Mercurial Repositories > eric / file comparison
comparison: Plugins/CheckerPlugins/CodeStyleChecker/pycodestyle.py
Plugins/CheckerPlugins/CodeStyleChecker/pycodestyle.py
- changeset 5147
- d39dd5cee0c8
- parent 5146
- f96c32abd120
- child 5288
- ff831e3e383e
equal
deleted
inserted
replaced
| 5146:f96c32abd120 | 5147:d39dd5cee0c8 |
|---|---|
| 1 #!/usr/bin/env python | |
| 2 # -*- coding: utf-8 -*- | |
| 3 | |
| 4 # pycodestyle.py - Check Python source code formatting, according to PEP 8 | |
| 5 # Copyright (C) 2006-2009 Johann C. Rocholl <johann@rocholl.net> | |
| 6 # Copyright (C) 2009-2014 Florent Xicluna <florent.xicluna@gmail.com> | |
| 7 # Copyright (C) 2014-2016 Ian Lee <ianlee1521@gmail.com> | |
| 8 # | |
| 9 # Permission is hereby granted, free of charge, to any person | |
| 10 # obtaining a copy of this software and associated documentation files | |
| 11 # (the "Software"), to deal in the Software without restriction, | |
| 12 # including without limitation the rights to use, copy, modify, merge, | |
| 13 # publish, distribute, sublicense, and/or sell copies of the Software, | |
| 14 # and to permit persons to whom the Software is furnished to do so, | |
| 15 # subject to the following conditions: | |
| 16 # | |
| 17 # The above copyright notice and this permission notice shall be | |
| 18 # included in all copies or substantial portions of the Software. | |
| 19 # | |
| 20 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, | |
| 21 # EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF | |
| 22 # MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND | |
| 23 # NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS | |
| 24 # BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN | |
| 25 # ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN | |
| 26 # CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE | |
| 27 # SOFTWARE. | |
| 28 | |
| 29 r""" | |
| 30 Check Python source code formatting, according to PEP 8. | |
| 31 | |
| 32 For usage and a list of options, try this: | |
| 33 $ python pycodestyle.py -h | |
| 34 | |
| 35 This program and its regression test suite live here: | |
| 36 https://github.com/pycqa/pycodestyle | |
| 37 | |
| 38 Groups of errors and warnings: | |
| 39 E errors | |
| 40 W warnings | |
| 41 100 indentation | |
| 42 200 whitespace | |
| 43 300 blank lines | |
| 44 400 imports | |
| 45 500 line length | |
| 46 600 deprecation | |
| 47 700 statements | |
| 48 900 syntax error | |
| 49 """ | |
| 50 from __future__ import with_statement | |
| 51 | |
| 52 # | |
| 53 # This is a modified version to make the original pycodestyle.py better suitable | |
| 54 # for being called from within the eric6 IDE. The modifications are as | |
| 55 # follows: | |
| 56 # | |
| 57 # - made messages translatable via Qt | |
| 58 # - added code for eric6 integration | |
| 59 # | |
| 60 # Copyright (c) 2011 - 2016 Detlev Offenbach <detlev@die-offenbachs.de> | |
| 61 # | |
| 62 | |
| 63 import inspect | |
| 64 import keyword | |
| 65 import os | |
| 66 import re | |
| 67 import sys | |
| 68 import time | |
| 69 import tokenize | |
| 70 ##import ast | |
| 71 from fnmatch import fnmatch | |
| 72 from optparse import OptionParser | |
| 73 try: | |
| 74 from configparser import RawConfigParser | |
| 75 from io import TextIOWrapper | |
| 76 except ImportError: | |
| 77 from ConfigParser import RawConfigParser # __IGNORE_WARNING__ | |
| 78 | |
| 79 __version__ = '2.1.0.dev0' | |
| 80 | |
| 81 DEFAULT_EXCLUDE = '.svn,CVS,.bzr,.hg,.git,__pycache__,.tox' | |
| 82 DEFAULT_IGNORE = 'E121,E123,E126,E226,E24,E704,W503' | |
| 83 try: | |
| 84 if sys.platform == 'win32': | |
| 85 USER_CONFIG = os.path.expanduser(r'~\.pycodestyle') | |
| 86 else: | |
| 87 USER_CONFIG = os.path.join( | |
| 88 os.getenv('XDG_CONFIG_HOME') or os.path.expanduser('~/.config'), | |
| 89 'pycodestyle' | |
| 90 ) | |
| 91 except ImportError: | |
| 92 USER_CONFIG = None | |
| 93 | |
| 94 PROJECT_CONFIG = ('setup.cfg', 'tox.ini') | |
| 95 TESTSUITE_PATH = os.path.join(os.path.dirname(__file__), 'testsuite') | |
| 96 MAX_LINE_LENGTH = 79 | |
| 97 REPORT_FORMAT = { | |
| 98 'default': '%(path)s:%(row)d:%(col)d: %(code)s %(text)s', | |
| 99 'pylint': '%(path)s:%(row)d: [%(code)s] %(text)s', | |
| 100 } | |
| 101 | |
| 102 PyCF_ONLY_AST = 1024 | |
| 103 SINGLETONS = frozenset(['False', 'None', 'True']) | |
| 104 KEYWORDS = frozenset(keyword.kwlist + ['print']) - SINGLETONS | |
| 105 UNARY_OPERATORS = frozenset(['>>', '**', '*', '+', '-']) | |
| 106 ARITHMETIC_OP = frozenset(['**', '*', '/', '//', '+', '-']) | |
| 107 WS_OPTIONAL_OPERATORS = ARITHMETIC_OP.union(['^', '&', '|', '<<', '>>', '%']) | |
| 108 WS_NEEDED_OPERATORS = frozenset([ | |
| 109 '**=', '*=', '/=', '//=', '+=', '-=', '!=', '<>', '<', '>', | |
| 110 '%=', '^=', '&=', '|=', '==', '<=', '>=', '<<=', '>>=', '=']) | |
| 111 WHITESPACE = frozenset(' \t') | |
| 112 NEWLINE = frozenset([tokenize.NL, tokenize.NEWLINE]) | |
| 113 SKIP_TOKENS = NEWLINE.union([tokenize.INDENT, tokenize.DEDENT]) | |
| 114 # ERRORTOKEN is triggered by backticks in Python 3 | |
| 115 SKIP_COMMENTS = SKIP_TOKENS.union([tokenize.COMMENT, tokenize.ERRORTOKEN]) | |
| 116 BENCHMARK_KEYS = ['directories', 'files', 'logical lines', 'physical lines'] | |
| 117 | |
| 118 INDENT_REGEX = re.compile(r'([ \t]*)') | |
| 119 RAISE_COMMA_REGEX = re.compile(r'raise\s+\w+\s*,') | |
| 120 RERAISE_COMMA_REGEX = re.compile(r'raise\s+\w+\s*,.*,\s*\w+\s*$') | |
| 121 ERRORCODE_REGEX = re.compile(r'\b[A-Z]\d{3}\b') | |
| 122 DOCSTRING_REGEX = re.compile(r'u?r?["\']') | |
| 123 EXTRANEOUS_WHITESPACE_REGEX = re.compile(r'[[({] | []}),;:]') | |
| 124 WHITESPACE_AFTER_COMMA_REGEX = re.compile(r'[,;:]\s*(?: |\t)') | |
| 125 COMPARE_SINGLETON_REGEX = re.compile(r'(\bNone|\bFalse|\bTrue)?\s*([=!]=)' | |
| 126 r'\s*(?(1)|(None|False|True))\b') | |
| 127 COMPARE_NEGATIVE_REGEX = re.compile(r'\b(not)\s+[^][)(}{ ]+\s+(in|is)\s') | |
| 128 COMPARE_TYPE_REGEX = re.compile(r'(?:[=!]=|is(?:\s+not)?)\s*type(?:s.\w+Type' | |
| 129 r'|\s*\(\s*([^)]*[^ )])\s*\))') | |
| 130 KEYWORD_REGEX = re.compile(r'(\s*)\b(?:%s)\b(\s*)' % r'|'.join(KEYWORDS)) | |
| 131 OPERATOR_REGEX = re.compile(r'(?:[^,\s])(\s*)(?:[-+*/|!<=>%&^]+)(\s*)') | |
| 132 LAMBDA_REGEX = re.compile(r'\blambda\b') | |
| 133 HUNK_REGEX = re.compile(r'^@@ -\d+(?:,\d+)? \+(\d+)(?:,(\d+))? @@.*$') | |
| 134 | |
| 135 # Work around Python < 2.6 behaviour, which does not generate NL after | |
| 136 # a comment which is on a line by itself. | |
| 137 COMMENT_WITH_NL = tokenize.generate_tokens(['#\n'].pop).send(None)[1] == '#\n' | |
| 138 | |
| 139 | |
| 140 ############################################################################## | |
| 141 # Plugins (check functions) for physical lines | |
| 142 ############################################################################## | |
| 143 | |
| 144 | |
| 145 def tabs_or_spaces(physical_line, indent_char): | |
| 146 r"""Never mix tabs and spaces. | |
| 147 | |
| 148 The most popular way of indenting Python is with spaces only. The | |
| 149 second-most popular way is with tabs only. Code indented with a mixture | |
| 150 of tabs and spaces should be converted to using spaces exclusively. When | |
| 151 invoking the Python command line interpreter with the -t option, it issues | |
| 152 warnings about code that illegally mixes tabs and spaces. When using -tt | |
| 153 these warnings become errors. These options are highly recommended! | |
| 154 | |
| 155 Okay: if a == 0:\n a = 1\n b = 1 | |
| 156 E101: if a == 0:\n a = 1\n\tb = 1 | |
| 157 """ | |
| 158 indent = INDENT_REGEX.match(physical_line).group(1) | |
| 159 for offset, char in enumerate(indent): | |
| 160 if char != indent_char: | |
| 161 return offset, "E101 indentation contains mixed spaces and tabs" | |
| 162 | |
| 163 | |
| 164 def tabs_obsolete(physical_line): | |
| 165 r"""For new projects, spaces-only are strongly recommended over tabs. | |
| 166 | |
| 167 Okay: if True:\n return | |
| 168 W191: if True:\n\treturn | |
| 169 """ | |
| 170 indent = INDENT_REGEX.match(physical_line).group(1) | |
| 171 if '\t' in indent: | |
| 172 return indent.index('\t'), "W191 indentation contains tabs" | |
| 173 | |
| 174 | |
| 175 def trailing_whitespace(physical_line): | |
| 176 r"""Trailing whitespace is superfluous. | |
| 177 | |
| 178 The warning returned varies on whether the line itself is blank, for easier | |
| 179 filtering for those who want to indent their blank lines. | |
| 180 | |
| 181 Okay: spam(1)\n# | |
| 182 W291: spam(1) \n# | |
| 183 W293: class Foo(object):\n \n bang = 12 | |
| 184 """ | |
| 185 physical_line = physical_line.rstrip('\n') # chr(10), newline | |
| 186 physical_line = physical_line.rstrip('\r') # chr(13), carriage return | |
| 187 physical_line = physical_line.rstrip('\x0c') # chr(12), form feed, ^L | |
| 188 stripped = physical_line.rstrip(' \t\v') | |
| 189 if physical_line != stripped: | |
| 190 if stripped: | |
| 191 return len(stripped), "W291 trailing whitespace" | |
| 192 else: | |
| 193 return 0, "W293 blank line contains whitespace" | |
| 194 | |
| 195 | |
| 196 def trailing_blank_lines(physical_line, lines, line_number, total_lines): | |
| 197 r"""Trailing blank lines are superfluous. | |
| 198 | |
| 199 Okay: spam(1) | |
| 200 W391: spam(1)\n | |
| 201 | |
| 202 However the last line should end with a new line (warning W292). | |
| 203 """ | |
| 204 if line_number == total_lines: | |
| 205 stripped_last_line = physical_line.rstrip() | |
| 206 if not stripped_last_line: | |
| 207 return 0, "W391 blank line at end of file" | |
| 208 if stripped_last_line == physical_line: | |
| 209 return len(physical_line), "W292 no newline at end of file" | |
| 210 | |
| 211 | |
| 212 def maximum_line_length(physical_line, max_line_length, multiline, noqa): | |
| 213 r"""Limit all lines to a maximum of 79 characters. | |
| 214 | |
| 215 There are still many devices around that are limited to 80 character | |
| 216 lines; plus, limiting windows to 80 characters makes it possible to have | |
| 217 several windows side-by-side. The default wrapping on such devices looks | |
| 218 ugly. Therefore, please limit all lines to a maximum of 79 characters. | |
| 219 For flowing long blocks of text (docstrings or comments), limiting the | |
| 220 length to 72 characters is recommended. | |
| 221 | |
| 222 Reports error E501. | |
| 223 """ | |
| 224 line = physical_line.rstrip() | |
| 225 length = len(line) | |
| 226 if length > max_line_length and not noqa: | |
| 227 # Special case for long URLs in multi-line docstrings or comments, | |
| 228 # but still report the error when the 72 first chars are whitespaces. | |
| 229 chunks = line.split() | |
| 230 if ((len(chunks) == 1 and multiline) or | |
| 231 (len(chunks) == 2 and chunks[0] == '#')) and \ | |
| 232 len(line) - len(chunks[-1]) < max_line_length - 7: | |
| 233 return | |
| 234 if hasattr(line, 'decode'): # Python 2 | |
| 235 # The line could contain multi-byte characters | |
| 236 try: | |
| 237 length = len(line.decode('utf-8')) | |
| 238 except UnicodeError: | |
| 239 pass | |
| 240 if length > max_line_length: | |
| 241 return (max_line_length, "E501 line too long " | |
| 242 "(%d > %d characters)" % (length, max_line_length), | |
| 243 length, max_line_length) | |
| 244 | |
| 245 | |
| 246 ############################################################################## | |
| 247 # Plugins (check functions) for logical lines | |
| 248 ############################################################################## | |
| 249 | |
| 250 | |
| 251 def blank_lines(logical_line, blank_lines, indent_level, line_number, | |
| 252 blank_before, previous_logical, | |
| 253 previous_unindented_logical_line, previous_indent_level, | |
| 254 lines): | |
| 255 r"""Separate top-level function and class definitions with two blank lines. | |
| 256 | |
| 257 Method definitions inside a class are separated by a single blank line. | |
| 258 | |
| 259 Extra blank lines may be used (sparingly) to separate groups of related | |
