Mercurial Repositories > eric / file comparison
comparison: Plugins/CheckerPlugins/CodeStyleChecker/pep8.py
Plugins/CheckerPlugins/CodeStyleChecker/pep8.py
- changeset 2982
- 556adfe76ba7
- parent 2910
- cdc56e9d9f12
- child 3004
- c4bf32c791d0
equal
deleted
inserted
replaced
| 2981:0c5205aa7a5a | 2982:556adfe76ba7 |
|---|---|
| 1 # -*- coding: utf-8 -*- | |
| 2 | |
| 3 # | |
| 4 # pep8.py - Check Python source code formatting, according to PEP 8 | |
| 5 # Copyright (C) 2006-2009 Johann C. Rocholl <johann@rocholl.net> | |
| 6 # Copyright (C) 2009-2013 Florent Xicluna <florent.xicluna@gmail.com> | |
| 7 # | |
| 8 # Permission is hereby granted, free of charge, to any person | |
| 9 # obtaining a copy of this software and associated documentation files | |
| 10 # (the "Software"), to deal in the Software without restriction, | |
| 11 # including without limitation the rights to use, copy, modify, merge, | |
| 12 # publish, distribute, sublicense, and/or sell copies of the Software, | |
| 13 # and to permit persons to whom the Software is furnished to do so, | |
| 14 # subject to the following conditions: | |
| 15 # | |
| 16 # The above copyright notice and this permission notice shall be | |
| 17 # included in all copies or substantial portions of the Software. | |
| 18 # | |
| 19 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, | |
| 20 # EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF | |
| 21 # MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND | |
| 22 # NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS | |
| 23 # BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN | |
| 24 # ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN | |
| 25 # CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE | |
| 26 # SOFTWARE. | |
| 27 | |
| 28 """ | |
| 29 Check Python source code formatting, according to PEP 8: | |
| 30 http://www.python.org/dev/peps/pep-0008/ | |
| 31 | |
| 32 For usage and a list of options, try this: | |
| 33 $ python pep8.py -h | |
| 34 | |
| 35 This program and its regression test suite live here: | |
| 36 http://github.com/jcrocholl/pep8 | |
| 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 | |
| 51 # | |
| 52 # This is a modified version to make the original pep8.py better suitable | |
| 53 # for being called from within the eric5 IDE. The modifications are as | |
| 54 # follows: | |
| 55 # | |
| 56 # - made messages translatable via Qt | |
| 57 # - added code for eric5 integration | |
| 58 # | |
| 59 # Copyright (c) 2011 - 2013 Detlev Offenbach <detlev@die-offenbachs.de> | |
| 60 # | |
| 61 | |
| 62 __version__ = '1.4.6' | |
| 63 | |
| 64 import os | |
| 65 import sys | |
| 66 import re | |
| 67 import time | |
| 68 import inspect | |
| 69 import keyword | |
| 70 import tokenize | |
| 71 from optparse import OptionParser | |
| 72 from fnmatch import fnmatch | |
| 73 try: | |
| 74 from configparser import RawConfigParser | |
| 75 from io import TextIOWrapper | |
| 76 except ImportError: | |
| 77 from ConfigParser import RawConfigParser # __IGNORE_WARNING__ | |
| 78 | |
| 79 from PyQt4.QtCore import QCoreApplication, QT_TRANSLATE_NOOP | |
| 80 | |
| 81 DEFAULT_EXCLUDE = '.svn,CVS,.bzr,.hg,.git,__pycache__' | |
| 82 DEFAULT_IGNORE = 'E123,E226,E24' | |
| 83 if sys.platform == 'win32': | |
| 84 DEFAULT_CONFIG = os.path.expanduser(r'~\.pep8') | |
| 85 else: | |
| 86 DEFAULT_CONFIG = os.path.join(os.getenv('XDG_CONFIG_HOME') or | |
| 87 os.path.expanduser('~/.config'), 'pep8') | |
| 88 PROJECT_CONFIG = ('setup.cfg', 'tox.ini', '.pep8') | |
| 89 TESTSUITE_PATH = os.path.join(os.path.dirname(__file__), 'testsuite') | |
| 90 MAX_LINE_LENGTH = 79 | |
| 91 REPORT_FORMAT = { | |
| 92 'default': '%(path)s:%(row)d:%(col)d: %(code)s %(text)s', | |
| 93 'pylint': '%(path)s:%(row)d: [%(code)s] %(text)s', | |
| 94 } | |
| 95 | |
| 96 PyCF_ONLY_AST = 1024 | |
| 97 SINGLETONS = frozenset(['False', 'None', 'True']) | |
| 98 KEYWORDS = frozenset(keyword.kwlist + ['print']) - SINGLETONS | |
| 99 UNARY_OPERATORS = frozenset(['>>', '**', '*', '+', '-']) | |
| 100 ARITHMETIC_OP = frozenset(['**', '*', '/', '//', '+', '-']) | |
| 101 WS_OPTIONAL_OPERATORS = ARITHMETIC_OP.union(['^', '&', '|', '<<', '>>', '%']) | |
| 102 WS_NEEDED_OPERATORS = frozenset([ | |
| 103 '**=', '*=', '/=', '//=', '+=', '-=', '!=', '<>', '<', '>', | |
| 104 '%=', '^=', '&=', '|=', '==', '<=', '>=', '<<=', '>>=', '=']) | |
| 105 WHITESPACE = frozenset(' \t') | |
| 106 SKIP_TOKENS = frozenset([tokenize.COMMENT, tokenize.NL, tokenize.NEWLINE, | |
| 107 tokenize.INDENT, tokenize.DEDENT]) | |
| 108 BENCHMARK_KEYS = ['directories', 'files', 'logical lines', 'physical lines'] | |
| 109 | |
| 110 INDENT_REGEX = re.compile(r'([ \t]*)') | |
| 111 RAISE_COMMA_REGEX = re.compile(r'raise\s+\w+\s*,') | |
| 112 RERAISE_COMMA_REGEX = re.compile(r'raise\s+\w+\s*,\s*\w+\s*,\s*\w+') | |
| 113 ERRORCODE_REGEX = re.compile(r'\b[A-Z]\d{3}\b') | |
| 114 DOCSTRING_REGEX = re.compile(r'u?r?["\']') | |
| 115 EXTRANEOUS_WHITESPACE_REGEX = re.compile(r'[[({] | []}),;:]') | |
| 116 WHITESPACE_AFTER_COMMA_REGEX = re.compile(r'[,;:]\s*(?: |\t)') | |
| 117 COMPARE_SINGLETON_REGEX = re.compile(r'([=!]=)\s*(None|False|True)') | |
| 118 COMPARE_TYPE_REGEX = re.compile(r'(?:[=!]=|is(?:\s+not)?)\s*type(?:s.\w+Type' | |
| 119 r'|\s*\(\s*([^)]*[^ )])\s*\))') | |
| 120 KEYWORD_REGEX = re.compile(r'(\s*)\b(?:%s)\b(\s*)' % r'|'.join(KEYWORDS)) | |
| 121 OPERATOR_REGEX = re.compile(r'(?:[^,\s])(\s*)(?:[-+*/|!<=>%&^]+)(\s*)') | |
| 122 LAMBDA_REGEX = re.compile(r'\blambda\b') | |
| 123 HUNK_REGEX = re.compile(r'^@@ -\d+(?:,\d+)? \+(\d+)(?:,(\d+))? @@.*$') | |
| 124 | |
| 125 # Work around Python < 2.6 behaviour, which does not generate NL after | |
| 126 # a comment which is on a line by itself. | |
| 127 COMMENT_WITH_NL = tokenize.generate_tokens(['#\n'].pop).send(None)[1] == '#\n' | |
| 128 | |
| 129 | |
| 130 ############################################################################## | |
| 131 # Helper functions for translated and formatted messages | |
| 132 ############################################################################## | |
| 133 | |
