokenize.COMMENT}): |
|
153 | 152 | tokens.pop() |
|
154 | 153 | |
|
155 | 154 | if tokens[-1].type == INCOMPLETE_STRING: |
|
156 | 155 | # Inside a multiline string |
|
157 | 156 | return 0 |
|
158 | 157 | |
|
159 | 158 | # Find the indents used before |
|
160 | 159 | prev_indents = [0] |
|
161 | 160 | def _add_indent(n): |
|
162 | 161 | if n != prev_indents[-1]: |
|
163 | 162 | prev_indents.append(n) |
|
164 | 163 | |
|
165 | 164 | tokiter = iter(tokens) |
|
166 | 165 | for tok in tokiter: |
|
167 | 166 | if tok.type in {tokenize.INDENT, tokenize.DEDENT}: |
|
168 | 167 | _add_indent(tok.end[1]) |
|
169 | 168 | elif (tok.type == tokenize.NL): |
|
170 | 169 | try: |
|
171 | 170 | _add_indent(next(tokiter).start[1]) |
|
172 | 171 | except StopIteration: |
|
173 | 172 | break |
|
174 | 173 | |
|
175 | 174 | last_indent = prev_indents.pop() |
|
176 | 175 | |
|
177 | 176 | # If we've just opened a multiline statement (e.g. 'a = ['), indent more |
|
178 | 177 | if tokens[-1].type == IN_MULTILINE_STATEMENT: |
|
179 | 178 | if tokens[-2].exact_type in {tokenize.LPAR, tokenize.LSQB, tokenize.LBRACE}: |
|
180 | 179 | return last_indent + 4 |
|
181 | 180 | return last_indent |
|
182 | 181 | |
|
183 | 182 | if tokens[-1].exact_type == tokenize.COLON: |
|
184 | 183 | # Line ends with colon - indent |
|
185 | 184 | return last_indent + 4 |
|
186 | 185 | |
|
187 | 186 | if last_indent: |
|
188 | 187 | # Examine the last line for dedent cues - statements like return or |
|
189 | 188 | # raise which normally end a block of code. |
|
190 | 189 | last_line_starts = 0 |
|
191 | 190 | for i, tok in enumerate(tokens): |
|
192 | 191 | if tok.type == tokenize.NEWLINE: |
|
193 | 192 | last_line_starts = i + 1 |
|
194 | 193 | |
|
195 | 194 | last_line_tokens = tokens[last_line_starts:] |
|
196 | 195 | names = [t.string for t in last_line_tokens if t.type == tokenize.NAME] |
|
197 | 196 | if names and names[0] in {'raise', 'return', 'pass', 'break', 'continue'}: |
|
198 | 197 | # Find the most recent indentation less than the current level |
|
199 | 198 | for indent in reversed(prev_indents): |
|
200 | 199 | if indent < last_indent: |
|
201 | 200 | return indent |
|
202 | 201 | |
|
203 | 202 | return last_indent |
|
204 | 203 | |
|
205 | 204 | |
|
206 | 205 | def last_blank(src): |
|
207 | 206 | """Determine if the input source ends in a blank. |
|
208 | 207 | |
|
209 | 208 | A blank is either a newline or a line consisting of whitespace. |
|
210 | 209 | |
|
211 | 210 | Parameters |
|
212 | 211 | ---------- |
|
213 | 212 | src : string |
|
214 | 213 | A single or multiline string. |
|
215 | 214 | """ |
|
216 | 215 | if not src: return False |
|
217 | 216 | ll = src.splitlines()[-1] |
|
218 | 217 | return (ll == '') or ll.isspace() |
|
219 | 218 | |
|
220 | 219 | |
|
221 | 220 | last_two_blanks_re = re.compile(r'\n\s*\n\s*$', re.MULTILINE) |
|
222 | 221 | last_two_blanks_re2 = re.compile(r'.+\n\s*\n\s+$', re.MULTILINE) |
|
223 | 222 | |
|
224 | 223 | def last_two_blanks(src): |
|
225 | 224 | """Determine if the input source ends in two blanks. |
|
226 | 225 | |
|
227 | 226 | A blank is either a newline or a line consisting of whitespace. |
|
228 | 227 | |
|
229 | 228 | Parameters |
|
230 | 229 | ---------- |
|
231 | 230 | src : string |
|
232 | 231 | A single or multiline string. |
|
233 | 232 | """ |
|
234 | 233 | if not src: return False |
|
235 | 234 | # The logic here is tricky: I couldn't get a regexp to work and pass all |