| 260 functions. Blank lines may be omitted between a bunch of related | |
| 261 one-liners (e.g. a set of dummy implementations). | |
| 262 | |
| 263 Use blank lines in functions, sparingly, to indicate logical sections. | |
| 264 | |
| 265 Okay: def a():\n pass\n\n\ndef b():\n pass | |
| 266 Okay: def a():\n pass\n\n\nasync def b():\n pass | |
| 267 Okay: def a():\n pass\n\n\n# Foo\n# Bar\n\ndef b():\n pass | |
| 268 | |
| 269 E301: class Foo:\n b = 0\n def bar():\n pass | |
| 270 E302: def a():\n pass\n\ndef b(n):\n pass | |
| 271 E302: def a():\n pass\n\nasync def b(n):\n pass | |
| 272 E303: def a():\n pass\n\n\n\ndef b(n):\n pass | |
| 273 E303: def a():\n\n\n\n pass | |
| 274 E304: @decorator\n\ndef a():\n pass | |
| 275 E305: def a():\n pass\na() | |
| 276 """ | |
| 277 if line_number < 3 and not previous_logical: | |
| 278 return # Don't expect blank lines before the first line | |
| 279 if previous_logical.startswith('@'): | |
| 280 if blank_lines: | |
| 281 yield 0, "E304 blank lines found after function decorator" | |
| 282 elif blank_lines > 2 or (indent_level and blank_lines == 2): | |
| 283 yield 0, "E303 too many blank lines (%d)", blank_lines | |
| 284 elif logical_line.startswith(('def ', 'async def ', 'class ', '@')): | |
| 285 if indent_level: | |
| 286 if not (blank_before or previous_indent_level < indent_level or | |
| 287 DOCSTRING_REGEX.match(previous_logical)): | |
| 288 ancestor_level = indent_level | |
| 289 nested = False | |
| 290 # Search backwards for a def ancestor or tree root (top level). | |
| 291 for line in lines[line_number - 2::-1]: | |
| 292 if line.strip() and expand_indent(line) < ancestor_level: | |
| 293 ancestor_level = expand_indent(line) | |
| 294 nested = line.lstrip().startswith('def ') | |
| 295 if nested or ancestor_level == 0: | |
| 296 break | |
| 297 if nested: | |
| 298 yield 0, "E306 expected 1 blank line before a " \ | |
| 299 "nested definition, found 0" | |
| 300 else: | |
| 301 yield 0, "E301 expected 1 blank line, found 0" | |
| 302 elif blank_before != 2: | |
| 303 yield 0, "E302 expected 2 blank lines, found %d", blank_before | |
| 304 elif (logical_line and not indent_level and blank_before != 2 and | |
| 305 previous_unindented_logical_line.startswith(('def', 'class'))): | |
| 306 yield 0, "E305 expected 2 blank lines after " \ | |
| 307 "class or function definition, found %d", blank_before | |
| 308 | |
| 309 | |
| 310 def extraneous_whitespace(logical_line): | |
| 311 r"""Avoid extraneous whitespace. | |
| 312 | |
| 313 Avoid extraneous whitespace in these situations: | |
| 314 - Immediately inside parentheses, brackets or braces. | |
| 315 - Immediately before a comma, semicolon, or colon. | |
| 316 | |
| 317 Okay: spam(ham[1], {eggs: 2}) | |
| 318 E201: spam( ham[1], {eggs: 2}) | |
| 319 E201: spam(ham[ 1], {eggs: 2}) | |
| 320 E201: spam(ham[1], { eggs: 2}) | |
| 321 E202: spam(ham[1], {eggs: 2} ) | |
| 322 E202: spam(ham[1 ], {eggs: 2}) | |
| 323 E202: spam(ham[1], {eggs: 2 }) | |
| 324 | |
| 325 E203: if x == 4: print x, y; x, y = y , x | |
| 326 E203: if x == 4: print x, y ; x, y = y, x | |
| 327 E203: if x == 4 : print x, y; x, y = y, x | |
| 328 """ | |
| 329 line = logical_line | |
| 330 for match in EXTRANEOUS_WHITESPACE_REGEX.finditer(line): | |
| 331 text = match.group() | |
| 332 char = text.strip() | |
| 333 found = match.start() | |
| 334 if text == char + ' ': | |
| 335 # assert char in '([{' | |
| 336 yield found + 1, "E201 whitespace after '%s'", char | |
| 337 elif line[found - 1] != ',': | |
| 338 code = ('E202' if char in '}])' else 'E203') # if char in ',;:' | |
| 339 yield found, "%s whitespace before '%s'" % (code, char), char | |
| 340 | |
| 341 | |
| 342 def whitespace_around_keywords(logical_line): | |
| 343 r"""Avoid extraneous whitespace around keywords. | |
| 344 | |
| 345 Okay: True and False | |
| 346 E271: True and False | |
| 347 E272: True and False | |
| 348 E273: True and\tFalse | |
| 349 E274: True\tand False | |
| 350 """ | |
| 351 for match in KEYWORD_REGEX.finditer(logical_line): | |
| 352 before, after = match.groups() | |
| 353 | |
| 354 if '\t' in before: | |
| 355 yield match.start(1), "E274 tab before keyword" | |
| 356 elif len(before) > 1: | |
| 357 yield match.start(1), "E272 multiple spaces before keyword" | |
| 358 | |
| 359 if '\t' in after: | |
| 360 yield match.start(2), "E273 tab after keyword" | |
| 361 elif len(after) > 1: | |
| 362 yield match.start(2), "E271 multiple spaces after keyword" | |
| 363 | |
| 364 | |
| 365 def missing_whitespace_after_import_keyword(logical_line): | |
| 366 r"""Multiple imports in form from x import (a, b, c) should have space | |
| 367 between import statement and parenthesised name list. | |
| 368 | |
| 369 Okay: from foo import (bar, baz) | |
| 370 E275: from foo import(bar, baz) | |
| 371 E275: from importable.module import(bar, baz) | |
| 372 """ | |
| 373 line = logical_line | |
| 374 indicator = ' import(' | |
| 375 if line.startswith('from '): | |
| 376 found = line.find(indicator) | |
| 377 if -1 < found: | |
| 378 pos = found + len(indicator) - 1 | |
| 379 yield pos, "E275 missing whitespace after keyword" | |
| 380 | |
| 381 | |
| 382 def missing_whitespace(logical_line): | |
| 383 r"""Each comma, semicolon or colon should be followed by whitespace. | |
| 384 | |
| 385 Okay: [a, b] | |
| 386 Okay: (3,) | |
| 387 Okay: a[1:4] | |
| 388 Okay: a[:4] | |
| 389 Okay: a[1:] | |
| 390 Okay: a[1:4:2] | |
| 391 E231: ['a','b'] | |
| 392 E231: foo(bar,baz) | |
| 393 E231: [{'a':'b'}] | |
| 394 """ | |
| 395 line = logical_line | |
| 396 for index in range(len(line) - 1): | |
| 397 char = line[index] | |
| 398 if char in ',;:' and line[index + 1] not in WHITESPACE: | |
| 399 before = line[:index] | |
| 400 if char == ':' and before.count('[') > before.count(']') and \ | |
| 401 before.rfind('{') < before.rfind('['): | |
| 402 continue # Slice syntax, no space required | |
| 403 if char == ',' and line[index + 1] == ')': | |
| 404 continue # Allow tuple with only one element: (3,) | |
| 405 yield index, "E231 missing whitespace after '%s'", char | |
| 406 | |
| 407 | |
| 408 def indentation(logical_line, previous_logical, indent_char, | |
| 409 indent_level, previous_indent_level): | |
| 410 r"""Use 4 spaces per indentation level. | |
| 411 | |
| 412 For really old code that you don't want to mess up, you can continue to | |
| 413 use 8-space tabs. | |
| 414 | |
| 415 Okay: a = 1 | |
| 416 Okay: if a == 0:\n a = 1 | |
| 417 E111: a = 1 | |
| 418 E114: # a = 1 | |
| 419 | |
| 420 Okay: for item in items:\n pass | |
| 421 E112: for item in items:\npass | |
| 422 E115: for item in items:\n# Hi\n pass | |
| 423 | |
| 424 Okay: a = 1\nb = 2 | |
| 425 E113: a = 1\n b = 2 | |
| 426 E116: a = 1\n # b = 2 | |
| 427 """ | |
| 428 c = 0 if logical_line else 3 | |
| 429 tmpl = "E11%d %s" if logical_line else "E11%d %s (comment)" | |
| 430 if indent_level % 4: | |
| 431 yield 0, tmpl % (1 + c, "indentation is not a multiple of four") | |
| 432 indent_expect = previous_logical.endswith(':') | |
| 433 if indent_expect and indent_level <= previous_indent_level: | |
| 434 yield 0, tmpl % (2 + c, "expected an indented block") | |
| 435 elif not indent_expect and indent_level > previous_indent_level: | |
| 436 yield 0, tmpl % (3 + c, "unexpected indentation") | |
| 437 | |
| 438 | |
| 439 def continued_indentation(logical_line, tokens, indent_level, hang_closing, | |
| 440 indent_char, noqa, verbose): | |
| 441 r"""Continuation lines indentation. | |
| 442 | |
| 443 Continuation lines should align wrapped elements either vertically | |
| 444 using Python's implicit line joining inside parentheses, brackets | |
| 445 and braces, or using a hanging indent. | |
| 446 | |
| 447 When using a hanging indent these considerations should be applied: | |
| 448 - there should be no arguments on the first line, and | |
| 449 - further indentation should be used to clearly distinguish itself as a | |
| 450 continuation line. | |
| 451 | |
| 452 Okay: a = (\n) | |
| 453 E123: a = (\n ) | |
| 454 | |
| 455 Okay: a = (\n 42) | |
| 456 E121: a = (\n 42) | |
| 457 E122: a = (\n42) | |
| 458 E123: a = (\n 42\n ) | |
| 459 E124: a = (24,\n 42\n) | |
| 460 E125: if (\n b):\n pass | |
| 461 E126: a = (\n 42) | |
| 462 E127: a = (24,\n 42) | |
| 463 E128: a = (24,\n 42) | |
| 464 E129: if (a or\n b):\n pass | |
| 465 E131: a = (\n 42\n 24) | |
| 466 """ | |
| 467 first_row = tokens[0][2][0] | |
| 468 nrows = 1 + tokens[-1][2][0] - first_row | |
| 469 if noqa or nrows == 1: | |
| 470 return | |
| 471 | |
| 472 # indent_next tells us whether the next block is indented; assuming | |
| 473 # that it is indented by 4 spaces, then we should not allow 4-space | |
| 474 # indents on the final continuation line; in turn, some other | |
| 475 # indents are allowed to have an extra 4 spaces. | |
| 476 indent_next = logical_line.endswith(':') | |
| 477 | |
| 478 row = depth = 0 | |
| 479 valid_hangs = (4,) if indent_char != '\t' else (4, 8) | |
| 480 # remember how many brackets were opened on each line | |
| 481 parens = [0] * nrows | |
| 482 # relative indents of physical lines | |
| 483 rel_indent = [0] * nrows | |
| 484 # for each depth, collect a list of opening rows | |
| 485 open_rows = [[0]] | |
| 486 # for each depth, memorize the hanging indentation | |
| 487 hangs = [None] | |
| 488 # visual indents | |
| 489 indent_chances = {} | |
| 490 last_indent = tokens[0][2] | |
| 491 visual_indent = None | |
| 492 last_token_multiline = False | |
| 493 # for each depth, memorize the visual indent column | |
| 494 indent = [last_indent[1]] | |
| 495 if verbose >= 3: | |
| 496 print(">>> " + tokens[0][4].rstrip()) | |
| 497 | |
| 498 for token_type, text, start, end, line in tokens: | |
| 499 | |
| 500 newline = row < start[0] - first_row | |
| 501 if newline: | |
| 502 row = start[0] - first_row | |
| 503 newline = not last_token_multiline and token_type not in NEWLINE | |
| 504 | |
| 505 if newline: | |
| 506 # this is the beginning of a continuation line. | |
| 507 last_indent = start | |
| 508 if verbose >= 3: | |
| 509 print("... " + line.rstrip()) | |
| 510 | |
| 511 # record the initial indent. | |
| 512 rel_indent[row] = expand_indent(line) - indent_level | |
| 513 | |
| 514 # identify closing bracket | |
| 515 close_bracket = (token_type == tokenize.OP and text in ']})') | |
| 516 | |
| 517 # is the indent relative to an opening bracket line? | |
| 518 for open_row in reversed(open_rows[depth]): | |
| 519 hang = rel_indent[row] - rel_indent[open_row] | |
| 520 hanging_indent = hang in valid_hangs | |
| 521 if hanging_indent: | |
| 522 break | |
| 523 if hangs[depth]: | |
| 524 hanging_indent = (hang == hangs[depth]) | |
| 525 # is there any chance of visual indent? | |
| 526 visual_indent = (not close_bracket and hang > 0 and | |
| 527 indent_chances.get(start[1])) | |
| 528 | |
| 529 if close_bracket and indent[depth]: | |
| 530 # closing bracket for visual indent | |
| 531 if start[1] != indent[depth]: | |
| 532 yield (start, "E124 closing bracket does not match " | |
| 533 "visual indentation") | |
| 534 elif close_bracket and not hang: | |
| 535 # closing bracket matches indentation of opening bracket's line | |
| 536 if hang_closing: | |
| 537 yield start, "E133 closing bracket is missing indentation" | |
| 538 elif indent[depth] and start[1] < indent[depth]: | |
| 539 if visual_indent is not True: | |
| 540 # visual indent is broken | |
| 541 yield (start, "E128 continuation line " | |
| 542 "under-indented for visual indent") | |
| 543 elif hanging_indent or (indent_next and rel_indent[row] == 8): | |
| 544 # hanging indent is verified | |
| 545 if close_bracket and not hang_closing: | |
| 546 yield (start, "E123 closing bracket does not match " | |
| 547 "indentation of opening bracket's line") | |
| 548 hangs[depth] = hang | |
| 549 elif visual_indent is True: | |
| 550 # visual indent is verified | |
| 551 indent[depth] = start[1] | |
| 552 elif visual_indent in (text, str): | |