| 134 | |
| 135 pep8_messages = { | |
| 136 "E101": QT_TRANSLATE_NOOP("pep8", | |
| 137 "indentation contains mixed spaces and tabs"), | |
| 138 "E111": QT_TRANSLATE_NOOP("pep8", | |
| 139 "indentation is not a multiple of four"), | |
| 140 "E112": QT_TRANSLATE_NOOP("pep8", | |
| 141 "expected an indented block"), | |
| 142 "E113": QT_TRANSLATE_NOOP("pep8", | |
| 143 "unexpected indentation"), | |
| 144 "E121": QT_TRANSLATE_NOOP("pep8", | |
| 145 "continuation line indentation is not a multiple of four"), | |
| 146 "E122": QT_TRANSLATE_NOOP("pep8", | |
| 147 "continuation line missing indentation or outdented"), | |
| 148 "E123": QT_TRANSLATE_NOOP("pep8", | |
| 149 "closing bracket does not match indentation of opening bracket's line"), | |
| 150 "E124": QT_TRANSLATE_NOOP("pep8", | |
| 151 "closing bracket does not match visual indentation"), | |
| 152 "E125": QT_TRANSLATE_NOOP("pep8", | |
| 153 "continuation line does not distinguish itself from next logical line"), | |
| 154 "E126": QT_TRANSLATE_NOOP("pep8", | |
| 155 "continuation line over-indented for hanging indent"), | |
| 156 "E127": QT_TRANSLATE_NOOP("pep8", | |
| 157 "continuation line over-indented for visual indent"), | |
| 158 "E128": QT_TRANSLATE_NOOP("pep8", | |
| 159 "continuation line under-indented for visual indent"), | |
| 160 "E133": QT_TRANSLATE_NOOP("pep8", | |
| 161 "closing bracket is missing indentation"), | |
| 162 "W191": QT_TRANSLATE_NOOP("pep8", | |
| 163 "indentation contains tabs"), | |
| 164 "E201": QT_TRANSLATE_NOOP("pep8", | |
| 165 "whitespace after '{0}'"), | |
| 166 "E202": QT_TRANSLATE_NOOP("pep8", | |
| 167 "whitespace before '{0}'"), | |
| 168 "E203": QT_TRANSLATE_NOOP("pep8", | |
| 169 "whitespace before '{0}'"), | |
| 170 "E211": QT_TRANSLATE_NOOP("pep8", | |
| 171 "whitespace before '{0}'"), | |
| 172 "E221": QT_TRANSLATE_NOOP("pep8", | |
| 173 "multiple spaces before operator"), | |
| 174 "E222": QT_TRANSLATE_NOOP("pep8", | |
| 175 "multiple spaces after operator"), | |
| 176 "E223": QT_TRANSLATE_NOOP("pep8", | |
| 177 "tab before operator"), | |
| 178 "E224": QT_TRANSLATE_NOOP("pep8", | |
| 179 "tab after operator"), | |
| 180 "E225": QT_TRANSLATE_NOOP("pep8", | |
| 181 "missing whitespace around operator"), | |
| 182 "E226": QT_TRANSLATE_NOOP("pep8", | |
| 183 "missing whitespace around arithmetic operator"), | |
| 184 "E227": QT_TRANSLATE_NOOP("pep8", | |
| 185 "missing whitespace around bitwise or shift operator"), | |
| 186 "E228": QT_TRANSLATE_NOOP("pep8", | |
| 187 "missing whitespace around modulo operator"), | |
| 188 "E231": QT_TRANSLATE_NOOP("pep8", | |
| 189 "missing whitespace after '{0}'"), | |
| 190 "E241": QT_TRANSLATE_NOOP("pep8", | |
| 191 "multiple spaces after '{0}'"), | |
| 192 "E242": QT_TRANSLATE_NOOP("pep8", | |
| 193 "tab after '{0}'"), | |
| 194 "E251": QT_TRANSLATE_NOOP("pep8", | |
| 195 "unexpected spaces around keyword / parameter equals"), | |
| 196 "E261": QT_TRANSLATE_NOOP("pep8", | |
| 197 "at least two spaces before inline comment"), | |
| 198 "E262": QT_TRANSLATE_NOOP("pep8", | |
| 199 "inline comment should start with '# '"), | |
| 200 "E271": QT_TRANSLATE_NOOP("pep8", | |
| 201 "multiple spaces after keyword"), | |
| 202 "E272": QT_TRANSLATE_NOOP("pep8", | |
| 203 "multiple spaces before keyword"), | |
| 204 "E273": QT_TRANSLATE_NOOP("pep8", | |
| 205 "tab after keyword"), | |
| 206 "E274": QT_TRANSLATE_NOOP("pep8", | |
| 207 "tab before keyword"), | |
| 208 "W291": QT_TRANSLATE_NOOP("pep8", | |
| 209 "trailing whitespace"), | |
| 210 "W292": QT_TRANSLATE_NOOP("pep8", | |
| 211 "no newline at end of file"), | |
| 212 "W293": QT_TRANSLATE_NOOP("pep8", | |
| 213 "blank line contains whitespace"), | |
| 214 "E301": QT_TRANSLATE_NOOP("pep8", | |
| 215 "expected 1 blank line, found 0"), | |
| 216 "E302": QT_TRANSLATE_NOOP("pep8", | |
| 217 "expected 2 blank lines, found {0}"), | |
| 218 "E303": QT_TRANSLATE_NOOP("pep8", | |
| 219 "too many blank lines ({0})"), | |
| 220 "E304": QT_TRANSLATE_NOOP("pep8", | |
| 221 "blank lines found after function decorator"), | |
| 222 "W391": QT_TRANSLATE_NOOP("pep8", | |
| 223 "blank line at end of file"), | |
| 224 "E401": QT_TRANSLATE_NOOP("pep8", | |
| 225 "multiple imports on one line"), | |
| 226 "E501": QT_TRANSLATE_NOOP("pep8", | |
| 227 "line too long ({0} > {1} characters)"), | |
| 228 "E502": QT_TRANSLATE_NOOP("pep8", | |
| 229 "the backslash is redundant between brackets"), | |
| 230 "W601": QT_TRANSLATE_NOOP("pep8", | |
| 231 ".has_key() is deprecated, use 'in'"), | |
| 232 "W602": QT_TRANSLATE_NOOP("pep8", | |
| 233 "deprecated form of raising exception"), | |
| 234 "W603": QT_TRANSLATE_NOOP("pep8", | |
| 235 "'<>' is deprecated, use '!='"), | |
| 236 "W604": QT_TRANSLATE_NOOP("pep8", | |
| 237 "backticks are deprecated, use 'repr()'"), | |
| 238 "E701": QT_TRANSLATE_NOOP("pep8", | |
| 239 "multiple statements on one line (colon)"), | |
| 240 "E702": QT_TRANSLATE_NOOP("pep8", | |
| 241 "multiple statements on one line (semicolon)"), | |
| 242 "E703": QT_TRANSLATE_NOOP("pep8", | |
| 243 "statement ends with a semicolon"), | |
| 244 "E711": QT_TRANSLATE_NOOP("pep8", | |
| 245 "comparison to {0} should be {1}"), | |
| 246 "E712": QT_TRANSLATE_NOOP("pep8", | |
| 247 "comparison to {0} should be {1}"), | |
| 248 "E721": QT_TRANSLATE_NOOP("pep8", | |
| 249 "do not compare types, use 'isinstance()'"), | |
| 250 "E901": QT_TRANSLATE_NOOP("pep8", | |
| 251 "{0}: {1}"), | |
| 252 } | |
| 253 | |
| 254 pep8_messages_sample_args = { | |
| 255 "E201": ["([{"], | |
| 256 "E202": ["}])"], | |
| 257 "E203": [",;:"], | |
| 258 "E211": ["(["], | |
| 259 "E231": [",;:"], | |
| 260 "E241": [",;:"], | |
| 261 "E242": [",;:"], | |
| 262 "E302": [1], | |
| 263 "E303": [3], | |
| 264 "E501": [85, 79], | |
| 265 "E711": ["None", "'if cond is None:'"], | |
| 266 "E712": ["True", "'if cond is True:' or 'if cond:'"], | |
| 267 "E901": ["SyntaxError", "Invalid Syntax"], | |
| 268 } | |
| 269 | |
| 270 | |
| 271 def getMessage(code, *args): | |
| 272 """ | |
| 273 Function to get a translated and formatted message for a given code. | |
| 274 | |
| 275 @param code message code (string) | |