|
236 | 235 | # the tests, so I took a different approach: split the source by lines, |
|
237 | 236 | # grab the last two and prepend '###\n' as a stand-in for whatever was in |
|
238 | 237 | # the body before the last two lines. Then, with that structure, it's |
|
239 | 238 | # possible to analyze with two regexps. Not the most elegant solution, but |
|
240 | 239 | # it works. If anyone tries to change this logic, make sure to validate |
|
241 | 240 | # the whole test suite first! |
|
242 | 241 | new_src = '\n'.join(['###\n'] + src.splitlines()[-2:]) |
|
243 | 242 | return (bool(last_two_blanks_re.match(new_src)) or |
|
244 | 243 | bool(last_two_blanks_re2.match(new_src)) ) |
|
245 | 244 | |
|
246 | 245 | |
|
247 | 246 | def remove_comments(src): |
|
248 | 247 | """Remove all comments from input source. |
|
249 | 248 | |
|
250 | 249 | Note: comments are NOT recognized inside of strings! |
|
251 | 250 | |
|
252 | 251 | Parameters |
|
253 | 252 | ---------- |
|
254 | 253 | src : string |
|
255 | 254 | A single or multiline input string. |
|
256 | 255 | |
|
257 | 256 | Returns |
|
258 | 257 | ------- |
|
259 | 258 | String with all Python comments removed. |
|
260 | 259 | """ |
|
261 | 260 | |
|
262 | 261 | return re.sub('#.*', '', src) |
|
263 | 262 | |
|
264 | 263 | |
|
265 | 264 | def get_input_encoding(): |
|
266 | 265 | """Return the default standard input encoding. |
|
267 | 266 | |
|
268 | 267 | If sys.stdin has no encoding, 'ascii' is returned.""" |
|
269 | 268 | # There are strange environments for which sys.stdin.encoding is None. We |
|
270 | 269 | # ensure that a valid encoding is returned. |
|
271 | 270 | encoding = getattr(sys.stdin, 'encoding', None) |
|
272 | 271 | if encoding is None: |
|
273 | 272 | encoding = 'ascii' |
|
274 | 273 | return encoding |
|
275 | 274 | |
|
276 | 275 | #----------------------------------------------------------------------------- |
|
277 | 276 | # Classes and functions for normal Python syntax handling |
|
278 | 277 | #----------------------------------------------------------------------------- |
|
279 | 278 | |
|
280 | 279 | class InputSplitter(object): |
|
281 | 280 | r"""An object that can accumulate lines of Python source before execution. |
|
282 | 281 | |
|
283 | 282 | This object is designed to be fed python source line-by-line, using |
|
284 | 283 | :meth:`push`. It will return on each push whether the currently pushed |
|
285 | 284 | code could be executed already. In addition, it provides a method called |
|
286 | 285 | :meth:`push_accepts_more` that can be used to query whether more input |
|
287 | 286 | can be pushed into a single interactive block. |
|
288 | 287 | |
|
289 | 288 | This is a simple example of how an interactive terminal-based client can use |
|
290 | 289 | this tool:: |
|
291 | 290 | |
|
292 | 291 | isp = InputSplitter() |
|
293 | 292 | while isp.push_accepts_more(): |
|
294 | 293 | indent = ' '*isp.indent_spaces |
|
295 | 294 | prompt = '>>> ' + indent |
|
296 | 295 | line = indent + raw_input(prompt) |
|
297 | 296 | isp.push(line) |
|
298 | 297 | print 'Input source was:\n', isp.source_reset(), |
|
299 | 298 | """ |
|
300 | 299 | # A cache for storing the current indentation |
|
301 | 300 | # The first value stores the most recently processed source input |
|
302 | 301 | # The second value is the number of spaces for the current indentation |
|
303 | 302 | # If self.source matches the first value, the second value is a valid |