| 553 # ignore token lined up with matching one from a previous line | |
| 554 pass | |
| 555 else: | |
| 556 # indent is broken | |
| 557 if hang <= 0: | |
| 558 error = "E122", "missing indentation or outdented" | |
| 559 elif indent[depth]: | |
| 560 error = "E127", "over-indented for visual indent" | |
| 561 elif not close_bracket and hangs[depth]: | |
| 562 error = "E131", "unaligned for hanging indent" | |
| 563 else: | |
| 564 hangs[depth] = hang | |
| 565 if hang > 4: | |
| 566 error = "E126", "over-indented for hanging indent" | |
| 567 else: | |
| 568 error = "E121", "under-indented for hanging indent" | |
| 569 yield start, "%s continuation line %s" % error | |
| 570 | |
| 571 # look for visual indenting | |
| 572 if (parens[row] and | |
| 573 token_type not in (tokenize.NL, tokenize.COMMENT) and | |
| 574 not indent[depth]): | |
| 575 indent[depth] = start[1] | |
| 576 indent_chances[start[1]] = True | |
| 577 if verbose >= 4: | |
| 578 print("bracket depth %s indent to %s" % (depth, start[1])) | |
| 579 # deal with implicit string concatenation | |
| 580 elif (token_type in (tokenize.STRING, tokenize.COMMENT) or | |
| 581 text in ('u', 'ur', 'b', 'br')): | |
| 582 indent_chances[start[1]] = str | |
| 583 # special case for the "if" statement because len("if (") == 4 | |
| 584 elif not indent_chances and not row and not depth and text == 'if': | |
| 585 indent_chances[end[1] + 1] = True | |
| 586 elif text == ':' and line[end[1]:].isspace(): | |
| 587 open_rows[depth].append(row) | |
| 588 | |
| 589 # keep track of bracket depth | |
| 590 if token_type == tokenize.OP: | |
| 591 if text in '([{': | |
| 592 depth += 1 | |
| 593 indent.append(0) | |
| 594 hangs.append(None) | |
| 595 if len(open_rows) == depth: | |
| 596 open_rows.append([]) | |
| 597 open_rows[depth].append(row) | |
| 598 parens[row] += 1 | |
| 599 if verbose >= 4: | |
| 600 print("bracket depth %s seen, col %s, visual min = %s" % | |
| 601 (depth, start[1], indent[depth])) | |
| 602 elif text in ')]}' and depth > 0: | |
| 603 # parent indents should not be more than this one | |
| 604 prev_indent = indent.pop() or last_indent[1] | |
| 605 hangs.pop() | |
| 606 for d in range(depth): | |
| 607 if indent[d] > prev_indent: | |
| 608 indent[d] = 0 | |
| 609 for ind in list(indent_chances): | |
| 610 if ind >= prev_indent: | |
| 611 del indent_chances[ind] | |
| 612 del open_rows[depth + 1:] | |
| 613 depth -= 1 | |
| 614 if depth: | |
| 615 indent_chances[indent[depth]] = True | |
| 616 for idx in range(row, -1, -1): | |
| 617 if parens[idx]: | |
| 618 parens[idx] -= 1 | |
| 619 break | |
| 620 assert len(indent) == depth + 1 | |
| 621 if start[1] not in indent_chances: | |
| 622 # allow lining up tokens | |
| 623 indent_chances[start[1]] = text | |
| 624 | |
| 625 last_token_multiline = (start[0] != end[0]) | |
| 626 if last_token_multiline: | |
| 627 rel_indent[end[0] - first_row] = rel_indent[row] | |
| 628 | |
| 629 if indent_next and expand_indent(line) == indent_level + 4: | |
| 630 pos = (start[0], indent[0] + 4) | |
| 631 if visual_indent: | |
| 632 code = "E129 visually indented line" | |
| 633 else: | |
| 634 code = "E125 continuation line" | |
| 635 yield pos, "%s with same indent as next logical line" % code | |
| 636 | |
| 637 | |
| 638 def whitespace_before_parameters(logical_line, tokens): | |
| 639 r"""Avoid extraneous whitespace. | |
| 640 | |
| 641 Avoid extraneous whitespace in the following situations: | |
| 642 - before the open parenthesis that starts the argument list of a | |
| 643 function call. | |
| 644 - before the open parenthesis that starts an indexing or slicing. | |
| 645 | |
| 646 Okay: spam(1) | |
| 647 E211: spam (1) | |
| 648 | |
| 649 Okay: dict['key'] = list[index] | |
| 650 E211: dict ['key'] = list[index] | |
| 651 E211: dict['key'] = list [index] | |
| 652 """ | |
| 653 prev_type, prev_text, __, prev_end, __ = tokens[0] | |
| 654 for index in range(1, len(tokens)): | |
| 655 token_type, text, start, end, __ = tokens[index] | |
| 656 if (token_type == tokenize.OP and | |
| 657 text in '([' and | |
| 658 start != prev_end and | |
| 659 (prev_type == tokenize.NAME or prev_text in '}])') and | |
| 660 # Syntax "class A (B):" is allowed, but avoid it | |
| 661 (index < 2 or tokens[index - 2][1] != 'class') and | |
| 662 # Allow "return (a.foo for a in range(5))" | |
| 663 not keyword.iskeyword(prev_text)): | |
| 664 yield prev_end, "E211 whitespace before '%s'", text | |
| 665 prev_type = token_type | |
| 666 prev_text = text | |
| 667 prev_end = end | |
| 668 | |
| 669 | |
| 670 def whitespace_around_operator(logical_line): | |
| 671 r"""Avoid extraneous whitespace around an operator. | |
| 672 | |
| 673 Okay: a = 12 + 3 | |
| 674 E221: a = 4 + 5 | |
| 675 E222: a = 4 + 5 | |
| 676 E223: a = 4\t+ 5 | |
| 677 E224: a = 4 +\t5 | |
| 678 """ | |
| 679 for match in OPERATOR_REGEX.finditer(logical_line): | |
| 680 before, after = match.groups() | |
| 681 | |
| 682 if '\t' in before: | |
| 683 yield match.start(1), "E223 tab before operator" | |
| 684 elif len(before) > 1: | |
| 685 yield match.start(1), "E221 multiple spaces before operator" | |
| 686 | |
| 687 if '\t' in after: | |
| 688 yield match.start(2), "E224 tab after operator" | |
| 689 elif len(after) > 1: | |
| 690 yield match.start(2), "E222 multiple spaces after operator" | |
| 691 | |
| 692 | |
| 693 def missing_whitespace_around_operator(logical_line, tokens): | |
| 694 r"""Surround operators with a single space on either side. | |
| 695 | |
| 696 - Always surround these binary operators with a single space on | |
| 697 either side: assignment (=), augmented assignment (+=, -= etc.), | |
| 698 comparisons (==, <, >, !=, <=, >=, in, not in, is, is not), | |
| 699 Booleans (and, or, not). | |
| 700 | |
| 701 - If operators with different priorities are used, consider adding | |
| 702 whitespace around the operators with the lowest priorities. | |
| 703 | |
| 704 Okay: i = i + 1 | |
| 705 Okay: submitted += 1 | |
| 706 Okay: x = x * 2 - 1 | |
| 707 Okay: hypot2 = x * x + y * y | |
| 708 Okay: c = (a + b) * (a - b) | |
| 709 Okay: foo(bar, key='word', *args, **kwargs) | |
| 710 Okay: alpha[:-i] | |
| 711 | |
| 712 E225: i=i+1 | |
| 713 E225: submitted +=1 | |
| 714 E225: x = x /2 - 1 | |
| 715 E225: z = x **y | |
| 716 E226: c = (a+b) * (a-b) | |
| 717 E226: hypot2 = x*x + y*y | |
| 718 E227: c = a|b | |
| 719 E228: msg = fmt%(errno, errmsg) | |
| 720 """ | |
| 721 parens = 0 | |
| 722 need_space = False | |
| 723 prev_type = tokenize.OP | |
| 724 prev_text = prev_end = None | |
| 725 for token_type, text, start, end, line in tokens: | |
| 726 if token_type in SKIP_COMMENTS: | |
| 727 continue | |
| 728 if text in ('(', 'lambda'): | |
| 729 parens += 1 | |
| 730 elif text == ')': | |
| 731 parens -= 1 | |
| 732 if need_space: | |
| 733 if start != prev_end: | |
| 734 # Found a (probably) needed space | |
| 735 if need_space is not True and not need_space[1]: | |
| 736 yield (need_space[0], | |
| 737 "E225 missing whitespace around operator") | |
| 738 need_space = False | |
| 739 elif text == '>' and prev_text in ('<', '-'): | |
| 740 # Tolerate the "<>" operator, even if running Python 3 | |
| 741 # Deal with Python 3's annotated return value "->" | |
| 742 pass | |
| 743 else: | |
| 744 if need_space is True or need_space[1]: | |
| 745 # A needed trailing space was not found | |
| 746 yield prev_end, "E225 missing whitespace around operator" | |
| 747 elif prev_text != '**': | |
| 748 code, optype = 'E226', 'arithmetic' | |
| 749 if prev_text == '%': | |
| 750 code, optype = 'E228', 'modulo' | |
| 751 elif prev_text not in ARITHMETIC_OP: | |
| 752 code, optype = 'E227', 'bitwise or shift' | |
| 753 yield (need_space[0], "%s missing whitespace " | |
| 754 "around %s operator" % (code, optype)) | |
| 755 need_space = False | |
| 756 elif token_type == tokenize.OP and prev_end is not None: | |
| 757 if text == '=' and parens: | |
| 758 # Allow keyword args or defaults: foo(bar=None). | |
| 759 pass | |
| 760 elif text in WS_NEEDED_OPERATORS: | |
| 761 need_space = True | |
| 762 elif text in UNARY_OPERATORS: | |
| 763 # Check if the operator is being used as a binary operator | |
| 764 # Allow unary operators: -123, -x, +1. | |
| 765 # Allow argument unpacking: foo(*args, **kwargs). | |
| 766 if (prev_text in '}])' if prev_type == tokenize.OP | |
| 767 else prev_text not in KEYWORDS): | |
| 768 need_space = None | |
| 769 elif text in WS_OPTIONAL_OPERATORS: | |
| 770 need_space = None | |
| 771 | |
| 772 if need_space is None: | |
| 773 # Surrounding space is optional, but ensure that | |
| 774 # trailing space matches opening space | |
| 775 need_space = (prev_end, start != prev_end) | |
| 776 elif need_space and start == prev_end: | |
| 777 # A needed opening space was not found | |
| 778 yield prev_end, "E225 missing whitespace around operator" | |
| 779 need_space = False | |
| 780 prev_type = token_type | |
| 781 prev_text = text | |
| 782 prev_end = end | |
| 783 | |
| 784 | |
| 785 def whitespace_around_comma(logical_line): | |
| 786 r"""Avoid extraneous whitespace after a comma or a colon. | |
| 787 | |
| 788 Note: these checks are disabled by default | |
| 789 | |
| 790 Okay: a = (1, 2) | |
| 791 E241: a = (1, 2) | |
| 792 E242: a = (1,\t2) | |
| 793 """ | |
| 794 line = logical_line | |
| 795 for m in WHITESPACE_AFTER_COMMA_REGEX.finditer(line): | |
| 796 found = m.start() + 1 | |
| 797 if '\t' in m.group(): | |
| 798 yield found, "E242 tab after '%s'", m.group()[0] | |
| 799 else: | |
| 800 yield found, "E241 multiple spaces after '%s'", m.group()[0] | |
| 801 | |
| 802 | |
| 803 def whitespace_around_named_parameter_equals(logical_line, tokens): | |
| 804 r"""Don't use spaces around the '=' sign in function arguments. | |
| 805 | |
| 806 Don't use spaces around the '=' sign when used to indicate a | |
| 807 keyword argument or a default parameter value. | |
| 808 | |
| 809 Okay: def complex(real, imag=0.0): | |
| 810 Okay: return magic(r=real, i=imag) | |
| 811 Okay: boolean(a == b) | |
| 812 Okay: boolean(a != b) | |
| 813 Okay: boolean(a <= b) | |
| 814 Okay: boolean(a >= b) | |
| 815 Okay: def foo(arg: int = 42): | |
| 816 Okay: async def foo(arg: int = 42): | |
| 817 | |
| 818 E251: def complex(real, imag = 0.0): | |
| 819 E251: return magic(r = real, i = imag) | |
| 820 """ | |
| 821 parens = 0 | |
| 822 no_space = False | |
| 823 prev_end = None | |
| 824 annotated_func_arg = False | |
| 825 in_def = logical_line.startswith(('def', 'async def')) | |
| 826 message = "E251 unexpected spaces around keyword / parameter equals" | |
| 827 for token_type, text, start, end, line in tokens: | |
| 828 if token_type == tokenize.NL: | |
| 829 continue | |
| 830 if no_space: | |
| 831 no_space = False | |
| 832 if start != prev_end: | |
| 833 yield (prev_end, message) | |
| 834 if token_type == tokenize.OP: | |
| 835 if text in '([': | |
| 836 parens += 1 | |
| 837 elif text in ')]': | |
| 838 parens -= 1 | |
| 839 elif in_def and text == ':' and parens == 1: | |
| 840 annotated_func_arg = True | |
| 841 elif parens and text == ',' and parens == 1: | |
| 842 annotated_func_arg = False | |
| 843 elif parens and text == '=' and not annotated_func_arg: | |
| 844 no_space = True | |
| 845 if start != prev_end: | |
| 846 yield (prev_end, message) | |
| 847 if not parens: | |
| 848 annotated_func_arg = False | |
| 849 | |
| 850 prev_end = end | |
| 851 | |
| 852 | |
| 853 def whitespace_before_comment(logical_line, tokens): | |
| 854 r"""Separate inline comments by at least two spaces. | |
| 855 | |
| 856 An inline comment is a comment on the same line as a statement. Inline | |
| 857 comments should be separated by at least two spaces from the statement. | |
| 858 They should start with a # and a single space. | |
| 859 | |
| 860 Each line of a block comment starts with a # and a single space | |
| 861 (unless it is indented text inside the comment). | |
| 862 | |
| 863 Okay: x = x + 1 # Increment x | |
| 864 Okay: x = x + 1 # Increment x | |
| 865 Okay: # Block comment | |