| 276 @param args arguments for a formatted message (list) | |
| 277 @return translated and formatted message (string) | |
| 278 """ | |
| 279 if code in pep8_messages: | |
| 280 return code + " " + QCoreApplication.translate("pep8", | |
| 281 pep8_messages[code]).format(*args) | |
| 282 else: | |
| 283 return code + " " + QCoreApplication.translate("pep8", | |
| 284 "no message for this code defined") | |
| 285 | |
| 286 ############################################################################## | |
| 287 # Plugins (check functions) for physical lines | |
| 288 ############################################################################## | |
| 289 | |
| 290 | |
| 291 def tabs_or_spaces(physical_line, indent_char): | |
| 292 r""" | |
| 293 Never mix tabs and spaces. | |
| 294 | |
| 295 The most popular way of indenting Python is with spaces only. The | |
| 296 second-most popular way is with tabs only. Code indented with a mixture | |
| 297 of tabs and spaces should be converted to using spaces exclusively. When | |
| 298 invoking the Python command line interpreter with the -t option, it issues | |
| 299 warnings about code that illegally mixes tabs and spaces. When using -tt | |
| 300 these warnings become errors. These options are highly recommended! | |
| 301 | |
| 302 Okay: if a == 0:\n a = 1\n b = 1 | |
| 303 E101: if a == 0:\n a = 1\n\tb = 1 | |
| 304 """ | |
| 305 indent = INDENT_REGEX.match(physical_line).group(1) | |
| 306 for offset, char in enumerate(indent): | |
| 307 if char != indent_char: | |
| 308 return offset, "E101" | |
| 309 | |
| 310 | |
| 311 def tabs_obsolete(physical_line): | |
| 312 r""" | |
| 313 For new projects, spaces-only are strongly recommended over tabs. Most | |
| 314 editors have features that make this easy to do. | |
| 315 | |
| 316 Okay: if True:\n return | |
| 317 W191: if True:\n\treturn | |
| 318 """ | |
| 319 indent = INDENT_REGEX.match(physical_line).group(1) | |
| 320 if '\t' in indent: | |
| 321 return indent.index('\t'), "W191" | |
| 322 | |
| 323 | |
| 324 def trailing_whitespace(physical_line): | |
| 325 r""" | |
| 326 JCR: Trailing whitespace is superfluous. | |
| 327 FBM: Except when it occurs as part of a blank line (i.e. the line is | |
| 328 nothing but whitespace). According to Python docs[1] a line with only | |
| 329 whitespace is considered a blank line, and is to be ignored. However, | |
| 330 matching a blank line to its indentation level avoids mistakenly | |
| 331 terminating a multi-line statement (e.g. class declaration) when | |
| 332 pasting code into the standard Python interpreter. | |
| 333 | |
| 334 [1] http://docs.python.org/reference/lexical_analysis.html#blank-lines | |
| 335 | |
| 336 The warning returned varies on whether the line itself is blank, for easier | |
| 337 filtering for those who want to indent their blank lines. | |
| 338 | |
| 339 Okay: spam(1)\n# | |
| 340 W291: spam(1) \n# | |
| 341 W293: class Foo(object):\n \n bang = 12 | |
| 342 """ | |
| 343 physical_line = physical_line.rstrip('\n') # chr(10), newline | |
| 344 physical_line = physical_line.rstrip('\r') # chr(13), carriage return | |
| 345 physical_line = physical_line.rstrip('\x0c') # chr(12), form feed, ^L | |
| 346 stripped = physical_line.rstrip(' \t\v') | |
| 347 if physical_line != stripped: | |
| 348 if stripped: | |
| 349 return len(stripped), "W291" | |
| 350 else: | |
| 351 return 0, "W293" | |
| 352 | |
| 353 | |
| 354 def trailing_blank_lines(physical_line, lines, line_number): | |
| 355 r""" | |
| 356 JCR: Trailing blank lines are superfluous. | |
| 357 | |
| 358 Okay: spam(1) | |
| 359 W391: spam(1)\n | |
| 360 """ | |
| 361 if not physical_line.rstrip() and line_number == len(lines): | |
| 362 return 0, "W391" | |
| 363 | |
| 364 | |
| 365 def missing_newline(physical_line): | |
| 366 """ | |
| 367 JCR: The last line should have a newline. | |
| 368 | |
| 369 Reports warning W292. | |
| 370 """ | |
| 371 if physical_line.rstrip() == physical_line: | |
| 372 return len(physical_line), "W292" | |
| 373 | |
| 374 | |
| 375 def maximum_line_length(physical_line, max_line_length): | |
| 376 """ | |
| 377 Limit all lines to a maximum of 79 characters. | |
| 378 | |
| 379 There are still many devices around that are limited to 80 character | |
| 380 lines; plus, limiting windows to 80 characters makes it possible to have | |
| 381 several windows side-by-side. The default wrapping on such devices looks | |
| 382 ugly. Therefore, please limit all lines to a maximum of 79 characters. | |
| 383 For flowing long blocks of text (docstrings or comments), limiting the | |
| 384 length to 72 characters is recommended. | |
| 385 | |
| 386 Reports error E501. | |
| 387 """ | |
| 388 line = physical_line.rstrip() | |
| 389 length = len(line) | |
| 390 if length > max_line_length and not noqa(line): | |
| 391 if hasattr(line, 'decode'): # Python 2 | |
| 392 # The line could contain multi-byte characters | |
| 393 try: | |
| 394 length = len(line.decode('utf-8')) | |
| 395 except UnicodeError: | |
| 396 pass | |
| 397 if length > max_line_length: | |
| 398 return max_line_length, "E501", length, max_line_length | |
| 399 | |
| 400 | |
| 401 ############################################################################## | |
| 402 # Plugins (check functions) for logical lines | |
| 403 ############################################################################## | |
| 404 | |
| 405 | |
| 406 def blank_lines(logical_line, blank_lines, indent_level, line_number, | |
| 407 previous_logical, previous_indent_level): | |
| 408 r""" | |
| 409 Separate top-level function and class definitions with two blank lines. | |
| 410 | |
| 411 Method definitions inside a class are separated by a single blank line. | |
| 412 | |
| 413 Extra blank lines may be used (sparingly) to separate groups of related | |
| 414 functions. Blank lines may be omitted between a bunch of related | |
| 415 one-liners (e.g. a set of dummy implementations). | |
| 416 | |
| 417 Use blank lines in functions, sparingly, to indicate logical sections. | |
| 418 | |
| 419 Okay: def a():\n pass\n\n\ndef b():\n pass | |