|
304 | 303 | # current indentation. Otherwise, the cache is invalid and the indentation |
|
305 | 304 | # must be recalculated. |
|
306 | 305 | _indent_spaces_cache = None, None |
|
307 | 306 | # String, indicating the default input encoding. It is computed by default |
|
308 | 307 | # at initialization time via get_input_encoding(), but it can be reset by a |
|
309 | 308 | # client with specific knowledge of the encoding. |
|
310 | 309 | encoding = '' |
|
311 | 310 | # String where the current full source input is stored, properly encoded. |
|
312 | 311 | # Reading this attribute is the normal way of querying the currently pushed |
|
313 | 312 | # source code, that has been properly encoded. |
|
314 | 313 | source = '' |
|
315 | 314 | # Code object corresponding to the current source. It is automatically |
|
316 | 315 | # synced to the source, so it can be queried at any time to obtain the code |
|
317 | 316 | # object; it will be None if the source doesn't compile to valid Python. |
|
318 | 317 | code = None |
|
319 | 318 | |
|
320 | 319 | # Private attributes |
|
321 | 320 | |
|
322 | 321 | # List with lines of input accumulated so far |
|
323 | 322 | _buffer = None |
|
324 | 323 | # Command compiler |
|
325 | 324 | _compile = None |
|
326 | 325 | # Boolean indicating whether the current block is complete |
|
327 | 326 | _is_complete = None |
|
328 | 327 | # Boolean indicating whether the current block has an unrecoverable syntax error |
|
329 | 328 | _is_invalid = False |
|
330 | 329 | |
|
331 | 330 | def __init__(self): |
|
332 | 331 | """Create a new InputSplitter instance. |
|
333 | 332 | """ |
|
334 | 333 | self._buffer = [] |
|
335 | 334 | self._compile = codeop.CommandCompiler() |
|
336 | 335 | self.encoding = get_input_encoding() |
|
337 | 336 | |
|
338 | 337 | def reset(self): |
|
339 | 338 | """Reset the input buffer and associated state.""" |
|
340 | 339 | self._buffer[:] = [] |
|
341 | 340 | self.source = '' |
|
342 | 341 | self.code = None |
|
343 | 342 | self._is_complete = False |
|
344 | 343 | self._is_invalid = False |
|
345 | 344 | |
|
346 | 345 | def source_reset(self): |
|
347 | 346 | """Return the input source and perform a full reset. |
|
348 | 347 | """ |
|
349 | 348 | out = self.source |
|
350 | 349 | self.reset() |
|
351 | 350 | return out |
|
352 | 351 | |
|
353 | 352 | def check_complete(self, source): |
|
354 | 353 | """Return whether a block of code is ready to execute, or should be continued |
|
355 | 354 | |
|
356 | 355 | This is a non-stateful API, and will reset the state of this InputSplitter. |
|
357 | 356 | |
|
358 | 357 | Parameters |
|
359 | 358 | ---------- |
|
360 | 359 | source : string |
|
361 | 360 | Python input code, which can be multiline. |
|
362 | 361 | |
|
363 | 362 | Returns |
|
364 | 363 | ------- |
|
365 | 364 | status : str |
|
366 | 365 | One of 'complete', 'incomplete', or 'invalid' if source is not a |
|
367 | 366 | prefix of valid code. |
|
368 | 367 | indent_spaces : int or None |
|
369 | 368 | The number of spaces by which to indent the next line of code. If |
|
370 | 369 | status is not 'incomplete', this is None. |
|
371 | 370 | """ |
|
372 | 371 | self.reset() |
|
373 | 372 | try: |
|
374 | 373 | self.push(source) |
|
375 | 374 | except SyntaxError: |
|
376 | 375 | # Transformers in IPythonInputSplitter can raise SyntaxError, |
|
377 | 376 | # which push() will not catch. |
|