| 866 E261: x = x + 1 # Increment x | |
| 867 E262: x = x + 1 #Increment x | |
| 868 E262: x = x + 1 # Increment x | |
| 869 E265: #Block comment | |
| 870 E266: ### Block comment | |
| 871 """ | |
| 872 prev_end = (0, 0) | |
| 873 for token_type, text, start, end, line in tokens: | |
| 874 if token_type == tokenize.COMMENT: | |
| 875 inline_comment = line[:start[1]].strip() | |
| 876 if inline_comment: | |
| 877 if prev_end[0] == start[0] and start[1] < prev_end[1] + 2: | |
| 878 yield (prev_end, | |
| 879 "E261 at least two spaces before inline comment") | |
| 880 symbol, sp, comment = text.partition(' ') | |
| 881 bad_prefix = symbol not in '#:' and (symbol.lstrip('#')[:1] or '#') | |
| 882 if inline_comment: | |
| 883 if bad_prefix or comment[:1] in WHITESPACE: | |
| 884 yield start, "E262 inline comment should start with '# '" | |
| 885 elif bad_prefix and (bad_prefix != '!' or start[0] > 1): | |
| 886 if bad_prefix != '#': | |
| 887 yield start, "E265 block comment should start with '# '" | |
| 888 elif comment: | |
| 889 yield start, "E266 too many leading '#' for block comment" | |
| 890 elif token_type != tokenize.NL: | |
| 891 prev_end = end | |
| 892 | |
| 893 | |
| 894 def imports_on_separate_lines(logical_line): | |
| 895 r"""Place imports on separate lines. | |
| 896 | |
| 897 Okay: import os\nimport sys | |
| 898 E401: import sys, os | |
| 899 | |
| 900 Okay: from subprocess import Popen, PIPE | |
| 901 Okay: from myclas import MyClass | |
| 902 Okay: from foo.bar.yourclass import YourClass | |
| 903 Okay: import myclass | |
| 904 Okay: import foo.bar.yourclass | |
| 905 """ | |
| 906 line = logical_line | |
| 907 if line.startswith('import '): | |
| 908 found = line.find(',') | |
| 909 if -1 < found and ';' not in line[:found]: | |
| 910 yield found, "E401 multiple imports on one line" | |
| 911 | |
| 912 | |
| 913 def module_imports_on_top_of_file( | |
| 914 logical_line, indent_level, checker_state, noqa): | |
| 915 r"""Place imports at the top of the file. | |
| 916 | |
| 917 Always put imports at the top of the file, just after any module comments | |
| 918 and docstrings, and before module globals and constants. | |
| 919 | |
| 920 Okay: import os | |
| 921 Okay: # this is a comment\nimport os | |
| 922 Okay: '''this is a module docstring'''\nimport os | |
| 923 Okay: r'''this is a module docstring'''\nimport os | |
| 924 Okay: try:\n import x\nexcept:\n pass\nelse:\n pass\nimport y | |
| 925 Okay: try:\n import x\nexcept:\n pass\nfinally:\n pass\nimport y | |
| 926 E402: a=1\nimport os | |
| 927 E402: 'One string'\n"Two string"\nimport os | |
| 928 E402: a=1\nfrom sys import x | |
| 929 | |
| 930 Okay: if x:\n import os | |
| 931 """ | |
| 932 def is_string_literal(line): | |
| 933 if line[0] in 'uUbB': | |
| 934 line = line[1:] | |
| 935 if line and line[0] in 'rR': | |
| 936 line = line[1:] | |
| 937 return line and (line[0] == '"' or line[0] == "'") | |
| 938 | |
| 939 allowed_try_keywords = ('try', 'except', 'else', 'finally') | |
| 940 | |
| 941 if indent_level: # Allow imports in conditional statements or functions | |
| 942 return | |
| 943 if not logical_line: # Allow empty lines or comments | |
| 944 return | |
| 945 if noqa: | |
| 946 return | |
| 947 line = logical_line | |
| 948 if line.startswith('import ') or line.startswith('from '): | |
| 949 if checker_state.get('seen_non_imports', False): | |
| 950 yield 0, "E402 module level import not at top of file" | |
| 951 elif any(line.startswith(kw) for kw in allowed_try_keywords): | |
| 952 # Allow try, except, else, finally keywords intermixed with imports in | |
| 953 # order to support conditional importing | |
| 954 return | |
| 955 elif is_string_literal(line): | |
| 956 # The first literal is a docstring, allow it. Otherwise, report error. | |
| 957 if checker_state.get('seen_docstring', False): | |
| 958 checker_state['seen_non_imports'] = True | |
| 959 else: | |
| 960 checker_state['seen_docstring'] = True | |
| 961 else: | |
| 962 checker_state['seen_non_imports'] = True | |
| 963 | |
| 964 | |
| 965 def compound_statements(logical_line): | |
| 966 r"""Compound statements (on the same line) are generally discouraged. | |
| 967 | |
| 968 While sometimes it's okay to put an if/for/while with a small body | |
| 969 on the same line, never do this for multi-clause statements. | |
| 970 Also avoid folding such long lines! | |
| 971 | |
| 972 Always use a def statement instead of an assignment statement that | |
| 973 binds a lambda expression directly to a name. | |
| 974 | |
| 975 Okay: if foo == 'blah':\n do_blah_thing() | |
| 976 Okay: do_one() | |
| 977 Okay: do_two() | |
| 978 Okay: do_three() | |
| 979 | |
| 980 E701: if foo == 'blah': do_blah_thing() | |
| 981 E701: for x in lst: total += x | |
| 982 E701: while t < 10: t = delay() | |
| 983 E701: if foo == 'blah': do_blah_thing() | |
| 984 E701: else: do_non_blah_thing() | |
| 985 E701: try: something() | |
| 986 E701: finally: cleanup() | |
| 987 E701: if foo == 'blah': one(); two(); three() | |
| 988 E702: do_one(); do_two(); do_three() | |
| 989 E703: do_four(); # useless semicolon | |
| 990 E704: def f(x): return 2*x | |
| 991 E705: async def f(x): return 2*x | |
| 992 E731: f = lambda x: 2*x | |
| 993 """ | |
| 994 line = logical_line | |
| 995 last_char = len(line) - 1 | |
| 996 found = line.find(':') | |
| 997 prev_found = 0 | |
| 998 counts = dict((char, 0) for char in '{}[]()') | |
| 999 while -1 < found < last_char: | |
| 1000 update_counts(line[prev_found:found], counts) | |
| 1001 if ((counts['{'] <= counts['}'] and # {'a': 1} (dict) | |
| 1002 counts['['] <= counts[']'] and # [1:2] (slice) | |
| 1003 counts['('] <= counts[')'])): # (annotation) | |
| 1004 lambda_kw = LAMBDA_REGEX.search(line, 0, found) | |
| 1005 if lambda_kw: | |
| 1006 before = line[:lambda_kw.start()].rstrip() | |
| 1007 if before[-1:] == '=' and isidentifier(before[:-1].strip()): | |
| 1008 yield 0, ("E731 do not assign a lambda expression, use a " | |
| 1009 "def") | |
| 1010 break | |
| 1011 if line.startswith('def '): | |
| 1012 yield 0, "E704 multiple statements on one line (def)" | |
| 1013 elif line.startswith('async def '): | |
| 1014 yield 0, "E705 multiple statements on one line (async def)" | |
| 1015 else: | |
| 1016 yield found, "E701 multiple statements on one line (colon)" | |
| 1017 prev_found = found | |
| 1018 found = line.find(':', found + 1) | |
| 1019 found = line.find(';') | |
| 1020 while -1 < found: | |
| 1021 if found < last_char: | |
| 1022 yield found, "E702 multiple statements on one line (semicolon)" | |
| 1023 else: | |
| 1024 yield found, "E703 statement ends with a semicolon" | |
| 1025 found = line.find(';', found + 1) | |
| 1026 | |
| 1027 | |
| 1028 def explicit_line_join(logical_line, tokens): | |
| 1029 r"""Avoid explicit line join between brackets. | |
| 1030 | |
| 1031 The preferred way of wrapping long lines is by using Python's implied line | |
| 1032 continuation inside parentheses, brackets and braces. Long lines can be | |
| 1033 broken over multiple lines by wrapping expressions in parentheses. These | |
| 1034 should be used in preference to using a backslash for line continuation. | |
| 1035 | |
| 1036 E502: aaa = [123, \\n 123] | |
| 1037 E502: aaa = ("bbb " \\n "ccc") | |
| 1038 | |
| 1039 Okay: aaa = [123,\n 123] | |
| 1040 Okay: aaa = ("bbb "\n "ccc") | |
| 1041 Okay: aaa = "bbb " \\n "ccc" | |
| 1042 Okay: aaa = 123 # \\ | |
| 1043 """ | |
| 1044 prev_start = prev_end = parens = 0 | |
| 1045 comment = False | |
| 1046 backslash = None | |
| 1047 for token_type, text, start, end, line in tokens: | |
| 1048 if token_type == tokenize.COMMENT: | |
| 1049 comment = True | |
| 1050 if start[0] != prev_start and parens and backslash and not comment: | |
| 1051 yield backslash, "E502 the backslash is redundant between brackets" | |
| 1052 if end[0] != prev_end: | |
| 1053 if line.rstrip('\r\n').endswith('\\'): | |
| 1054 backslash = (end[0], len(line.splitlines()[-1]) - 1) | |
| 1055 else: | |
| 1056 backslash = None | |
| 1057 prev_start = prev_end = end[0] | |
| 1058 else: | |
| 1059 prev_start = start[0] | |
| 1060 if token_type == tokenize.OP: | |
| 1061 if text in '([{': | |
| 1062 parens += 1 | |
| 1063 elif text in ')]}': | |
| 1064 parens -= 1 | |
| 1065 | |
| 1066 | |
| 1067 def break_around_binary_operator(logical_line, tokens): | |
| 1068 r""" | |
| 1069 Avoid breaks before binary operators. | |
| 1070 | |
| 1071 The preferred place to break around a binary operator is after the | |
| 1072 operator, not before it. | |
| 1073 | |
| 1074 W503: (width == 0\n + height == 0) | |
| 1075 W503: (width == 0\n and height == 0) | |
| 1076 | |
| 1077 Okay: (width == 0 +\n height == 0) | |
| 1078 Okay: foo(\n -x) | |
| 1079 Okay: foo(x\n []) | |
| 1080 Okay: x = '''\n''' + '' | |
| 1081 Okay: foo(x,\n -y) | |
| 1082 Okay: foo(x, # comment\n -y) | |
| 1083 Okay: var = (1 &\n ~2) | |
| 1084 Okay: var = (1 /\n -2) | |
| 1085 Okay: var = (1 +\n -1 +\n -2) | |
| 1086 """ | |
| 1087 def is_binary_operator(token_type, text): | |
| 1088 # The % character is strictly speaking a binary operator, but the | |
| 1089 # common usage seems to be to put it next to the format parameters, | |
| 1090 # after a line break. | |
| 1091 return ((token_type == tokenize.OP or text in ['and', 'or']) and | |
| 1092 text not in "()[]{},:.;@=%~") | |
| 1093 | |
| 1094 line_break = False | |
| 1095 unary_context = True | |
| 1096 # Previous non-newline token types and text | |
| 1097 previous_token_type = None | |
| 1098 previous_text = None | |
| 1099 for token_type, text, start, end, line in tokens: | |
| 1100 if token_type == tokenize.COMMENT: | |
| 1101 continue | |
| 1102 if ('\n' in text or '\r' in text) and token_type != tokenize.STRING: | |
| 1103 line_break = True | |
| 1104 else: | |
| 1105 if (is_binary_operator(token_type, text) and line_break and | |
| 1106 not unary_context and | |
| 1107 not is_binary_operator(previous_token_type, | |
| 1108 previous_text)): | |
| 1109 yield start, "W503 line break before binary operator" | |
| 1110 unary_context = text in '([{,;' | |
| 1111 line_break = False | |
| 1112 previous_token_type = token_type | |
| 1113 previous_text = text | |
| 1114 | |
| 1115 | |
| 1116 def comparison_to_singleton(logical_line, noqa): | |
| 1117 r"""Comparison to singletons should use "is" or "is not". | |
| 1118 | |
| 1119 Comparisons to singletons like None should always be done | |
| 1120 with "is" or "is not", never the equality operators. | |
| 1121 | |
| 1122 Okay: if arg is not None: | |
| 1123 E711: if arg != None: | |
| 1124 E711: if None == arg: | |
| 1125 E712: if arg == True: | |
| 1126 E712: if False == arg: | |
| 1127 | |
| 1128 Also, beware of writing if x when you really mean if x is not None -- | |
| 1129 e.g. when testing whether a variable or argument that defaults to None was | |
| 1130 set to some other value. The other value might have a type (such as a | |
| 1131 container) that could be false in a boolean context! | |
| 1132 """ | |
| 1133 match = not noqa and COMPARE_SINGLETON_REGEX.search(logical_line) | |
| 1134 if match: | |
| 1135 singleton = match.group(1) or match.group(3) | |
| 1136 same = (match.group(2) == '==') | |
| 1137 | |
| 1138 msg = "'if cond is %s:'" % (('' if same else 'not ') + singleton) | |
| 1139 if singleton in ('None',): | |
| 1140 code = 'E711' | |
| 1141 else: | |
| 1142 code = 'E712' | |
| 1143 nonzero = ((singleton == 'True' and same) or | |
| 1144 (singleton == 'False' and not same)) | |
| 1145 msg += " or 'if %scond:'" % ('' if nonzero else 'not ') | |
| 1146 yield (match.start(2), ("%s comparison to %s should be %s" % | |
| 1147 (code, singleton, msg)), singleton, msg) | |
| 1148 | |
| 1149 | |
| 1150 def comparison_negative(logical_line): | |
| 1151 r"""Negative comparison should be done using "not in" and "is not". | |
| 1152 | |
| 1153 Okay: if x not in y:\n pass | |
| 1154 Okay: assert (X in Y or X is Z) | |
| 1155 Okay: if not (X in Y):\n pass | |
| 1156 Okay: zz = x is not y | |