| 420 Okay: def a():\n pass\n\n\n# Foo\n# Bar\n\ndef b():\n pass | |
| 421 | |
| 422 E301: class Foo:\n b = 0\n def bar():\n pass | |
| 423 E302: def a():\n pass\n\ndef b(n):\n pass | |
| 424 E303: def a():\n pass\n\n\n\ndef b(n):\n pass | |
| 425 E303: def a():\n\n\n\n pass | |
| 426 E304: @decorator\n\ndef a():\n pass | |
| 427 """ | |
| 428 if line_number < 3 and not previous_logical: | |
| 429 return # Don't expect blank lines before the first line | |
| 430 if previous_logical.startswith('@'): | |
| 431 if blank_lines: | |
| 432 yield 0, "E304" | |
| 433 elif blank_lines > 2 or (indent_level and blank_lines == 2): | |
| 434 yield 0, "E303", blank_lines | |
| 435 elif logical_line.startswith(('def ', 'class ', '@')): | |
| 436 if indent_level: | |
| 437 if not (blank_lines or previous_indent_level < indent_level or | |
| 438 DOCSTRING_REGEX.match(previous_logical)): | |
| 439 yield 0, "E301" | |
| 440 elif blank_lines != 2: | |
| 441 yield 0, "E302", blank_lines | |
| 442 | |
| 443 | |
| 444 def extraneous_whitespace(logical_line): | |
| 445 """ | |
| 446 Avoid extraneous whitespace in the following situations: | |
| 447 | |
| 448 - Immediately inside parentheses, brackets or braces. | |
| 449 | |
| 450 - Immediately before a comma, semicolon, or colon. | |
| 451 | |
| 452 Okay: spam(ham[1], {eggs: 2}) | |
| 453 E201: spam( ham[1], {eggs: 2}) | |
| 454 E201: spam(ham[ 1], {eggs: 2}) | |
| 455 E201: spam(ham[1], { eggs: 2}) | |
| 456 E202: spam(ham[1], {eggs: 2} ) | |
| 457 E202: spam(ham[1 ], {eggs: 2}) | |
| 458 E202: spam(ham[1], {eggs: 2 }) | |
| 459 | |
| 460 E203: if x == 4: print x, y; x, y = y , x | |
| 461 E203: if x == 4: print x, y ; x, y = y, x | |
| 462 E203: if x == 4 : print x, y; x, y = y, x | |
| 463 """ | |
| 464 line = logical_line | |
| 465 for match in EXTRANEOUS_WHITESPACE_REGEX.finditer(line): | |
| 466 text = match.group() | |
| 467 char = text.strip() | |
| 468 found = match.start() | |
| 469 if text == char + ' ': | |
| 470 # assert char in '([{' | |
| 471 yield found + 1, "E201", char | |
| 472 elif line[found - 1] != ',': | |
| 473 code = ('E202' if char in '}])' else 'E203') # if char in ',;:' | |
| 474 yield found, code, char | |
| 475 | |
| 476 | |
| 477 def whitespace_around_keywords(logical_line): | |
| 478 r""" | |
| 479 Avoid extraneous whitespace around keywords. | |
| 480 | |
| 481 Okay: True and False | |
| 482 E271: True and False | |
| 483 E272: True and False | |
| 484 E273: True and\tFalse | |
| 485 E274: True\tand False | |
| 486 """ | |
| 487 for match in KEYWORD_REGEX.finditer(logical_line): | |
| 488 before, after = match.groups() | |
| 489 | |
| 490 if '\t' in before: | |
| 491 yield match.start(1), "E274" | |
| 492 elif len(before) > 1: | |
| 493 yield match.start(1), "E272" | |
| 494 | |
| 495 if '\t' in after: | |
| 496 yield match.start(2), "E273" | |
| 497 elif len(after) > 1: | |
| 498 yield match.start(2), "E271" | |
| 499 | |
| 500 | |
| 501 def missing_whitespace(logical_line): | |
| 502 """ | |
| 503 JCR: Each comma, semicolon or colon should be followed by whitespace. | |
| 504 | |
| 505 Okay: [a, b] | |
| 506 Okay: (3,) | |
| 507 Okay: a[1:4] | |
| 508 Okay: a[:4] | |
| 509 Okay: a[1:] | |
| 510 Okay: a[1:4:2] | |
| 511 E231: ['a','b'] | |
| 512 E231: foo(bar,baz) | |
| 513 E231: [{'a':'b'}] | |
| 514 """ | |
| 515 line = logical_line | |
| 516 for index in range(len(line) - 1): | |
| 517 char = line[index] | |
| 518 if char in ',;:' and line[index + 1] not in WHITESPACE: | |
| 519 before = line[:index] | |
| 520 if char == ':' and before.count('[') > before.count(']') and \ | |
| 521 before.rfind('{') < before.rfind('['): | |
| 522 continue # Slice syntax, no space required | |
| 523 if char == ',' and line[index + 1] == ')': | |
| 524 continue # Allow tuple with only one element: (3,) | |
| 525 yield index, "E231", char | |
| 526 | |
| 527 | |
| 528 def indentation(logical_line, previous_logical, indent_char, | |
| 529 indent_level, previous_indent_level): | |
| 530 r""" | |
| 531 Use 4 spaces per indentation level. | |
| 532 | |
| 533 For really old code that you don't want to mess up, you can continue to | |
| 534 use 8-space tabs. | |
| 535 | |
| 536 Okay: a = 1 | |
| 537 Okay: if a == 0:\n a = 1 | |
| 538 E111: a = 1 | |
| 539 | |
| 540 Okay: for item in items:\n pass | |
| 541 E112: for item in items:\npass | |
| 542 | |
| 543 Okay: a = 1\nb = 2 | |
| 544 E113: a = 1\n b = 2 | |
| 545 """ | |
| 546 if indent_char == ' ' and indent_level % 4: | |
| 547 yield 0, "E111" | |
| 548 indent_expect = previous_logical.endswith(':') | |
| 549 if indent_expect and indent_level <= previous_indent_level: | |
| 550 yield 0, "E112" | |
| 551 if indent_level > previous_indent_level and not indent_expect: | |
| 552 yield 0, "E113" | |
| 553 | |
| 554 | |
| 555 def continued_indentation(logical_line, tokens, indent_level, hang_closing, | |
| 556 noqa, verbose): | |
| 557 r""" | |
| 558 Continuation lines should align wrapped elements either vertically using | |
| 559 Python's implicit line joining inside parentheses, brackets and braces, or | |
| 560 using a hanging indent. | |
| 561 | |
| 562 When using a hanging indent the following considerations should be applied: | |
| 563 | |
| 564 - there should be no arguments on the first line, and | |
| 565 | |
| 566 - further indentation should be used to clearly distinguish itself as a | |
| 567 continuation line. | |
| 568 | |
| 569 Okay: a = (\n) | |
| 570 E123: a = (\n ) | |
| 571 | |
| 572 Okay: a = (\n 42) | |
| 573 E121: a = (\n 42) | |
| 574 E122: a = (\n42) | |
| 575 E123: a = (\n 42\n ) | |
| 576 E124: a = (24,\n 42\n) | |
| 577 E125: if (a or\n b):\n pass | |
| 578 E126: a = (\n 42) | |
| 579 E127: a = (24,\n 42) | |
| 580 E128: a = (24,\n 42) | |
| 581 """ | |
| 582 first_row = tokens[0][2][0] | |
| 583 nrows = 1 + tokens[-1][2][0] - first_row | |
| 584 if noqa or nrows == 1: | |
| 585 return | |
| 586 | |
| 587 # indent_next tells us whether the next block is indented; assuming | |
| 588 # that it is indented by 4 spaces, then we should not allow 4-space | |