378 | 377 | return 'invalid', None |
|
379 | 378 | else: |
|
380 | 379 | if self._is_invalid: |
|
381 | 380 | return 'invalid', None |
|
382 | 381 | elif self.push_accepts_more(): |
|
383 | 382 | return 'incomplete', self.get_indent_spaces() |
|
384 | 383 | else: |
|
385 | 384 | return 'complete', None |
|
386 | 385 | finally: |
|
387 | 386 | self.reset() |
|
388 | 387 | |
|
389 | 388 | def push(self, lines:str) -> bool: |
|
390 | 389 | """Push one or more lines of input. |
|
391 | 390 | |
|
392 | 391 | This stores the given lines and returns a status code indicating |
|
393 | 392 | whether the code forms a complete Python block or not. |
|
394 | 393 | |
|
395 | 394 | Any exceptions generated in compilation are swallowed, but if an |
|
396 | 395 | exception was produced, the method returns True. |
|
397 | 396 | |
|
398 | 397 | Parameters |
|
399 | 398 | ---------- |
|
400 | 399 | lines : string |
|
401 | 400 | One or more lines of Python input. |
|
402 | 401 | |
|
403 | 402 | Returns |
|
404 | 403 | ------- |
|
405 | 404 | is_complete : boolean |
|
406 | 405 | True if the current input source (the result of the current input |
|
407 | 406 | plus prior inputs) forms a complete Python execution block. Note that |
|
408 | 407 | this value is also stored as a private attribute (``_is_complete``), so it |
|
409 | 408 | can be queried at any time. |
|
410 | 409 | """ |
|
411 | 410 | assert isinstance(lines, str) |
|
412 | 411 | self._store(lines) |
|
413 | 412 | source = self.source |
|
414 | 413 | |
|
415 | 414 | # Before calling _compile(), reset the code object to None so that if an |
|
416 | 415 | # exception is raised in compilation, we don't mislead by having |
|
417 | 416 | # inconsistent code/source attributes. |
|
418 | 417 | self.code, self._is_complete = None, None |
|
419 | 418 | self._is_invalid = False |
|
420 | 419 | |
|
421 | 420 | # Honor termination lines properly |
|
422 | 421 | if source.endswith('\\\n'): |
|
423 | 422 | return False |
|
424 | 423 | |
|
425 | 424 | try: |
|
426 | 425 | with warnings.catch_warnings(): |
|
427 | 426 | warnings.simplefilter('error', SyntaxWarning) |
|
428 | 427 | self.code = self._compile(source, symbol="exec") |
|
429 | 428 | # Invalid syntax can produce any of a number of different errors from |
|
430 | 429 | # inside the compiler, so we have to catch them all. Syntax errors |
|
431 | 430 | # immediately produce a 'ready' block, so the invalid Python can be |
|
432 | 431 | # sent to the kernel for evaluation with possible ipython |
|
433 | 432 | # special-syntax conversion. |
|
434 | 433 | except (SyntaxError, OverflowError, ValueError, TypeError, |
|
435 | 434 | MemoryError, SyntaxWarning): |
|
436 | 435 | self._is_complete = True |
|
437 | 436 | self._is_invalid = True |
|
438 | 437 | else: |
|
439 | 438 | # Compilation didn't produce any exceptions (though it may not have |
|
440 | 439 | # given a complete code object) |
|
441 | 440 | self._is_complete = self.code is not None |
|
442 | 441 | |
|
443 | 442 | return self._is_complete |
|
444 | 443 | |
|
445 | 444 | def push_accepts_more(self): |
|
446 | 445 | """Return whether a block of interactive input can accept more input. |
|
447 | 446 | |
|
448 | 447 | This method is meant to be used by line-oriented frontends, who need to |
|
449 | 448 | guess whether a block is complete or not based solely on prior and |
|
450 | 449 | current input lines. The InputSplitter considers it has a complete |