| 1157 E713: Z = not X in Y | |
| 1158 E713: if not X.B in Y:\n pass | |
| 1159 E714: if not X is Y:\n pass | |
| 1160 E714: Z = not X.B is Y | |
| 1161 """ | |
| 1162 match = COMPARE_NEGATIVE_REGEX.search(logical_line) | |
| 1163 if match: | |
| 1164 pos = match.start(1) | |
| 1165 if match.group(2) == 'in': | |
| 1166 yield pos, "E713 test for membership should be 'not in'" | |
| 1167 else: | |
| 1168 yield pos, "E714 test for object identity should be 'is not'" | |
| 1169 | |
| 1170 | |
| 1171 def comparison_type(logical_line, noqa): | |
| 1172 r"""Object type comparisons should always use isinstance(). | |
| 1173 | |
| 1174 Do not compare types directly. | |
| 1175 | |
| 1176 Okay: if isinstance(obj, int): | |
| 1177 E721: if type(obj) is type(1): | |
| 1178 | |
| 1179 When checking if an object is a string, keep in mind that it might be a | |
| 1180 unicode string too! In Python 2.3, str and unicode have a common base | |
| 1181 class, basestring, so you can do: | |
| 1182 | |
| 1183 Okay: if isinstance(obj, basestring): | |
| 1184 Okay: if type(a1) is type(b1): | |
| 1185 """ | |
| 1186 match = COMPARE_TYPE_REGEX.search(logical_line) | |
| 1187 if match and not noqa: | |
| 1188 inst = match.group(1) | |
| 1189 if inst and isidentifier(inst) and inst not in SINGLETONS: | |
| 1190 return # Allow comparison for types which are not obvious | |
| 1191 yield match.start(), "E721 do not compare types, use 'isinstance()'" | |
| 1192 | |
| 1193 | |
| 1194 def ambiguous_identifier(logical_line, tokens): | |
| 1195 r"""Never use the characters 'l', 'O', or 'I' as variable names. | |
| 1196 | |
| 1197 In some fonts, these characters are indistinguishable from the numerals | |
| 1198 one and zero. When tempted to use 'l', use 'L' instead. | |
| 1199 | |
| 1200 Okay: L = 0 | |
| 1201 Okay: o = 123 | |
| 1202 Okay: i = 42 | |
| 1203 E741: l = 0 | |
| 1204 E741: O = 123 | |
| 1205 E741: I = 42 | |
| 1206 | |
| 1207 Variables can be bound in several other contexts, including class and | |
| 1208 function definitions, 'global' and 'nonlocal' statements, exception | |
| 1209 handlers, and 'with' statements. | |
| 1210 | |
| 1211 Okay: except AttributeError as o: | |
| 1212 Okay: with lock as L: | |
| 1213 E741: except AttributeError as O: | |
| 1214 E741: with lock as l: | |
| 1215 E741: global I | |
| 1216 E741: nonlocal l | |
| 1217 E742: class I(object): | |
| 1218 E743: def l(x): | |
| 1219 """ | |
| 1220 idents_to_avoid = ('l', 'O', 'I') | |
| 1221 prev_type, prev_text, prev_start, prev_end, __ = tokens[0] | |
| 1222 for token_type, text, start, end, line in tokens[1:]: | |
| 1223 ident = pos = None | |
| 1224 # identifiers on the lhs of an assignment operator | |
| 1225 if token_type == tokenize.OP and '=' in text: | |
| 1226 if prev_text in idents_to_avoid: | |
| 1227 ident = prev_text | |
| 1228 pos = prev_start | |
| 1229 # identifiers bound to a value with 'as', 'global', or 'nonlocal' | |
| 1230 if prev_text in ('as', 'global', 'nonlocal'): | |
| 1231 if text in idents_to_avoid: | |
| 1232 ident = text | |
| 1233 pos = start | |
| 1234 if prev_text == 'class': | |
| 1235 if text in idents_to_avoid: | |
| 1236 yield start, "E742 ambiguous class definition '%s'", text | |
| 1237 if prev_text == 'def': | |
| 1238 if text in idents_to_avoid: | |
| 1239 yield start, "E743 ambiguous function definition '%s'", text | |
| 1240 if ident: | |
| 1241 yield pos, "E741 ambiguous variable name '%s'", ident | |
| 1242 prev_text = text | |
| 1243 prev_start = start | |
| 1244 | |
| 1245 | |
| 1246 def python_3000_has_key(logical_line, noqa): | |
| 1247 r"""The {}.has_key() method is removed in Python 3: use the 'in' operator. | |
| 1248 | |
| 1249 Okay: if "alph" in d:\n print d["alph"] | |
| 1250 W601: assert d.has_key('alph') | |
| 1251 """ | |
| 1252 pos = logical_line.find('.has_key(') | |
| 1253 if pos > -1 and not noqa: | |
| 1254 yield pos, "W601 .has_key() is deprecated, use 'in'" | |
| 1255 | |
| 1256 | |
| 1257 def python_3000_raise_comma(logical_line): | |
| 1258 r"""When raising an exception, use "raise ValueError('message')". | |
| 1259 | |
| 1260 The older form is removed in Python 3. | |
| 1261 | |
| 1262 Okay: raise DummyError("Message") | |
| 1263 W602: raise DummyError, "Message" | |
| 1264 """ | |
| 1265 match = RAISE_COMMA_REGEX.match(logical_line) | |
| 1266 if match and not RERAISE_COMMA_REGEX.match(logical_line): | |
| 1267 yield match.end() - 1, "W602 deprecated form of raising exception" | |
| 1268 | |
| 1269 | |
| 1270 def python_3000_not_equal(logical_line): | |
| 1271 r"""New code should always use != instead of <>. | |
| 1272 | |
| 1273 The older syntax is removed in Python 3. | |
| 1274 | |
| 1275 Okay: if a != 'no': | |
| 1276 W603: if a <> 'no': | |
| 1277 """ | |
| 1278 pos = logical_line.find('<>') | |
| 1279 if pos > -1: | |
| 1280 yield pos, "W603 '<>' is deprecated, use '!='" | |
| 1281 | |
| 1282 | |
| 1283 def python_3000_backticks(logical_line): | |
| 1284 r"""Use repr() instead of backticks in Python 3. | |
| 1285 | |
| 1286 Okay: val = repr(1 + 2) | |
| 1287 W604: val = `1 + 2` | |
| 1288 """ | |
| 1289 pos = logical_line.find('`') | |
| 1290 if pos > -1: | |
| 1291 yield pos, "W604 backticks are deprecated, use 'repr()'" | |
| 1292 | |
| 1293 | |
| 1294 ############################################################################## | |
| 1295 # Helper functions | |
| 1296 ############################################################################## | |
| 1297 | |
| 1298 | |
| 1299 if sys.version_info < (3,): | |
| 1300 # Python 2: implicit encoding. | |
| 1301 def readlines(filename): | |
| 1302 """Read the source code.""" | |
| 1303 with open(filename, 'rU') as f: | |
| 1304 return f.readlines() | |
| 1305 isidentifier = re.compile(r'[a-zA-Z_]\w*$').match | |
| 1306 stdin_get_value = sys.stdin.read | |
| 1307 else: | |
| 1308 # Python 3 | |
| 1309 def readlines(filename): | |
| 1310 """Read the source code.""" | |
| 1311 try: | |
| 1312 with open(filename, 'rb') as f: | |
| 1313 (coding, lines) = tokenize.detect_encoding(f.readline) | |
| 1314 f = TextIOWrapper(f, coding, line_buffering=True) | |
| 1315 return [line.decode(coding) for line in lines] + f.readlines() | |
| 1316 except (LookupError, SyntaxError, UnicodeError): | |
| 1317 # Fall back if file encoding is improperly declared | |
| 1318 with open(filename, encoding='latin-1') as f: | |
| 1319 return f.readlines() | |
| 1320 isidentifier = str.isidentifier | |
| 1321 | |
| 1322 def stdin_get_value(): | |
| 1323 """Read the value from stdin.""" | |
| 1324 return TextIOWrapper(sys.stdin.buffer, errors='ignore').read() | |
| 1325 | |
| 1326 noqa = re.compile(r'# no(?:qa|pep8)\b', re.I).search | |
| 1327 | |
| 1328 | |
| 1329 def expand_indent(line): | |
| 1330 r"""Return the amount of indentation. | |
| 1331 | |
| 1332 Tabs are expanded to the next multiple of 8. | |
| 1333 | |
| 1334 >>> expand_indent(' ') | |
| 1335 4 | |
| 1336 >>> expand_indent('\t') | |
| 1337 8 | |
| 1338 >>> expand_indent(' \t') | |
| 1339 8 | |
| 1340 >>> expand_indent(' \t') | |
| 1341 16 | |
| 1342 """ | |
| 1343 if '\t' not in line: | |
| 1344 return len(line) - len(line.lstrip()) | |
| 1345 result = 0 | |
| 1346 for char in line: | |
| 1347 if char == '\t': | |
| 1348 result = result // 8 * 8 + 8 | |
| 1349 elif char == ' ': | |
| 1350 result += 1 | |
| 1351 else: | |
| 1352 break | |
| 1353 return result | |
| 1354 | |
| 1355 | |
| 1356 def mute_string(text): | |
| 1357 """Replace contents with 'xxx' to prevent syntax matching. | |
| 1358 | |
| 1359 >>> mute_string('"abc"') | |
| 1360 '"xxx"' | |
| 1361 >>> mute_string("'''abc'''") | |
| 1362 "'''xxx'''" | |
| 1363 >>> mute_string("r'abc'") | |
| 1364 "r'xxx'" | |
| 1365 """ | |
| 1366 # String modifiers (e.g. u or r) | |
| 1367 start = text.index(text[-1]) + 1 | |
| 1368 end = len(text) - 1 | |
| 1369 # Triple quotes | |
| 1370 if text[-3:] in ('"""', "'''"): | |
| 1371 start += 2 | |
| 1372 end -= 2 | |
| 1373 return text[:start] + 'x' * (end - start) + text[end:] | |
| 1374 | |
| 1375 | |
| 1376 def parse_udiff(diff, patterns=None, parent='.'): | |
| 1377 """Return a dictionary of matching lines.""" | |
| 1378 # For each file of the diff, the entry key is the filename, | |
| 1379 # and the value is a set of row numbers to consider. | |
| 1380 rv = {} | |
| 1381 path = nrows = None | |
| 1382 for line in diff.splitlines(): | |
| 1383 if nrows: | |
| 1384 if line[:1] != '-': | |
| 1385 nrows -= 1 | |
| 1386 continue | |
| 1387 if line[:3] == '@@ ': | |
| 1388 hunk_match = HUNK_REGEX.match(line) | |
| 1389 (row, nrows) = [int(g or '1') for g in hunk_match.groups()] | |
| 1390 rv[path].update(range(row, row + nrows)) | |
| 1391 elif line[:3] == '+++': | |
| 1392 path = line[4:].split('\t', 1)[0] | |
| 1393 if path[:2] == 'b/': | |
| 1394 path = path[2:] | |
| 1395 rv[path] = set() | |
| 1396 return dict([(os.path.join(parent, path), rows) | |
| 1397 for (path, rows) in rv.items() | |
| 1398 if rows and filename_match(path, patterns)]) | |
| 1399 | |
| 1400 | |
| 1401 def normalize_paths(value, parent=os.curdir): | |
| 1402 """Parse a comma-separated list of paths. | |
| 1403 | |
| 1404 Return a list of absolute paths. | |
| 1405 """ | |
| 1406 if not value: | |
| 1407 return [] | |
| 1408 if isinstance(value, list): | |
| 1409 return value | |
| 1410 paths = [] | |
| 1411 for path in value.split(','): | |
| 1412 path = path.strip() | |
| 1413 if '/' in path: | |
| 1414 path = os.path.abspath(os.path.join(parent, path)) | |
| 1415 paths.append(path.rstrip('/')) | |
| 1416 return paths | |
| 1417 | |
| 1418 | |
| 1419 def filename_match(filename, patterns, default=True): | |
| 1420 """Check if patterns contains a pattern that matches filename. | |
| 1421 | |
| 1422 If patterns is unspecified, this always returns True. | |
| 1423 """ | |
| 1424 if not patterns: | |
| 1425 return default | |
| 1426 return any(fnmatch(filename, pattern) for pattern in patterns) | |
| 1427 | |
| 1428 | |
| 1429 def update_counts(s, counts): | |
| 1430 r"""Adds one to the counts of each appearance of characters in s, | |
| 1431 for characters in counts""" | |
| 1432 for char in s: | |
| 1433 if char in counts: | |
| 1434 counts[char] += 1 | |
| 1435 | |
| 1436 | |
| 1437 def _is_eol_token(token): | |
| 1438 return token[0] in NEWLINE or token[4][token[3][1]:].lstrip() == '\\\n' | |
| 1439 | |
| 1440 | |
| 1441 if COMMENT_WITH_NL: | |
| 1442 def _is_eol_token(token, _eol_token=_is_eol_token): | |
| 1443 return _eol_token(token) or (token[0] == tokenize.COMMENT and | |
| 1444 token[1] == token[4]) | |
| 1445 | |
| 1446 ############################################################################## | |
| 1447 # Framework to run all checks | |
| 1448 ############################################################################## | |
| 1449 | |
| 1450 | |
| 1451 _checks = {'physical_line': {}, 'logical_line': {}, 'tree': {}} | |
| 1452 | |
| 1453 | |
| 1454 def _get_parameters(function): | |
| 1455 if sys.version_info >= (3, 3): | |
| 1456 return [parameter.name | |
| 1457 for parameter | |
| 1458 in inspect.signature(function).parameters.values() | |
| 1459 if parameter.kind == parameter.POSITIONAL_OR_KEYWORD] | |
| 1460 else: | |
| 1461 return inspect.getargspec(function)[0] | |
| 1462 | |
| 1463 | |
| 1464 def register_check(check, codes=None): | |
| 1465 """Register a new check object.""" | |
| 1466 def _add_check(check, kind, codes, args): | |
| 1467 if check in _checks[kind]: | |
| 1468 _checks[kind][check][0].extend(codes or []) | |
| 1469 else: | |
| 1470 _checks[kind][check] = (codes or [''], args) | |
| 1471 if inspect.isfunction(check): | |
| 1472 args = _get_parameters(check) | |
| 1473 if args and args[0] in ('physical_line', 'logical_line'): | |
| 1474 if codes is None: | |
| 1475 codes = ERRORCODE_REGEX.findall(check.__doc__ or '') | |
| 1476 _add_check(check, args[0], codes, args) | |
| 1477 elif inspect.isclass(check): | |
| 1478 if _get_parameters(check.__init__)[:2] == ['self', 'tree']: | |
| 1479 _add_check(check, 'tree', codes, None) | |