| 589 # indents on the final continuation line; in turn, some other | |
| 590 # indents are allowed to have an extra 4 spaces. | |
| 591 indent_next = logical_line.endswith(':') | |
| 592 | |
| 593 row = depth = 0 | |
| 594 # remember how many brackets were opened on each line | |
| 595 parens = [0] * nrows | |
| 596 # relative indents of physical lines | |
| 597 rel_indent = [0] * nrows | |
| 598 # visual indents | |
| 599 indent_chances = {} | |
| 600 last_indent = tokens[0][2] | |
| 601 indent = [last_indent[1]] | |
| 602 if verbose >= 3: | |
| 603 print(">>> " + tokens[0][4].rstrip()) | |
| 604 | |
| 605 for token_type, text, start, end, line in tokens: | |
| 606 | |
| 607 last_token_multiline = (start[0] != end[0]) | |
| 608 newline = row < start[0] - first_row | |
| 609 if newline: | |
| 610 row = start[0] - first_row | |
| 611 newline = (not last_token_multiline and | |
| 612 token_type not in (tokenize.NL, tokenize.NEWLINE)) | |
| 613 | |
| 614 if newline: | |
| 615 # this is the beginning of a continuation line. | |
| 616 last_indent = start | |
| 617 if verbose >= 3: | |
| 618 print("... " + line.rstrip()) | |
| 619 | |
| 620 # record the initial indent. | |
| 621 rel_indent[row] = expand_indent(line) - indent_level | |
| 622 | |
| 623 if depth: | |
| 624 # a bracket expression in a continuation line. | |
| 625 # find the line that it was opened on | |
| 626 for open_row in range(row - 1, -1, -1): | |
| 627 if parens[open_row]: | |
| 628 break | |
| 629 else: | |
| 630 # an unbracketed continuation line (ie, backslash) | |
| 631 open_row = 0 | |
| 632 hang = rel_indent[row] - rel_indent[open_row] | |
| 633 close_bracket = (token_type == tokenize.OP and text in ']})') | |
| 634 visual_indent = (not close_bracket and hang > 0 and | |
| 635 indent_chances.get(start[1])) | |
| 636 | |
| 637 if close_bracket and indent[depth]: | |
| 638 # closing bracket for visual indent | |
| 639 if start[1] != indent[depth]: | |
| 640 yield start, "E124" | |
| 641 elif close_bracket and not hang: | |
| 642 # closing bracket matches indentation of opening bracket's line | |
| 643 if hang_closing: | |
| 644 yield start, "E133" | |
| 645 elif visual_indent is True: | |
| 646 # visual indent is verified | |
| 647 if not indent[depth]: | |
| 648 indent[depth] = start[1] | |
| 649 elif visual_indent in (text, str): | |
| 650 # ignore token lined up with matching one from a previous line | |
| 651 pass | |
| 652 elif indent[depth] and start[1] < indent[depth]: | |
| 653 # visual indent is broken | |
| 654 yield start, "E128" | |
| 655 elif hang == 4 or (indent_next and rel_indent[row] == 8): | |
| 656 # hanging indent is verified | |
| 657 if close_bracket and not hang_closing: | |
| 658 yield (start, "E123") | |
| 659 else: | |
| 660 # indent is broken | |
| 661 if hang <= 0: | |
| 662 error = "E122" | |
| 663 elif indent[depth]: | |
| 664 error = "E127" | |
| 665 elif hang % 4: | |
| 666 error = "E121" | |
| 667 else: | |
| 668 error = "E126" | |
| 669 yield start, error | |
| 670 | |
| 671 # look for visual indenting | |
| 672 if (parens[row] and token_type not in (tokenize.NL, tokenize.COMMENT) | |
| 673 and not indent[depth]): | |
| 674 indent[depth] = start[1] | |
| 675 indent_chances[start[1]] = True | |
| 676 if verbose >= 4: | |
| 677 print("bracket depth %s indent to %s" % (depth, start[1])) | |
| 678 # deal with implicit string concatenation | |
| 679 elif (token_type in (tokenize.STRING, tokenize.COMMENT) or | |
| 680 text in ('u', 'ur', 'b', 'br')): | |
| 681 indent_chances[start[1]] = str | |
| 682 # special case for the "if" statement because len("if (") == 4 | |
| 683 elif not indent_chances and not row and not depth and text == 'if': | |
| 684 indent_chances[end[1] + 1] = True | |
| 685 | |
| 686 # keep track of bracket depth | |
| 687 if token_type == tokenize.OP: | |
| 688 if text in '([{': | |
| 689 depth += 1 | |
| 690 indent.append(0) | |
| 691 parens[row] += 1 | |
| 692 if verbose >= 4: | |
| 693 print("bracket depth %s seen, col %s, visual min = %s" % | |
| 694 (depth, start[1], indent[depth])) | |
| 695 elif text in ')]}' and depth > 0: | |
| 696 # parent indents should not be more than this one | |
| 697 prev_indent = indent.pop() or last_indent[1] | |
| 698 for d in range(depth): | |
| 699 if indent[d] > prev_indent: | |
| 700 indent[d] = 0 | |
| 701 for ind in list(indent_chances): | |
| 702 if ind >= prev_indent: | |
| 703 del indent_chances[ind] | |
| 704 depth -= 1 | |
| 705 if depth: | |
| 706 indent_chances[indent[depth]] = True | |
| 707 for idx in range(row, -1, -1): | |
| 708 if parens[idx]: | |
| 709 parens[idx] -= 1 | |
| 710 rel_indent[row] = rel_indent[idx] | |
| 711 break | |
| 712 assert len(indent) == depth + 1 | |
| 713 if start[1] not in indent_chances: | |
| 714 # allow to line up tokens | |
| 715 indent_chances[start[1]] = text | |
| 716 | |
| 717 if indent_next and expand_indent(line) == indent_level + 4: | |
| 718 yield last_indent, "E125" | |
| 719 | |
| 720 | |
| 721 def whitespace_before_parameters(logical_line, tokens): | |
| 722 """ | |
| 723 Avoid extraneous whitespace in the following situations: | |
| 724 | |
| 725 - Immediately before the open parenthesis that starts the argument | |
| 726 list of a function call. | |
| 727 | |
| 728 - Immediately before the open parenthesis that starts an indexing or | |
| 729 slicing. | |
| 730 | |
| 731 Okay: spam(1) | |
| 732 E211: spam (1) | |
| 733 | |
| 734 Okay: dict['key'] = list[index] | |
| 735 E211: dict ['key'] = list[index] | |
| 736 E211: dict['key'] = list [index] | |
| 737 """ | |
| 738 prev_type, prev_text, __, prev_end, __ = tokens[0] | |
| 739 for index in range(1, len(tokens)): | |
| 740 token_type, text, start, end, __ = tokens[index] | |
| 741 if (token_type == tokenize.OP and | |
| 742 text in '([' and | |
| 743 start != prev_end and | |
| 744 (prev_type == tokenize.NAME or prev_text in '}])') and | |
| 745 # Syntax "class A (B):" is allowed, but avoid it | |
| 746 (index < 2 or tokens[index - 2][1] != 'class') and | |
| 747 # Allow "return (a.foo for a in range(5))" | |
| 748 not keyword.iskeyword(prev_text)): | |
| 749 yield prev_end, "E211", text | |