|
451 | 450 | interactive block and will not accept more input when either: |
|
452 | 451 | |
|
453 | 452 | * A SyntaxError is raised |
|
454 | 453 | |
|
455 | 454 | * The code is complete and consists of a single line or a single |
|
456 | 455 | non-compound statement |
|
457 | 456 | |
|
458 | 457 | * The code is complete and has a blank line at the end |
|
459 | 458 | |
|
460 | 459 | If the current input produces a syntax error, this method immediately |
|
461 | 460 | returns False but does *not* raise the syntax error exception, as |
|
462 | 461 | typically clients will want to send invalid syntax to an execution |
|
463 | 462 | backend which might convert the invalid syntax into valid Python via |
|
464 | 463 | one of the dynamic IPython mechanisms. |
|
465 | 464 | """ |
|
466 | 465 | |
|
467 | 466 | # With incomplete input, unconditionally accept more |
|
468 | 467 | # A syntax error also sets _is_complete to True - see push() |
|
469 | 468 | if not self._is_complete: |
|
470 | 469 | #print("Not complete") # debug |
|
471 | 470 | return True |
|
472 | 471 | |
|
473 | 472 | # The user can make any (complete) input execute by leaving a blank line |
|
474 | 473 | last_line = self.source.splitlines()[-1] |
|
475 | 474 | if (not last_line) or last_line.isspace(): |
|
476 | 475 | #print("Blank line") # debug |
|
477 | 476 | return False |
|
478 | 477 | |
|
479 | 478 | # If there's just a single line or AST node, and we're flush left, as is |
|
480 | 479 | # the case after a simple statement such as 'a=1', we want to execute it |
|
481 | 480 | # straight away. |
|
482 | 481 | if self.get_indent_spaces() == 0: |
|
483 | 482 | if len(self.source.splitlines()) <= 1: |
|
484 | 483 | return False |
|
485 | 484 | |
|
486 | 485 | try: |
|
487 | 486 | code_ast = ast.parse(u''.join(self._buffer)) |
|
488 | 487 | except Exception: |
|
489 | 488 | #print("Can't parse AST") # debug |
|
490 | 489 | return False |
|
491 | 490 | else: |
|
492 | 491 | if len(code_ast.body) == 1 and \ |
|
493 | 492 | not hasattr(code_ast.body[0], 'body'): |
|
494 | 493 | #print("Simple statement") # debug |
|
495 | 494 | return False |
|
496 | 495 | |
|
497 | 496 | # General fallback - accept more code |
|
498 | 497 | return True |
|
499 | 498 | |
|
500 | 499 | def get_indent_spaces(self): |
|
501 | 500 | sourcefor, n = self._indent_spaces_cache |
|
502 | 501 | if sourcefor == self.source: |
|
503 | 502 | return n |
|
504 | 503 | |
|
505 | 504 | # self.source always has a trailing newline |
|
506 | 505 | n = find_next_indent(self.source[:-1]) |
|
507 | 506 | self._indent_spaces_cache = (self.source, n) |
|
508 | 507 | return n |
|
509 | 508 | |
|
510 | 509 | # Backwards compatibility. I think all code that used .indent_spaces was |
|
511 | 510 | # inside IPython, but we can leave this here until IPython 7 in case any |
|
512 | 511 | # other modules are using it. -TK, November 2017 |
|
513 | 512 | indent_spaces = property(get_indent_spaces) |
|
514 | 513 | |
|
515 | 514 | def _store(self, lines, buffer=None, store='source'): |
|
516 | 515 | """Store one or more lines of input. |
|
517 | 516 | |
|
518 | 517 | If input lines are not newline-terminated, a newline is automatically |
|
519 | 518 | appended.""" |
|
520 | 519 | |
|
521 | 520 | if buffer is None: |
|
522 | 521 | buffer = self._buffer |
|
523 | 522 | |
|
524 | 523 | if lines.endswith('\n'): |
|
525 | 524 | buffer.append(lines) |
|
526 | 525 | else: |