| 1480 | |
| 1481 | |
| 1482 def init_checks_registry(): | |
| 1483 """Register all globally visible functions. | |
| 1484 | |
| 1485 The first argument name is either 'physical_line' or 'logical_line'. | |
| 1486 """ | |
| 1487 mod = inspect.getmodule(register_check) | |
| 1488 for (name, function) in inspect.getmembers(mod, inspect.isfunction): | |
| 1489 register_check(function) | |
| 1490 | |
| 1491 | |
| 1492 init_checks_registry() | |
| 1493 | |
| 1494 | |
| 1495 class Checker(object): | |
| 1496 """Load a Python source file, tokenize it, check coding style.""" | |
| 1497 | |
| 1498 def __init__(self, filename=None, lines=None, | |
| 1499 options=None, report=None, **kwargs): | |
| 1500 if options is None: | |
| 1501 options = StyleGuide(kwargs).options | |
| 1502 else: | |
| 1503 assert not kwargs | |
| 1504 self._io_error = None | |
| 1505 self._physical_checks = options.physical_checks | |
| 1506 self._logical_checks = options.logical_checks | |
| 1507 self._ast_checks = options.ast_checks | |
| 1508 self.max_line_length = options.max_line_length | |
| 1509 self.multiline = False # in a multiline string? | |
| 1510 self.hang_closing = options.hang_closing | |
| 1511 self.verbose = options.verbose | |
| 1512 self.filename = filename | |
| 1513 # Dictionary where a checker can store its custom state. | |
| 1514 self._checker_states = {} | |
| 1515 if filename is None: | |
| 1516 self.filename = 'stdin' | |
| 1517 self.lines = lines or [] | |
| 1518 elif filename == '-': | |
| 1519 self.filename = 'stdin' | |
| 1520 self.lines = stdin_get_value().splitlines(True) | |
| 1521 elif lines is None: | |
| 1522 try: | |
| 1523 self.lines = readlines(filename) | |
| 1524 except IOError: | |
| 1525 (exc_type, exc) = sys.exc_info()[:2] | |
| 1526 self._io_error = '%s: %s' % (exc_type.__name__, exc) | |
| 1527 self.lines = [] | |
| 1528 else: | |
| 1529 self.lines = lines | |
| 1530 if self.lines: | |
| 1531 ord0 = ord(self.lines[0][0]) | |
| 1532 if ord0 in (0xef, 0xfeff): # Strip the UTF-8 BOM | |
| 1533 if ord0 == 0xfeff: | |
| 1534 self.lines[0] = self.lines[0][1:] | |
| 1535 elif self.lines[0][:3] == '\xef\xbb\xbf': | |
| 1536 self.lines[0] = self.lines[0][3:] | |
| 1537 self.report = report or options.report | |
| 1538 self.report_error = self.report.error | |
| 1539 self.report_error_args = self.report.error_args | |
| 1540 self.noqa = False | |
| 1541 | |
| 1542 # added for eric6 integration | |
| 1543 self.options = options | |
| 1544 | |
| 1545 def report_invalid_syntax(self): | |
| 1546 """Check if the syntax is valid.""" | |
| 1547 (exc_type, exc) = sys.exc_info()[:2] | |
| 1548 if len(exc.args) > 1: | |
| 1549 offset = exc.args[1] | |
| 1550 if len(offset) > 2: | |
| 1551 offset = offset[1:3] | |
| 1552 else: | |
| 1553 offset = (1, 0) | |
| 1554 self.report_error_args(offset[0], offset[1] or 0, | |
| 1555 'E901', self.report_invalid_syntax, | |
| 1556 exc_type.__name__, exc.args[0]) | |
| 1557 | |
| 1558 def readline(self): | |
| 1559 """Get the next line from the input buffer.""" | |
| 1560 if self.line_number >= self.total_lines: | |
| 1561 return '' | |
| 1562 line = self.lines[self.line_number] | |
| 1563 self.line_number += 1 | |
| 1564 if self.indent_char is None and line[:1] in WHITESPACE: | |
| 1565 self.indent_char = line[0] | |
| 1566 return line | |
| 1567 | |
| 1568 def run_check(self, check, argument_names): | |
| 1569 """Run a check plugin.""" | |
| 1570 arguments = [] | |
| 1571 for name in argument_names: | |
| 1572 arguments.append(getattr(self, name)) | |
| 1573 return check(*arguments) | |
| 1574 | |
| 1575 def init_checker_state(self, name, argument_names): | |
| 1576 """ Prepare custom state for the specific checker plugin.""" | |
| 1577 if 'checker_state' in argument_names: | |
| 1578 self.checker_state = self._checker_states.setdefault(name, {}) | |
| 1579 | |
| 1580 def check_physical(self, line): | |
| 1581 """Run all physical checks on a raw input line.""" | |
| 1582 self.physical_line = line | |
| 1583 for name, check, argument_names in self._physical_checks: | |
| 1584 self.init_checker_state(name, argument_names) | |
| 1585 result = self.run_check(check, argument_names) | |
| 1586 if result is not None: | |
| 1587 (offset, text) = result[:2] | |
| 1588 args = result[2:] | |
| 1589 self.report_error_args( | |
| 1590 self.line_number, offset, text, check, *args) | |
| 1591 if text[:4] == 'E101': | |
| 1592 self.indent_char = line[0] | |
| 1593 | |
| 1594 def build_tokens_line(self): | |
| 1595 """Build a logical line from tokens.""" | |
| 1596 logical = [] | |
| 1597 comments = [] | |
| 1598 length = 0 | |
| 1599 prev_row = prev_col = mapping = None | |
| 1600 for token_type, text, start, end, line in self.tokens: | |
| 1601 if token_type in SKIP_TOKENS: | |
| 1602 continue | |
| 1603 if not mapping: | |
| 1604 mapping = [(0, start)] | |
| 1605 if token_type == tokenize.COMMENT: | |
| 1606 comments.append(text) | |
| 1607 continue | |
| 1608 if token_type == tokenize.STRING: | |
| 1609 text = mute_string(text) | |
| 1610 if prev_row: | |
| 1611 (start_row, start_col) = start | |
| 1612 if prev_row != start_row: # different row | |
| 1613 prev_text = self.lines[prev_row - 1][prev_col - 1] | |
| 1614 if prev_text == ',' or (prev_text not in '{[(' and | |
| 1615 text not in '}])'): | |
| 1616 text = ' ' + text | |
| 1617 elif prev_col != start_col: # different column | |
| 1618 text = line[prev_col:start_col] + text | |
| 1619 logical.append(text) | |
| 1620 length += len(text) | |
| 1621 mapping.append((length, end)) | |
| 1622 (prev_row, prev_col) = end | |
| 1623 self.logical_line = ''.join(logical) | |
| 1624 self.noqa = comments and noqa(''.join(comments)) | |
| 1625 return mapping | |
| 1626 | |
| 1627 def check_logical(self): | |
| 1628 """Build a line from tokens and run all logical checks on it.""" | |
| 1629 self.report.increment_logical_line() | |
| 1630 mapping = self.build_tokens_line() | |
| 1631 | |
| 1632 if not mapping: | |
| 1633 return | |
| 1634 | |
| 1635 (start_row, start_col) = mapping[0][1] | |
| 1636 start_line = self.lines[start_row - 1] | |
| 1637 self.indent_level = expand_indent(start_line[:start_col]) | |
| 1638 if self.blank_before < self.blank_lines: | |
| 1639 self.blank_before = self.blank_lines | |
| 1640 if self.verbose >= 2: | |
| 1641 print(self.logical_line[:80].rstrip()) | |
| 1642 for name, check, argument_names in self._logical_checks: | |
| 1643 if self.verbose >= 4: | |
| 1644 print(' ' + name) | |
| 1645 self.init_checker_state(name, argument_names) | |
| 1646 for result in self.run_check(check, argument_names) or (): | |
| 1647 offset, text = result[:2] | |
| 1648 args = result[2:] | |
| 1649 if not isinstance(offset, tuple): | |
| 1650 for token_offset, pos in mapping: | |
| 1651 if offset <= token_offset: | |
| 1652 break | |
| 1653 offset = (pos[0], pos[1] + offset - token_offset) | |
| 1654 self.report_error_args( | |
| 1655 offset[0], offset[1], text, check, *args) | |
| 1656 if self.logical_line: | |
| 1657 self.previous_indent_level = self.indent_level | |
| 1658 self.previous_logical = self.logical_line | |
| 1659 if not self.indent_level: | |
| 1660 self.previous_unindented_logical_line = self.logical_line | |
| 1661 self.blank_lines = 0 | |
| 1662 self.tokens = [] | |
| 1663 | |
| 1664 def check_ast(self): | |
| 1665 """Build the file's AST and run all AST checks.""" | |
| 1666 try: | |
| 1667 tree = compile(''.join(self.lines), '', 'exec', PyCF_ONLY_AST) | |
| 1668 except (ValueError, SyntaxError, TypeError): | |
| 1669 return self.report_invalid_syntax() | |
| 1670 for name, cls, __ in self._ast_checks: | |
| 1671 # extended API for eric6 integration | |
| 1672 checker = cls(tree, self.filename, self.options) | |
| 1673 for args in checker.run(): | |
| 1674 lineno = args[0] | |
| 1675 if not self.lines or not noqa(self.lines[lineno - 1]): | |
| 1676 self.report_error_args(lineno, *args[1:]) | |
| 1677 | |
| 1678 def generate_tokens(self): | |
| 1679 """Tokenize the file, run physical line checks and yield tokens.""" | |
| 1680 if self._io_error: | |
| 1681 self.report_error_args(1, 0, 'E902', self._io_error, readlines) | |
| 1682 tokengen = tokenize.generate_tokens(self.readline) | |
| 1683 try: | |
| 1684 for token in tokengen: | |
| 1685 if token[2][0] > self.total_lines: | |
| 1686 return | |
| 1687 self.noqa = token[4] and noqa(token[4]) | |
| 1688 self.maybe_check_physical(token) | |
| 1689 yield token | |
| 1690 except (SyntaxError, tokenize.TokenError): | |
| 1691 self.report_invalid_syntax() | |
| 1692 | |
| 1693 def maybe_check_physical(self, token): | |
| 1694 """If appropriate (based on token), check current physical line(s).""" | |
| 1695 # Called after every token, but act only on end of line. | |
| 1696 if _is_eol_token(token): | |
| 1697 # Obviously, a newline token ends a single physical line. | |
| 1698 self.check_physical(token[4]) | |
| 1699 elif token[0] == tokenize.STRING and '\n' in token[1]: | |
| 1700 # Less obviously, a string that contains newlines is a | |
| 1701 # multiline string, either triple-quoted or with internal | |
| 1702 # newlines backslash-escaped. Check every physical line in the | |
| 1703 # string *except* for the last one: its newline is outside of | |
| 1704 # the multiline string, so we consider it a regular physical | |
| 1705 # line, and will check it like any other physical line. | |
| 1706 # | |
| 1707 # Subtleties: | |
| 1708 # - we don't *completely* ignore the last line; if it contains | |
| 1709 # the magical "# noqa" comment, we disable all physical | |
| 1710 # checks for the entire multiline string | |
| 1711 # - have to wind self.line_number back because initially it | |
| 1712 # points to the last line of the string, and we want | |
| 1713 # check_physical() to give accurate feedback | |
| 1714 if noqa(token[4]): | |
| 1715 return | |
| 1716 self.multiline = True | |
| 1717 self.line_number = token[2][0] | |
| 1718 for line in token[1].split('\n')[:-1]: | |
| 1719 self.check_physical(line + '\n') | |
| 1720 self.line_number += 1 | |
| 1721 self.multiline = False | |
| 1722 | |
| 1723 def check_all(self, expected=None, line_offset=0): | |
| 1724 """Run all checks on the input file.""" | |
| 1725 self.report.init_file(self.filename, self.lines, expected, line_offset) | |
| 1726 self.total_lines = len(self.lines) | |
| 1727 if self._ast_checks: | |
| 1728 self.check_ast() | |
| 1729 self.line_number = 0 | |
| 1730 self.indent_char = None | |
| 1731 self.indent_level = self.previous_indent_level = 0 | |
| 1732 self.previous_logical = '' | |
| 1733 self.previous_unindented_logical_line = '' | |
| 1734 self.tokens = [] | |
| 1735 self.blank_lines = self.blank_before = 0 | |
| 1736 parens = 0 | |
| 1737 for token in self.generate_tokens(): | |
| 1738 self.tokens.append(token) | |
| 1739 token_type, text = token[0:2] | |
| 1740 if self.verbose >= 3: | |
| 1741 if token[2][0] == token[3][0]: | |
| 1742 pos = '[%s:%s]' % (token[2][1] or '', token[3][1]) | |
| 1743 else: | |
| 1744 pos = 'l.%s' % token[3][0] | |
| 1745 print('l.%s\t%s\t%s\t%r' % | |
| 1746 (token[2][0], pos, tokenize.tok_name[token[0]], text)) | |
| 1747 if token_type == tokenize.OP: | |
| 1748 if text in '([{': | |
| 1749 parens += 1 | |
| 1750 elif text in '}])': | |
| 1751 parens -= 1 | |
| 1752 elif not parens: | |
| 1753 if token_type in NEWLINE: | |
| 1754 if token_type == tokenize.NEWLINE: | |
| 1755 self.check_logical() | |
| 1756 self.blank_before = 0 | |
| 1757 elif len(self.tokens) == 1: | |
| 1758 # The physical line contains only this token. | |
| 1759 self.blank_lines += 1 | |
| 1760 del self.tokens[0] | |
| 1761 else: | |
| 1762 self.check_logical() | |
| 1763 elif COMMENT_WITH_NL and token_type == tokenize.COMMENT: | |
| 1764 if len(self.tokens) == 1: | |
| 1765 # The comment also ends a physical line | |
| 1766 token = list(token) | |
| 1767 token[1] = text.rstrip('\r\n') | |
| 1768 token[3] = (token[2][0], token[2][1] + len(token[1])) | |
| 1769 self.tokens = [tuple(token)] | |