| 750 prev_type = token_type | |
| 751 prev_text = text | |
| 752 prev_end = end | |
| 753 | |
| 754 | |
| 755 def whitespace_around_operator(logical_line): | |
| 756 """ | |
| 757 Avoid extraneous whitespace in the following situations: | |
| 758 | |
| 759 - More than one space around an assignment (or other) operator to | |
| 760 align it with another. | |
| 761 | |
| 762 Okay: a = 12 + 3 | |
| 763 E221: a = 4 + 5 | |
| 764 E222: a = 4 + 5 | |
| 765 E223: a = 4\t+ 5 | |
| 766 E224: a = 4 +\t5 | |
| 767 """ | |
| 768 for match in OPERATOR_REGEX.finditer(logical_line): | |
| 769 before, after = match.groups() | |
| 770 | |
| 771 if '\t' in before: | |
| 772 yield match.start(1), "E223" | |
| 773 elif len(before) > 1: | |
| 774 yield match.start(1), "E221" | |
| 775 | |
| 776 if '\t' in after: | |
| 777 yield match.start(2), "E224" | |
| 778 elif len(after) > 1: | |
| 779 yield match.start(2), "E222" | |
| 780 | |
| 781 def missing_whitespace_around_operator(logical_line, tokens): | |
| 782 r""" | |
| 783 - Always surround these binary operators with a single space on | |
| 784 either side: assignment (=), augmented assignment (+=, -= etc.), | |
| 785 comparisons (==, <, >, !=, <>, <=, >=, in, not in, is, is not), | |
| 786 Booleans (and, or, not). | |
| 787 | |
| 788 - Use spaces around arithmetic operators. | |
| 789 | |
| 790 Okay: i = i + 1 | |
| 791 Okay: submitted += 1 | |
| 792 Okay: x = x * 2 - 1 | |
| 793 Okay: hypot2 = x * x + y * y | |
| 794 Okay: c = (a + b) * (a - b) | |
| 795 Okay: foo(bar, key='word', *args, **kwargs) | |
| 796 Okay: alpha[:-i] | |
| 797 | |
| 798 E225: i=i+1 | |
| 799 E225: submitted +=1 | |
| 800 E225: x = x /2 - 1 | |
| 801 E225: z = x **y | |
| 802 E226: c = (a+b) * (a-b) | |
| 803 E226: hypot2 = x*x + y*y | |
| 804 E227: c = a|b | |
| 805 E228: msg = fmt%(errno, errmsg) | |
| 806 """ | |
| 807 parens = 0 | |
| 808 need_space = False | |
| 809 prev_type = tokenize.OP | |
| 810 prev_text = prev_end = None | |
| 811 for token_type, text, start, end, line in tokens: | |
| 812 if token_type in (tokenize.NL, tokenize.NEWLINE, tokenize.ERRORTOKEN): | |
| 813 # ERRORTOKEN is triggered by backticks in Python 3 | |
| 814 continue | |
| 815 if text in ('(', 'lambda'): | |
| 816 parens += 1 | |
| 817 elif text == ')': | |
| 818 parens -= 1 | |
| 819 if need_space: | |
| 820 if start != prev_end: | |
| 821 # Found a (probably) needed space | |
| 822 if need_space is not True and not need_space[1]: | |
| 823 yield need_space[0], "E225" | |
| 824 need_space = False | |
| 825 elif text == '>' and prev_text in ('<', '-'): | |
| 826 # Tolerate the "<>" operator, even if running Python 3 | |
| 827 # Deal with Python 3's annotated return value "->" | |
| 828 pass | |
| 829 else: | |
| 830 if need_space is True or need_space[1]: | |
| 831 # A needed trailing space was not found | |
| 832 yield prev_end, "E225" | |
| 833 else: | |
| 834 code = 'E226' | |
| 835 if prev_text == '%': | |
| 836 code = 'E228' | |
| 837 elif prev_text not in ARITHMETIC_OP: | |
| 838 code = 'E227' | |
| 839 yield need_space[0], code | |
| 840 need_space = False | |
| 841 elif token_type == tokenize.OP and prev_end is not None: | |
| 842 if text == '=' and parens: | |
| 843 # Allow keyword args or defaults: foo(bar=None). | |
| 844 pass | |
| 845 elif text in WS_NEEDED_OPERATORS: | |
| 846 need_space = True | |
| 847 elif text in UNARY_OPERATORS: | |
| 848 # Check if the operator is being used as a binary operator | |
| 849 # Allow unary operators: -123, -x, +1. | |
| 850 # Allow argument unpacking: foo(*args, **kwargs). | |
| 851 if prev_type == tokenize.OP: | |
| 852 binary_usage = (prev_text in '}])') | |
| 853 elif prev_type == tokenize.NAME: | |
| 854 binary_usage = (prev_text not in KEYWORDS) | |
| 855 else: | |
| 856 binary_usage = (prev_type not in SKIP_TOKENS) | |
| 857 | |
| 858 if binary_usage: | |
| 859 need_space = None | |
| 860 elif text in WS_OPTIONAL_OPERATORS: | |
| 861 need_space = None | |
| 862 | |
| 863 if need_space is None: | |
| 864 # Surrounding space is optional, but ensure that | |
| 865 # trailing space matches opening space | |
| 866 need_space = (prev_end, start != prev_end) | |
| 867 elif need_space and start == prev_end: | |
| 868 # A needed opening space was not found | |
| 869 yield prev_end, "E225" | |
| 870 need_space = False | |
| 871 prev_type = token_type | |
| 872 prev_text = text | |
| 873 prev_end = end | |
| 874 | |
| 875 | |
| 876 def whitespace_around_comma(logical_line): | |
| 877 """ | |
| 878 Avoid extraneous whitespace in the following situations: | |
| 879 | |
| 880 - More than one space around an assignment (or other) operator to | |
| 881 align it with another. | |
| 882 | |
| 883 JCR: This should also be applied around comma etc. | |
| 884 Note: these checks are disabled by default | |
| 885 | |
| 886 Okay: a = (1, 2) | |
| 887 E241: a = (1, 2) | |
| 888 E242: a = (1,\t2) | |
| 889 """ | |
| 890 line = logical_line | |
| 891 for m in WHITESPACE_AFTER_COMMA_REGEX.finditer(line): | |
| 892 found = m.start() + 1 | |
| 893 if '\t' in m.group(): | |
| 894 yield found, "E242", m.group()[0] | |
| 895 else: | |
| 896 yield found, "E241", m.group()[0] | |
| 897 | |
| 898 | |
| 899 def whitespace_around_named_parameter_equals(logical_line, tokens): | |
| 900 """ | |
| 901 Don't use spaces around the '=' sign when used to indicate a | |
| 902 keyword argument or a default parameter value. | |
| 903 | |
| 904 Okay: def complex(real, imag=0.0): | |
| 905 Okay: return magic(r=real, i=imag) | |
| 906 Okay: boolean(a == b) | |
| 907 Okay: boolean(a != b) | |
| 908 Okay: boolean(a <= b) | |
| 909 Okay: boolean(a >= b) | |
| 910 | |
| 911 E251: def complex(real, imag = 0.0): | |
| 912 E251: return magic(r = real, i = imag) | |
| 913 """ | |
| 914 parens = 0 | |
| 915 no_space = False | |
| 916 prev_end = None | |
| 917 message = "E251" | |
| 918 for token_type, text, start, end, line in tokens: | |
| 919 if no_space: | |
| 920 no_space = False | |
| 921 if start != prev_end: | |
| 922 yield prev_end, message | |
| 923 elif token_type == tokenize.OP: | |
| 924 if text == '(': | |