|
527 | 526 | buffer.append(lines+'\n') |
|
528 | 527 | setattr(self, store, self._set_source(buffer)) |
|
529 | 528 | |
|
530 | 529 | def _set_source(self, buffer): |
|
531 | 530 | return u''.join(buffer) |
|
532 | 531 | |
|
533 | 532 | |
|
534 | 533 | class IPythonInputSplitter(InputSplitter): |
|
535 | 534 | """An input splitter that recognizes all of IPython's special syntax.""" |
|
536 | 535 | |
|
537 | 536 | # String with raw, untransformed input. |
|
538 | 537 | source_raw = '' |
|
539 | 538 | |
|
540 | 539 | # Flag to track when a transformer has stored input that it hasn't given |
|
541 | 540 | # back yet. |
|
542 | 541 | transformer_accumulating = False |
|
543 | 542 | |
|
544 | 543 | # Flag to track when assemble_python_lines has stored input that it hasn't |
|
545 | 544 | # given back yet. |
|
546 | 545 | within_python_line = False |
|
547 | 546 | |
|
548 | 547 | # Private attributes |
|
549 | 548 | |
|
550 | 549 | # List with lines of raw input accumulated so far. |
|
551 | 550 | _buffer_raw = None |
|
552 | 551 | |
|
553 | 552 | def __init__(self, line_input_checker=True, physical_line_transforms=None, |
|
554 | 553 | logical_line_transforms=None, python_line_transforms=None): |
|
555 | 554 | super(IPythonInputSplitter, self).__init__() |
|
556 | 555 | self._buffer_raw = [] |
|
557 | 556 | self._validate = True |
|
558 | 557 | |
|
559 | 558 | if physical_line_transforms is not None: |
|
560 | 559 | self.physical_line_transforms = physical_line_transforms |
|
561 | 560 | else: |
|
562 | 561 | self.physical_line_transforms = [ |
|
563 | 562 | leading_indent(), |
|
564 | 563 | classic_prompt(), |
|
565 | 564 | ipy_prompt(), |
|
566 | 565 | cellmagic(end_on_blank_line=line_input_checker), |
|
567 | 566 | ] |
|
568 | 567 | |
|
569 | 568 | self.assemble_logical_lines = assemble_logical_lines() |
|
570 | 569 | if logical_line_transforms is not None: |
|
571 | 570 | self.logical_line_transforms = logical_line_transforms |
|
572 | 571 | else: |
|
573 | 572 | self.logical_line_transforms = [ |
|
574 | 573 | help_end(), |
|
575 | 574 | escaped_commands(), |
|
576 | 575 | assign_from_magic(), |
|
577 | 576 | assign_from_system(), |
|
578 | 577 | ] |
|
579 | 578 | |
|
580 | 579 | self.assemble_python_lines = assemble_python_lines() |
|
581 | 580 | if python_line_transforms is not None: |
|
582 | 581 | self.python_line_transforms = python_line_transforms |
|
583 | 582 | else: |
|
584 | 583 | # We don't use any of these at present |
|
585 | 584 | self.python_line_transforms = [] |
|
586 | 585 | |
|
587 | 586 | @property |
|
588 | 587 | def transforms(self): |
|
589 | 588 | "Quick access to all transformers." |
|
590 | 589 | return self.physical_line_transforms + \ |
|
591 | 590 | [self.assemble_logical_lines] + self.logical_line_transforms + \ |
|
592 | 591 | [self.assemble_python_lines] + self.python_line_transforms |
|
593 | 592 | |
|
594 | 593 | @property |
|
595 | 594 | def transforms_in_use(self): |
|
596 | 595 | """Transformers, excluding logical line transformers if we're in a |
|
597 | 596 | Python line.""" |
|
598 | 597 | t = self.physical_line_transforms[:] |
|
599 | 598 | if not self.within_python_line: |
|
600 | 599 | t += [self.assemble_logical_lines] + self.logical_line_transforms |
|
601 | 600 | return t + [self.assemble_python_lines] + self.python_line_transforms |
|
602 | 601 | |
|
603 | 602 | def reset(self): |
|
604 | 603 | """Reset the input buffer and associated state.""" |