| 1770 self.check_logical() | |
| 1771 if self.tokens: | |
| 1772 self.check_physical(self.lines[-1]) | |
| 1773 self.check_logical() | |
| 1774 return self.report.get_file_results() | |
| 1775 | |
| 1776 | |
| 1777 class BaseReport(object): | |
| 1778 """Collect the results of the checks.""" | |
| 1779 | |
| 1780 print_filename = False | |
| 1781 | |
| 1782 def __init__(self, options): | |
| 1783 self._benchmark_keys = options.benchmark_keys | |
| 1784 self._ignore_code = options.ignore_code | |
| 1785 # Results | |
| 1786 self.elapsed = 0 | |
| 1787 self.total_errors = 0 | |
| 1788 self.counters = dict.fromkeys(self._benchmark_keys, 0) | |
| 1789 self.messages = {} | |
| 1790 | |
| 1791 def start(self): | |
| 1792 """Start the timer.""" | |
| 1793 self._start_time = time.time() | |
| 1794 | |
| 1795 def stop(self): | |
| 1796 """Stop the timer.""" | |
| 1797 self.elapsed = time.time() - self._start_time | |
| 1798 | |
| 1799 def init_file(self, filename, lines, expected, line_offset): | |
| 1800 """Signal a new file.""" | |
| 1801 self.filename = filename | |
| 1802 self.lines = lines | |
| 1803 self.expected = expected or () | |
| 1804 self.line_offset = line_offset | |
| 1805 self.file_errors = 0 | |
| 1806 self.counters['files'] += 1 | |
| 1807 self.counters['physical lines'] += len(lines) | |
| 1808 | |
| 1809 def increment_logical_line(self): | |
| 1810 """Signal a new logical line.""" | |
| 1811 self.counters['logical lines'] += 1 | |
| 1812 | |
| 1813 def error(self, line_number, offset, text, check): | |
| 1814 """Report an error, according to options.""" | |
| 1815 code = text[:4] | |
| 1816 if self._ignore_code(code): | |
| 1817 return | |
| 1818 if code in self.counters: | |
| 1819 self.counters[code] += 1 | |
| 1820 else: | |
| 1821 self.counters[code] = 1 | |
| 1822 self.messages[code] = [] | |
| 1823 # Don't care about expected errors or warnings | |
| 1824 if code in self.expected: | |
| 1825 return | |
| 1826 if self.print_filename and not self.file_errors: | |
| 1827 print(self.filename) | |
| 1828 self.file_errors += 1 | |
| 1829 self.total_errors += 1 | |
| 1830 return code | |
| 1831 | |
| 1832 def error_args(self, line_number, offset, text, check, *args): | |
| 1833 """Report an error, according to options.""" | |
| 1834 code = text[:4] | |
| 1835 if self._ignore_code(code): | |
| 1836 return | |
| 1837 if code in self.counters: | |
| 1838 self.counters[code] += 1 | |
| 1839 else: | |
| 1840 self.counters[code] = 1 | |
| 1841 self.messages[code] = args | |
| 1842 # Don't care about expected errors or warnings | |
| 1843 if code in self.expected: | |
| 1844 return | |
| 1845 if self.print_filename and not self.file_errors: | |
| 1846 print(self.filename) | |
| 1847 self.file_errors += 1 | |
| 1848 self.total_errors += 1 | |
| 1849 return code | |
| 1850 | |
| 1851 def get_file_results(self): | |
| 1852 """Return the count of errors and warnings for this file.""" | |
| 1853 return self.file_errors | |
| 1854 | |
| 1855 def get_count(self, prefix=''): | |
| 1856 """Return the total count of errors and warnings.""" | |
| 1857 return sum([self.counters[key] | |
| 1858 for key in self.messages if key.startswith(prefix)]) | |
| 1859 | |
| 1860 def get_statistics(self, prefix=''): | |
| 1861 """Get statistics for message codes that start with the prefix. | |
| 1862 | |
| 1863 prefix='' matches all errors and warnings | |
| 1864 prefix='E' matches all errors | |
| 1865 prefix='W' matches all warnings | |
| 1866 prefix='E4' matches all errors that have to do with imports | |
| 1867 """ | |
| 1868 return ['%-7s %s %s' % (self.counters[key], key, self.messages[key]) | |
| 1869 for key in sorted(self.messages) if key.startswith(prefix)] | |
| 1870 | |
| 1871 def print_statistics(self, prefix=''): | |
| 1872 """Print overall statistics (number of errors and warnings).""" | |
| 1873 for line in self.get_statistics(prefix): | |
| 1874 print(line) | |
| 1875 | |
| 1876 def print_benchmark(self): | |
| 1877 """Print benchmark numbers.""" | |
| 1878 print('%-7.2f %s' % (self.elapsed, 'seconds elapsed')) | |
| 1879 if self.elapsed: | |
| 1880 for key in self._benchmark_keys: | |
| 1881 print('%-7d %s per second (%d total)' % | |
| 1882 (self.counters[key] / self.elapsed, key, | |
| 1883 self.counters[key])) | |
| 1884 | |
| 1885 | |
| 1886 class FileReport(BaseReport): | |
| 1887 """Collect the results of the checks and print only the filenames.""" | |
| 1888 print_filename = True | |
| 1889 | |
| 1890 | |
| 1891 class StandardReport(BaseReport): | |
| 1892 """Collect and print the results of the checks.""" | |
| 1893 | |
| 1894 def __init__(self, options): | |
| 1895 super(StandardReport, self).__init__(options) | |
| 1896 self._fmt = REPORT_FORMAT.get(options.format.lower(), | |
| 1897 options.format) | |
| 1898 self._repeat = options.repeat | |
| 1899 self._show_source = options.show_source | |
| 1900 self._show_pep8 = options.show_pep8 | |
| 1901 | |
| 1902 def init_file(self, filename, lines, expected, line_offset): | |
| 1903 """Signal a new file.""" | |
| 1904 self._deferred_print = [] | |
| 1905 return super(StandardReport, self).init_file( | |
| 1906 filename, lines, expected, line_offset) | |
| 1907 | |
| 1908 def error(self, line_number, offset, text, check): | |
| 1909 """Report an error, according to options.""" | |
| 1910 code = super(StandardReport, self).error(line_number, offset, | |
| 1911 text, check) | |
| 1912 if code and (self.counters[code] == 1 or self._repeat): | |
| 1913 self._deferred_print.append( | |
| 1914 (line_number, offset, code, text[5:], check.__doc__)) | |
| 1915 return code | |
| 1916 | |
| 1917 def error_args(self, line_number, offset, code, check, *args): | |
| 1918 """Report an error, according to options.""" | |
| 1919 code = super(StandardReport, self).error_args(line_number, offset, | |
| 1920 code, check, *args) | |
| 1921 if code and (self.counters[code] == 1 or self._repeat): | |
| 1922 self._deferred_print.append( | |
| 1923 (line_number, offset, code, args, check.__doc__)) | |
| 1924 return code | |
| 1925 | |
| 1926 def get_file_results(self): | |
| 1927 """Print the result and return the overall count for this file.""" | |
| 1928 self._deferred_print.sort() | |
| 1929 for line_number, offset, code, text, doc in self._deferred_print: | |
| 1930 print(self._fmt % { | |
| 1931 'path': self.filename, | |
| 1932 'row': self.line_offset + line_number, 'col': offset + 1, | |
| 1933 'code': code, 'text': text, | |
| 1934 }) | |
| 1935 if self._show_source: | |
| 1936 if line_number > len(self.lines): | |
| 1937 line = '' | |
| 1938 else: | |
| 1939 line = self.lines[line_number - 1] | |
| 1940 print(line.rstrip()) | |
| 1941 print(re.sub(r'\S', ' ', line[:offset]) + '^') | |
| 1942 if self._show_pep8 and doc: | |
| 1943 print(' ' + doc.strip()) | |
| 1944 | |
| 1945 # stdout is block buffered when not stdout.isatty(). | |
| 1946 # line can be broken where buffer boundary since other processes | |
| 1947 # write to same file. | |
| 1948 # flush() after print() to avoid buffer boundary. | |
| 1949 # Typical buffer size is 8192. line written safely when | |
| 1950 # len(line) < 8192. | |
| 1951 sys.stdout.flush() | |
| 1952 return self.file_errors | |
| 1953 | |
| 1954 | |
| 1955 class DiffReport(StandardReport): | |
| 1956 """Collect and print the results for the changed lines only.""" | |
| 1957 | |
| 1958 def __init__(self, options): | |
| 1959 super(DiffReport, self).__init__(options) | |
| 1960 self._selected = options.selected_lines | |
| 1961 | |
| 1962 def error(self, line_number, offset, text, check): | |
| 1963 if line_number not in self._selected[self.filename]: | |
| 1964 return | |
| 1965 return super(DiffReport, self).error(line_number, offset, text, check) | |
| 1966 | |
| 1967 | |
| 1968 class StyleGuide(object): | |
| 1969 """Initialize a PEP-8 instance with few options.""" | |
| 1970 | |
| 1971 def __init__(self, *args, **kwargs): | |
| 1972 # build options from the command line | |
| 1973 self.checker_class = kwargs.pop('checker_class', Checker) | |
| 1974 parse_argv = kwargs.pop('parse_argv', False) | |
| 1975 config_file = kwargs.pop('config_file', False) | |
| 1976 parser = kwargs.pop('parser', None) | |
| 1977 # build options from dict | |
| 1978 options_dict = dict(*args, **kwargs) | |
| 1979 arglist = None if parse_argv else options_dict.get('paths', None) | |
| 1980 options, self.paths = process_options( | |
| 1981 arglist, parse_argv, config_file, parser) | |
| 1982 if options_dict: | |
| 1983 options.__dict__.update(options_dict) | |
| 1984 if 'paths' in options_dict: | |
| 1985 self.paths = options_dict['paths'] | |
| 1986 | |
| 1987 self.runner = self.input_file | |
| 1988 self.options = options | |
| 1989 | |
| 1990 if not options.reporter: | |
| 1991 options.reporter = BaseReport if options.quiet else StandardReport | |
| 1992 | |
| 1993 options.select = tuple(options.select or ()) | |
| 1994 # if not (options.select or options.ignore or | |
| 1995 # options.testsuite or options.doctest) and DEFAULT_IGNORE: | |
| 1996 # # The default choice: ignore controversial checks | |
| 1997 # options.ignore = tuple(DEFAULT_IGNORE.split(',')) | |
| 1998 # else: | |
| 1999 # # Ignore all checks which are not explicitly selected or all if no | |
| 2000 # options.ignore = ('',) if options.select else tuple(options.ignore) | |
| 2001 | |
| 2002 # check is ignored or explicitly selected | |
| 2003 options.ignore = ('',) if options.select else tuple(options.ignore) | |
| 2004 options.benchmark_keys = BENCHMARK_KEYS[:] | |
| 2005 options.ignore_code = self.ignore_code | |
| 2006 options.physical_checks = self.get_checks('physical_line') | |
| 2007 options.logical_checks = self.get_checks('logical_line') | |
| 2008 options.ast_checks = self.get_checks('tree') | |
| 2009 self.init_report() | |
| 2010 | |
| 2011 def init_report(self, reporter=None): | |
| 2012 """Initialize the report instance.""" | |
| 2013 self.options.report = (reporter or self.options.reporter)(self.options) | |
| 2014 return self.options.report | |
| 2015 | |
| 2016 def check_files(self, paths=None): | |
| 2017 """Run all checks on the paths.""" | |
| 2018 if paths is None: | |
| 2019 paths = self.paths | |
| 2020 report = self.options.report | |
| 2021 runner = self.runner | |
| 2022 report.start() | |
| 2023 try: | |
| 2024 for path in paths: | |
| 2025 if os.path.isdir(path): | |
| 2026 self.input_dir(path) | |
| 2027 elif not self.excluded(path): | |
| 2028 runner(path) | |
| 2029 except KeyboardInterrupt: | |
| 2030 print('... stopped') | |
| 2031 report.stop() | |
| 2032 return report | |
| 2033 | |
| 2034 def input_file(self, filename, lines=None, expected=None, line_offset=0): | |
| 2035 """Run all checks on a Python source file.""" | |
| 2036 if self.options.verbose: | |
| 2037 print('checking %s' % filename) | |
| 2038 fchecker = self.checker_class( | |
| 2039 filename, lines=lines, options=self.options) | |
| 2040 return fchecker.check_all(expected=expected, line_offset=line_offset) | |
| 2041 | |
| 2042 def input_dir(self, dirname): | |
| 2043 """Check all files in this directory and all subdirectories.""" | |
| 2044 dirname = dirname.rstrip('/') | |
| 2045 if self.excluded(dirname): | |
| 2046 return 0 | |
| 2047 counters = self.options.report.counters | |
| 2048 verbose = self.options.verbose | |
| 2049 filepatterns = self.options.filename | |
| 2050 runner = self.runner | |
| 2051 for root, dirs, files in os.walk(dirname): | |
| 2052 if verbose: | |
| 2053 print('directory ' + root) | |
| 2054 counters['directories'] += 1 | |
| 2055 for subdir in sorted(dirs): | |
| 2056 if self.excluded(subdir, root): | |
| 2057 dirs.remove(subdir) | |
| 2058 for filename in sorted(files): | |
| 2059 # contain a pattern that matches? | |
| 2060 if ((filename_match(filename, filepatterns) and | |
| 2061 not self.excluded(filename, root))): | |
| 2062 runner(os.path.join(root, filename)) | |
| 2063 | |
| 2064 def excluded(self, filename, parent=None): | |
| 2065 """Check if the file should be excluded. | |