| 925 parens += 1 | |
| 926 elif text == ')': | |
| 927 parens -= 1 | |
| 928 elif parens and text == '=': | |
| 929 no_space = True | |
| 930 if start != prev_end: | |
| 931 yield prev_end, message | |
| 932 prev_end = end | |
| 933 | |
| 934 | |
| 935 def whitespace_before_inline_comment(logical_line, tokens): | |
| 936 """ | |
| 937 Separate inline comments by at least two spaces. | |
| 938 | |
| 939 An inline comment is a comment on the same line as a statement. Inline | |
| 940 comments should be separated by at least two spaces from the statement. | |
| 941 They should start with a # and a single space. | |
| 942 | |
| 943 Okay: x = x + 1 # Increment x | |
| 944 Okay: x = x + 1 # Increment x | |
| 945 E261: x = x + 1 # Increment x | |
| 946 E262: x = x + 1 #Increment x | |
| 947 E262: x = x + 1 # Increment x | |
| 948 """ | |
| 949 prev_end = (0, 0) | |
| 950 for token_type, text, start, end, line in tokens: | |
| 951 if token_type == tokenize.COMMENT: | |
| 952 if not line[:start[1]].strip(): | |
| 953 continue | |
| 954 if prev_end[0] == start[0] and start[1] < prev_end[1] + 2: | |
| 955 yield prev_end, "E261" | |
| 956 symbol, sp, comment = text.partition(' ') | |
| 957 if symbol not in ('#', '#:') or comment[:1].isspace(): | |
| 958 yield start, "E262" | |
| 959 elif token_type != tokenize.NL: | |
| 960 prev_end = end | |
| 961 | |
| 962 | |
| 963 def imports_on_separate_lines(logical_line): | |
| 964 r""" | |
| 965 Imports should usually be on separate lines. | |
| 966 | |
| 967 Okay: import os\nimport sys | |
| 968 E401: import sys, os | |
| 969 | |
| 970 Okay: from subprocess import Popen, PIPE | |
| 971 Okay: from myclas import MyClass | |
| 972 Okay: from foo.bar.yourclass import YourClass | |
| 973 Okay: import myclass | |
| 974 Okay: import foo.bar.yourclass | |
| 975 """ | |
| 976 line = logical_line | |
| 977 if line.startswith('import '): | |
| 978 found = line.find(',') | |
| 979 if -1 < found and ';' not in line[:found]: | |
| 980 yield found, "E401" | |
| 981 | |
| 982 | |
| 983 def compound_statements(logical_line): | |
| 984 r""" | |
| 985 Compound statements (multiple statements on the same line) are | |
| 986 generally discouraged. | |
| 987 | |
| 988 While sometimes it's okay to put an if/for/while with a small body | |
| 989 on the same line, never do this for multi-clause statements. Also | |
| 990 avoid folding such long lines! | |
| 991 | |
| 992 Okay: if foo == 'blah':\n do_blah_thing() | |
| 993 Okay: do_one() | |
| 994 Okay: do_two() | |
| 995 Okay: do_three() | |
| 996 | |
| 997 E701: if foo == 'blah': do_blah_thing() | |
| 998 E701: for x in lst: total += x | |
| 999 E701: while t < 10: t = delay() | |
| 1000 E701: if foo == 'blah': do_blah_thing() | |
| 1001 E701: else: do_non_blah_thing() | |
| 1002 E701: try: something() | |
| 1003 E701: finally: cleanup() | |
| 1004 E701: if foo == 'blah': one(); two(); three() | |
| 1005 | |
| 1006 E702: do_one(); do_two(); do_three() | |
| 1007 E703: do_four(); # useless semicolon | |
| 1008 """ | |
| 1009 line = logical_line | |
| 1010 last_char = len(line) - 1 | |
| 1011 found = line.find(':') | |
| 1012 while -1 < found < last_char: | |
| 1013 before = line[:found] | |
| 1014 if (before.count('{') <= before.count('}') and # {'a': 1} (dict) | |
| 1015 before.count('[') <= before.count(']') and # [1:2] (slice) | |
| 1016 before.count('(') <= before.count(')') and # (Python 3 annotation) | |
| 1017 not LAMBDA_REGEX.search(before)): # lambda x: x | |
| 1018 yield found, "E701" | |
| 1019 found = line.find(':', found + 1) | |
| 1020 found = line.find(';') | |
| 1021 while -1 < found: | |
| 1022 if found < last_char: | |
| 1023 yield found, "E702" | |
| 1024 else: | |
| 1025 yield found, "E703" | |
| 1026 found = line.find(';', found + 1) | |
| 1027 | |
| 1028 | |
| 1029 def explicit_line_join(logical_line, tokens): | |
| 1030 r""" | |
| 1031 Avoid explicit line join between brackets. | |
| 1032 | |
| 1033 The preferred way of wrapping long lines is by using Python's implied line | |
| 1034 continuation inside parentheses, brackets and braces. Long lines can be | |
| 1035 broken over multiple lines by wrapping expressions in parentheses. These | |
| 1036 should be used in preference to using a backslash for line continuation. | |
| 1037 | |
| 1038 E502: aaa = [123, \\n 123] | |
| 1039 E502: aaa = ("bbb " \\n "ccc") | |
| 1040 | |
| 1041 Okay: aaa = [123,\n 123] | |
| 1042 Okay: aaa = ("bbb "\n "ccc") | |
| 1043 Okay: aaa = "bbb " \\n "ccc" | |
| 1044 """ | |
| 1045 prev_start = prev_end = parens = 0 | |
| 1046 backslash = None | |
| 1047 for token_type, text, start, end, line in tokens: | |
| 1048 if start[0] != prev_start and parens and backslash: | |
| 1049 yield backslash, "E502" | |
| 1050 if end[0] != prev_end: | |
| 1051 if line.rstrip('\r\n').endswith('\\'): | |
| 1052 backslash = (end[0], len(line.splitlines()[-1]) - 1) | |
| 1053 else: | |
| 1054 backslash = None | |
| 1055 prev_start = prev_end = end[0] | |
| 1056 else: | |
| 1057 prev_start = start[0] | |
| 1058 if token_type == tokenize.OP: | |
| 1059 if text in '([{': | |
| 1060 parens += 1 | |
| 1061 elif text in ')]}': | |
| 1062 parens -= 1 | |
| 1063 | |
| 1064 | |
| 1065 def comparison_to_singleton(logical_line, noqa): | |
| 1066 """ | |
| 1067 Comparisons to singletons like None should always be done | |
| 1068 with "is" or "is not", never the equality operators. | |
| 1069 | |
| 1070 Okay: if arg is not None: | |
| 1071 E711: if arg != None: | |
| 1072 E712: if arg == True: | |
| 1073 | |
| 1074 Also, beware of writing if x when you really mean if x is not None -- | |
| 1075 e.g. when testing whether a variable or argument that defaults to None was | |
| 1076 set to some other value. The other value might have a type (such as a | |
| 1077 container) that could be false in a boolean context! | |
| 1078 """ | |
| 1079 match = not noqa and COMPARE_SINGLETON_REGEX.search(logical_line) | |
| 1080 if match: | |
| 1081 same = (match.group(1) == '==') | |
| 1082 singleton = match.group(2) | |
| 1083 msg = "'if cond is %s:'" % (('' if same else 'not ') + singleton) | |