|
605 | 604 | super(IPythonInputSplitter, self).reset() |
|
606 | 605 | self._buffer_raw[:] = [] |
|
607 | 606 | self.source_raw = '' |
|
608 | 607 | self.transformer_accumulating = False |
|
609 | 608 | self.within_python_line = False |
|
610 | 609 | |
|
611 | 610 | for t in self.transforms: |
|
612 | 611 | try: |
|
613 | 612 | t.reset() |
|
614 | 613 | except SyntaxError: |
|
615 | 614 | # Nothing that calls reset() expects to handle transformer |
|
616 | 615 | # errors |
|
617 | 616 | pass |
|
618 | 617 | |
|
619 | 618 | def flush_transformers(self): |
|
620 | 619 | def _flush(transform, outs): |
|
621 | 620 | """yield transformed lines |
|
622 | 621 | |
|
623 | 622 | always strings, never None |
|
624 | 623 | |
|
625 | 624 | transform: the current transform |
|
626 | 625 | outs: an iterable of previously transformed inputs. |
|
627 | 626 | Each may be multiline, which will be passed |
|
628 | 627 | one line at a time to transform. |
|
629 | 628 | """ |
|
630 | 629 | for out in outs: |
|
631 | 630 | for line in out.splitlines(): |
|
632 | 631 | # push one line at a time |
|
633 | 632 | tmp = transform.push(line) |
|
634 | 633 | if tmp is not None: |
|
635 | 634 | yield tmp |
|
636 | 635 | |
|
637 | 636 | # reset the transform |
|
638 | 637 | tmp = transform.reset() |
|
639 | 638 | if tmp is not None: |
|
640 | 639 | yield tmp |
|
641 | 640 | |
|
642 | 641 | out = [] |
|
643 | 642 | for t in self.transforms_in_use: |
|
644 | 643 | out = _flush(t, out) |
|
645 | 644 | |
|
646 | 645 | out = list(out) |
|
647 | 646 | if out: |
|
648 | 647 | self._store('\n'.join(out)) |
|
649 | 648 | |
|
650 | 649 | def raw_reset(self): |
|
651 | 650 | """Return raw input only and perform a full reset. |
|
652 | 651 | """ |
|
653 | 652 | out = self.source_raw |
|
654 | 653 | self.reset() |
|
655 | 654 | return out |
|
656 | 655 | |
|
657 | 656 | def source_reset(self): |
|
658 | 657 | try: |
|
659 | 658 | self.flush_transformers() |
|
660 | 659 | return self.source |
|
661 | 660 | finally: |
|
662 | 661 | self.reset() |
|
663 | 662 | |
|
664 | 663 | def push_accepts_more(self): |
|
665 | 664 | if self.transformer_accumulating: |
|
666 | 665 | return True |
|
667 | 666 | else: |
|
668 | 667 | return super(IPythonInputSplitter, self).push_accepts_more() |
|
669 | 668 | |
|
670 | 669 | def transform_cell(self, cell): |
|
671 | 670 | """Process and translate a cell of input. |
|
672 | 671 | """ |
|
673 | 672 | self.reset() |
|
674 | 673 | try: |
|
675 | 674 | self.push(cell) |
|
676 | 675 | self.flush_transformers() |
|
677 | 676 | return self.source |
|
678 | 677 | finally: |
|
679 | 678 | self.reset() |
|
680 | 679 | |
|
681 | 680 | def push(self, lines:str) -> bool: |
|
682 | 681 | """Push one or more lines of IPython input. |
|
683 | 682 | |
|
684 | 683 | This stores the given lines and returns a status code indicating |
|
685 | 684 | whether the code forms a complete Python block or not, after processing |
|
686 | 685 | all input lines for special IPython syntax. |
|
687 | 686 | |
|
688 | 687 | Any exceptions generated in compilation are swallowed, but if an |
|
689 | 688 | exception was produced, the method returns True. |
|
690 | 689 | |
|
691 | 690 | Parameters |
|
692 | 691 | ---------- |
|
693 | 692 | lines : string |
|
694 | 693 | One or more lines of Python input. |
|
695 | 694 | |
|
696 | 695 | Returns |
|
697 | 696 | ------- |
|
698 | 697 | is_complete : boolean |
|