| 2066 | |
| 2067 Check if 'options.exclude' contains a pattern that matches filename. | |
| 2068 """ | |
| 2069 if not self.options.exclude: | |
| 2070 return False | |
| 2071 basename = os.path.basename(filename) | |
| 2072 if filename_match(basename, self.options.exclude): | |
| 2073 return True | |
| 2074 if parent: | |
| 2075 filename = os.path.join(parent, filename) | |
| 2076 filename = os.path.abspath(filename) | |
| 2077 return filename_match(filename, self.options.exclude) | |
| 2078 | |
| 2079 def ignore_code(self, code): | |
| 2080 """Check if the error code should be ignored. | |
| 2081 | |
| 2082 If 'options.select' contains a prefix of the error code, | |
| 2083 return False. Else, if 'options.ignore' contains a prefix of | |
| 2084 the error code, return True. | |
| 2085 """ | |
| 2086 if len(code) < 4 and any(s.startswith(code) | |
| 2087 for s in self.options.select): | |
| 2088 return False | |
| 2089 return (code.startswith(self.options.ignore) and | |
| 2090 not code.startswith(self.options.select)) | |
| 2091 | |
| 2092 def get_checks(self, argument_name): | |
| 2093 """Get all the checks for this category. | |
| 2094 | |
| 2095 Find all globally visible functions where the first argument name | |
| 2096 starts with argument_name and which contain selected tests. | |
| 2097 """ | |
| 2098 checks = [] | |
| 2099 for check, attrs in _checks[argument_name].items(): | |
| 2100 (codes, args) = attrs | |
| 2101 if any(not (code and self.ignore_code(code)) for code in codes): | |
| 2102 checks.append((check.__name__, check, args)) | |
| 2103 return sorted(checks) | |
| 2104 | |
| 2105 | |
| 2106 def get_parser(prog='pycodestyle', version=__version__): | |
| 2107 """Create the parser for the program.""" | |
| 2108 parser = OptionParser(prog=prog, version=version, | |
| 2109 usage="%prog [options] input ...") | |
| 2110 parser.config_options = [ | |
| 2111 'exclude', 'filename', 'select', 'ignore', 'max-line-length', | |
| 2112 'hang-closing', 'count', 'format', 'quiet', 'show-pep8', | |
| 2113 'show-source', 'statistics', 'verbose'] | |
| 2114 parser.add_option('-v', '--verbose', default=0, action='count', | |
| 2115 help="print status messages, or debug with -vv") | |
| 2116 parser.add_option('-q', '--quiet', default=0, action='count', | |
| 2117 help="report only file names, or nothing with -qq") | |
| 2118 parser.add_option('-r', '--repeat', default=True, action='store_true', | |
| 2119 help="(obsolete) show all occurrences of the same error") | |
| 2120 parser.add_option('--first', action='store_false', dest='repeat', | |
| 2121 help="show first occurrence of each error") | |
| 2122 parser.add_option('--exclude', metavar='patterns', default=DEFAULT_EXCLUDE, | |
| 2123 help="exclude files or directories which match these " | |
| 2124 "comma separated patterns (default: %default)") | |
| 2125 parser.add_option('--filename', metavar='patterns', default='*.py', | |
| 2126 help="when parsing directories, only check filenames " | |
| 2127 "matching these comma separated patterns " | |
| 2128 "(default: %default)") | |
| 2129 parser.add_option('--select', metavar='errors', default='', | |
| 2130 help="select errors and warnings (e.g. E,W6)") | |
| 2131 parser.add_option('--ignore', metavar='errors', default='', | |
| 2132 help="skip errors and warnings (e.g. E4,W) " | |
| 2133 "(default: %s)" % DEFAULT_IGNORE) | |
| 2134 parser.add_option('--show-source', action='store_true', | |
| 2135 help="show source code for each error") | |
| 2136 parser.add_option('--show-pep8', action='store_true', | |
| 2137 help="show text of PEP 8 for each error " | |
| 2138 "(implies --first)") | |
| 2139 parser.add_option('--statistics', action='store_true', | |
| 2140 help="count errors and warnings") | |
| 2141 parser.add_option('--count', action='store_true', | |
| 2142 help="print total number of errors and warnings " | |
| 2143 "to standard error and set exit code to 1 if " | |
| 2144 "total is not null") | |
| 2145 parser.add_option('--max-line-length', type='int', metavar='n', | |
| 2146 default=MAX_LINE_LENGTH, | |
| 2147 help="set maximum allowed line length " | |
| 2148 "(default: %default)") | |
| 2149 parser.add_option('--hang-closing', action='store_true', | |
| 2150 help="hang closing bracket instead of matching " | |
| 2151 "indentation of opening bracket's line") | |
| 2152 parser.add_option('--format', metavar='format', default='default', | |
| 2153 help="set the error format [default|pylint|<custom>]") | |
| 2154 parser.add_option('--diff', action='store_true', | |
| 2155 help="report changes only within line number ranges in " | |
| 2156 "the unified diff received on STDIN") | |
| 2157 group = parser.add_option_group("Testing Options") | |
| 2158 if os.path.exists(TESTSUITE_PATH): | |
| 2159 group.add_option('--testsuite', metavar='dir', | |
| 2160 help="run regression tests from dir") | |
| 2161 group.add_option('--doctest', action='store_true', | |
| 2162 help="run doctest on myself") | |
| 2163 group.add_option('--benchmark', action='store_true', | |
| 2164 help="measure processing speed") | |
| 2165 return parser | |
| 2166 | |
| 2167 | |
| 2168 def read_config(options, args, arglist, parser): | |
| 2169 """Read and parse configurations | |
| 2170 | |
| 2171 If a config file is specified on the command line with the "--config" | |
| 2172 option, then only it is used for configuration. | |
| 2173 | |
| 2174 Otherwise, the user configuration (~/.config/pycodestyle) and any local | |
| 2175 configurations in the current directory or above will be merged together | |
| 2176 (in that order) using the read method of ConfigParser. | |
| 2177 """ | |
| 2178 config = RawConfigParser() | |
| 2179 | |
| 2180 cli_conf = options.config | |
| 2181 | |
| 2182 local_dir = os.curdir | |
| 2183 | |
| 2184 if USER_CONFIG and os.path.isfile(USER_CONFIG): | |
| 2185 if options.verbose: | |
| 2186 print('user configuration: %s' % USER_CONFIG) | |
| 2187 config.read(USER_CONFIG) | |
| 2188 | |
| 2189 parent = tail = args and os.path.abspath(os.path.commonprefix(args)) | |
| 2190 while tail: | |
| 2191 if config.read(os.path.join(parent, fn) for fn in PROJECT_CONFIG): | |
| 2192 local_dir = parent | |
| 2193 if options.verbose: | |
| 2194 print('local configuration: in %s' % parent) | |
| 2195 break | |
| 2196 (parent, tail) = os.path.split(parent) | |
| 2197 | |
| 2198 if cli_conf and os.path.isfile(cli_conf): | |
| 2199 if options.verbose: | |
| 2200 print('cli configuration: %s' % cli_conf) | |
| 2201 config.read(cli_conf) | |
| 2202 | |
| 2203 pep8_section = parser.prog | |
| 2204 if config.has_section(pep8_section): | |
| 2205 option_list = dict([(o.dest, o.type or o.action) | |
| 2206 for o in parser.option_list]) | |
| 2207 | |
| 2208 # First, read the default values | |
| 2209 (new_options, __) = parser.parse_args([]) | |
| 2210 | |
| 2211 # Second, parse the configuration | |
| 2212 for opt in config.options(pep8_section): | |
| 2213 if opt.replace('_', '-') not in parser.config_options: | |
| 2214 print(" unknown option '%s' ignored" % opt) | |
| 2215 continue | |
| 2216 if options.verbose > 1: | |
| 2217 print(" %s = %s" % (opt, config.get(pep8_section, opt))) | |
| 2218 normalized_opt = opt.replace('-', '_') | |
| 2219 opt_type = option_list[normalized_opt] | |
| 2220 if opt_type in ('int', 'count'): | |
| 2221 value = config.getint(pep8_section, opt) | |
| 2222 elif opt_type in ('store_true', 'store_false'): | |
| 2223 value = config.getboolean(pep8_section, opt) | |
| 2224 else: | |
| 2225 value = config.get(pep8_section, opt) | |
| 2226 if normalized_opt == 'exclude': | |
| 2227 value = normalize_paths(value, local_dir) | |
| 2228 setattr(new_options, normalized_opt, value) | |
| 2229 | |
| 2230 # Third, overwrite with the command-line options | |
| 2231 (options, __) = parser.parse_args(arglist, values=new_options) | |
| 2232 options.doctest = options.testsuite = False | |
| 2233 return options | |
| 2234 | |
| 2235 | |
| 2236 def process_options(arglist=None, parse_argv=False, config_file=None, | |
| 2237 parser=None): | |
| 2238 """Process options passed either via arglist or via command line args. | |
| 2239 | |
| 2240 Passing in the ``config_file`` parameter allows other tools, such as flake8 | |
| 2241 to specify their own options to be processed in pycodestyle. | |
| 2242 """ | |
| 2243 if not parser: | |
| 2244 parser = get_parser() | |
| 2245 if not parser.has_option('--config'): | |
| 2246 group = parser.add_option_group("Configuration", description=( | |
| 2247 "The project options are read from the [%s] section of the " | |
| 2248 "tox.ini file or the setup.cfg file located in any parent folder " | |
| 2249 "of the path(s) being processed. Allowed options are: %s." % | |
| 2250 (parser.prog, ', '.join(parser.config_options)))) | |
| 2251 group.add_option('--config', metavar='path', default=config_file, | |
| 2252 help="user config file location") | |
| 2253 # Don't read the command line if the module is used as a library. | |
| 2254 if not arglist and not parse_argv: | |
| 2255 arglist = [] | |
| 2256 # If parse_argv is True and arglist is None, arguments are | |
| 2257 # parsed from the command line (sys.argv) | |
| 2258 (options, args) = parser.parse_args(arglist) | |
| 2259 options.reporter = None | |
| 2260 | |
| 2261 if options.ensure_value('testsuite', False): | |
| 2262 args.append(options.testsuite) | |
| 2263 elif not options.ensure_value('doctest', False): | |
| 2264 if parse_argv and not args: | |
| 2265 if options.diff or any(os.path.exists(name) | |
| 2266 for name in PROJECT_CONFIG): | |
| 2267 args = ['.'] | |
| 2268 else: | |
| 2269 parser.error('input not specified') | |
| 2270 options = read_config(options, args, arglist, parser) | |
| 2271 options.reporter = parse_argv and options.quiet == 1 and FileReport | |
| 2272 | |
| 2273 options.filename = _parse_multi_options(options.filename) | |
| 2274 options.exclude = normalize_paths(options.exclude) | |
| 2275 options.select = _parse_multi_options(options.select) | |
| 2276 options.ignore = _parse_multi_options(options.ignore) | |
| 2277 | |
| 2278 if options.diff: | |
| 2279 options.reporter = DiffReport | |
| 2280 stdin = stdin_get_value() | |
| 2281 options.selected_lines = parse_udiff(stdin, options.filename, args[0]) | |
| 2282 args = sorted(options.selected_lines) | |
| 2283 | |
| 2284 return options, args | |
| 2285 | |
| 2286 | |
| 2287 def _parse_multi_options(options, split_token=','): | |
| 2288 r"""Split and strip and discard empties. | |
| 2289 | |
| 2290 Turns the following: | |
| 2291 | |
| 2292 A, | |
| 2293 B, | |
| 2294 | |
| 2295 into ["A", "B"] | |
| 2296 """ | |
| 2297 if options: | |
| 2298 return [o.strip() for o in options.split(split_token) if o.strip()] | |
| 2299 else: | |
| 2300 return options | |
| 2301 | |
| 2302 | |
| 2303 def _main(): | |
| 2304 """Parse options and run checks on Python source.""" | |
| 2305 import signal | |
| 2306 | |
| 2307 # Handle "Broken pipe" gracefully | |
| 2308 try: | |
| 2309 signal.signal(signal.SIGPIPE, lambda signum, frame: sys.exit(1)) | |
| 2310 except AttributeError: | |
| 2311 pass # not supported on Windows | |
| 2312 | |
| 2313 style_guide = StyleGuide(parse_argv=True) | |
| 2314 options = style_guide.options | |
| 2315 | |
| 2316 if options.doctest or options.testsuite: | |
| 2317 from testsuite.support import run_tests | |
| 2318 report = run_tests(style_guide) | |
| 2319 else: | |
| 2320 report = style_guide.check_files() | |
| 2321 | |
| 2322 if options.statistics: | |
| 2323 report.print_statistics() | |
| 2324 | |
| 2325 if options.benchmark: | |
| 2326 report.print_benchmark() | |
| 2327 | |
| 2328 if options.testsuite and not options.quiet: | |
| 2329 report.print_results() | |
| 2330 | |
| 2331 if report.total_errors: | |
| 2332 if options.count: | |
| 2333 sys.stderr.write(str(report.total_errors) + '\n') | |
| 2334 sys.exit(1) | |
| 2335 | |
| 2336 if __name__ == '__main__': | |
| 2337 _main() | |
| 2338 # | |
| 2339 # eflag: noqa = M702 |