699 | 698 | True if the current input source (the result of the current input |
|
700 | 699 | plus prior inputs) forms a complete Python execution block. Note that |
|
701 | 700 | this value is also stored as a private attribute (_is_complete), so it |
|
702 | 701 | can be queried at any time. |
|
703 | 702 | """ |
|
704 | 703 | assert isinstance(lines, str) |
|
705 | 704 | # We must ensure all input is pure unicode |
|
706 | 705 | # ''.splitlines() --> [], but we need to push the empty line to transformers |
|
707 | 706 | lines_list = lines.splitlines() |
|
708 | 707 | if not lines_list: |
|
709 | 708 | lines_list = [''] |
|
710 | 709 | |
|
711 | 710 | # Store raw source before applying any transformations to it. Note |
|
712 | 711 | # that this must be done *after* the reset() call that would otherwise |
|
713 | 712 | # flush the buffer. |
|
714 | 713 | self._store(lines, self._buffer_raw, 'source_raw') |
|
715 | 714 | |
|
716 | 715 | transformed_lines_list = [] |
|
717 | 716 | for line in lines_list: |
|
718 | 717 | transformed = self._transform_line(line) |
|
719 | 718 | if transformed is not None: |
|
720 | 719 | transformed_lines_list.append(transformed) |
|
721 | 720 | |
|
722 | 721 | if transformed_lines_list: |
|
723 | 722 | transformed_lines = '\n'.join(transformed_lines_list) |
|
724 | 723 | return super(IPythonInputSplitter, self).push(transformed_lines) |
|
725 | 724 | else: |
|
726 | 725 | # Got nothing back from transformers - they must be waiting for |
|
727 | 726 | # more input. |
|
728 | 727 | return False |
|
729 | 728 | |
|
730 | 729 | def _transform_line(self, line): |
|
731 | 730 | """Push a line of input code through the various transformers. |
|
732 | 731 | |
|
733 | 732 | Returns any output from the transformers, or None if a transformer |
|
734 | 733 | is accumulating lines. |
|
735 | 734 | |
|
736 | 735 | Sets self.transformer_accumulating as a side effect. |
|
737 | 736 | """ |
|
738 | 737 | def _accumulating(dbg): |
|
739 | 738 | #print(dbg) |
|
740 | 739 | self.transformer_accumulating = True |
|
741 | 740 | return None |
|
742 | 741 | |
|
743 | 742 | for transformer in self.physical_line_transforms: |
|
744 | 743 | line = transformer.push(line) |
|
745 | 744 | if line is None: |
|
746 | 745 | return _accumulating(transformer) |
|
747 | 746 | |
|
748 | 747 | if not self.within_python_line: |
|
749 | 748 | line = self.assemble_logical_lines.push(line) |
|
750 | 749 | if line is None: |
|
751 | 750 | return _accumulating('acc logical line') |
|
752 | 751 | |
|
753 | 752 | for transformer in self.logical_line_transforms: |
|
754 | 753 | line = transformer.push(line) |
|
755 | 754 | if line is None: |
|
756 | 755 | return _accumulating(transformer) |
|
757 | 756 | |
|
758 | 757 | line = self.assemble_python_lines.push(line) |
|
759 | 758 | if line is None: |
|
760 | 759 | self.within_python_line = True |
|
761 | 760 | return _accumulating('acc python line') |
|
762 | 761 | else: |
|
763 | 762 | self.within_python_line = False |
|
764 | 763 | |
|
765 | 764 | for transformer in self.python_line_transforms: |
|
766 | 765 | line = transformer.push(line) |
|
767 | 766 | if line is None: |
|
768 | 767 | return _accumulating(transformer) |
|
769 | 768 | |
|
770 | 769 | #print("transformers clear") #debug |
|
771 | 770 | self.transformer_accumulating = False |
|
772 | 771 | return line |
|
773 | 772 |
General Comments 0
You need to be logged in to leave comments.
Login now