| 1084 if singleton in ('None',): | |
| 1085 code = 'E711' | |
| 1086 else: | |
| 1087 code = 'E712' | |
| 1088 nonzero = ((singleton == 'True' and same) or | |
| 1089 (singleton == 'False' and not same)) | |
| 1090 msg += " or 'if %scond:'" % ('' if nonzero else 'not ') | |
| 1091 yield match.start(1), code, singleton, msg | |
| 1092 | |
| 1093 | |
| 1094 def comparison_type(logical_line): | |
| 1095 """ | |
| 1096 Object type comparisons should always use isinstance() instead of | |
| 1097 comparing types directly. | |
| 1098 | |
| 1099 Okay: if isinstance(obj, int): | |
| 1100 E721: if type(obj) is type(1): | |
| 1101 | |
| 1102 When checking if an object is a string, keep in mind that it might be a | |
| 1103 unicode string too! In Python 2.3, str and unicode have a common base | |
| 1104 class, basestring, so you can do: | |
| 1105 | |
| 1106 Okay: if isinstance(obj, basestring): | |
| 1107 Okay: if type(a1) is type(b1): | |
| 1108 """ | |
| 1109 match = COMPARE_TYPE_REGEX.search(logical_line) | |
| 1110 if match: | |
| 1111 inst = match.group(1) | |
| 1112 if inst and isidentifier(inst) and inst not in SINGLETONS: | |
| 1113 return # Allow comparison for types which are not obvious | |
| 1114 yield match.start(), "E721" | |
| 1115 | |
| 1116 | |
| 1117 def python_3000_has_key(logical_line): | |
| 1118 r""" | |
| 1119 The {}.has_key() method is removed in the Python 3. | |
| 1120 Use the 'in' operation instead. | |
| 1121 | |
| 1122 Okay: if "alph" in d:\n print d["alph"] | |
| 1123 W601: assert d.has_key('alph') | |
| 1124 """ | |
| 1125 pos = logical_line.find('.has_key(') | |
| 1126 if pos > -1: | |
| 1127 yield pos, "W601" | |
| 1128 | |
| 1129 | |
| 1130 def python_3000_raise_comma(logical_line): | |
| 1131 """ | |
| 1132 When raising an exception, use "raise ValueError('message')" | |
| 1133 instead of the older form "raise ValueError, 'message'". | |
| 1134 | |
| 1135 The paren-using form is preferred because when the exception arguments | |
| 1136 are long or include string formatting, you don't need to use line | |
| 1137 continuation characters thanks to the containing parentheses. The older | |
| 1138 form is removed in Python 3. | |
| 1139 | |
| 1140 Okay: raise DummyError("Message") | |
| 1141 W602: raise DummyError, "Message" | |
| 1142 """ | |
| 1143 match = RAISE_COMMA_REGEX.match(logical_line) | |
| 1144 if match and not RERAISE_COMMA_REGEX.match(logical_line): | |
| 1145 yield match.end() - 1, "W602" | |
| 1146 | |
| 1147 | |
| 1148 def python_3000_not_equal(logical_line): | |
| 1149 """ | |
| 1150 != can also be written <>, but this is an obsolete usage kept for | |
| 1151 backwards compatibility only. New code should always use !=. | |
| 1152 The older syntax is removed in Python 3. | |
| 1153 | |
| 1154 Okay: if a != 'no': | |
| 1155 W603: if a <> 'no': | |
| 1156 """ | |
| 1157 pos = logical_line.find('<>') | |
| 1158 if pos > -1: | |
| 1159 yield pos, "W603" | |
| 1160 | |
| 1161 | |
| 1162 def python_3000_backticks(logical_line): | |
| 1163 """ | |
| 1164 Backticks are removed in Python 3. | |
| 1165 Use repr() instead. | |
| 1166 | |
| 1167 Okay: val = repr(1 + 2) | |
| 1168 W604: val = `1 + 2` | |
| 1169 """ | |
| 1170 pos = logical_line.find('`') | |
| 1171 if pos > -1: | |
| 1172 yield pos, "W604" | |
| 1173 | |
| 1174 | |
| 1175 ############################################################################## | |
| 1176 # Helper functions | |
| 1177 ############################################################################## | |
| 1178 | |
| 1179 | |
| 1180 if '' == ''.encode(): | |
| 1181 # Python 2: implicit encoding. | |
| 1182 def readlines(filename): | |
| 1183 f = open(filename) | |
| 1184 try: | |
| 1185 return f.readlines() | |
| 1186 finally: | |
| 1187 f.close() | |
| 1188 isidentifier = re.compile(r'[a-zA-Z_]\w*').match | |
| 1189 stdin_get_value = sys.stdin.read | |
| 1190 else: | |
| 1191 # Python 3 | |
| 1192 def readlines(filename): # __IGNORE_WARNING__ | |
| 1193 f = open(filename, 'rb') | |
| 1194 try: | |
| 1195 coding, lines = tokenize.detect_encoding(f.readline) | |
| 1196 f = TextIOWrapper(f, coding, line_buffering=True) | |
| 1197 return [l.decode(coding) for l in lines] + f.readlines() | |
| 1198 except (LookupError, SyntaxError, UnicodeError): | |
| 1199 f.close() | |
| 1200 # Fall back if files are improperly declared | |
| 1201 f = open(filename, encoding='latin-1') | |
| 1202 return f.readlines() | |
| 1203 finally: | |
| 1204 f.close() | |
| 1205 isidentifier = str.isidentifier | |
| 1206 | |
| 1207 def stdin_get_value(): | |
| 1208 return TextIOWrapper(sys.stdin.buffer, errors='ignore').read() | |
| 1209 readlines.__doc__ = " Read the source code." | |
| 1210 noqa = re.compile(r'# no(?:qa|pep8)\b', re.I).search | |
| 1211 | |
| 1212 | |
| 1213 def expand_indent(line): | |
| 1214 r""" | |
| 1215 Return the amount of indentation. | |
| 1216 Tabs are expanded to the next multiple of 8. | |
| 1217 | |
| 1218 >>> expand_indent(' ') | |
| 1219 4 | |
| 1220 >>> expand_indent('\t') | |
| 1221 8 | |
| 1222 >>> expand_indent(' \t') | |
| 1223 8 | |
| 1224 >>> expand_indent(' \t') | |
| 1225 8 | |
| 1226 >>> expand_indent(' \t') | |
| 1227 16 | |
| 1228 """ | |
| 1229 if '\t' not in line: | |
| 1230 return len(line) - len(line.lstrip()) | |
| 1231 result = 0 | |
| 1232 for char in line: | |
| 1233 if char == '\t': | |
| 1234 result = result // 8 * 8 + 8 | |
| 1235 elif char == ' ': | |
| 1236 result += 1 | |
| 1237 else: | |
| 1238 break | |
| 1239 return result | |
| 1240 | |
| 1241 | |
| 1242 def mute_string(text): | |
| 1243 """ | |
| 1244 Replace contents with 'xxx' to prevent syntax matching. | |
| 1245 | |
| 1246 >>> mute_string('"abc"') | |
| 1247 '"xxx"' | |
| 1248 >>> mute_string("'''abc'''") | |
| 1249 "'''xxx'''" | |
| 1250 >>> mute_string("r'abc'") | |
| 1251 "r'xxx'" | |
| 1252 """ | |
| 1253 # String modifiers (e.g. u or r) | |
| 1254 start = text.index(text[-1]) + 1 | |
| 1255 end = len(text) - 1 | |
| 1256 # Triple quotes | |
| 1257 if text[-3:] in ('"""', "'''"): | |
| 1258 start += 2 | |
| 1259 end -= 2 | |
| 1260 return text[:start] + 'x' * (end - start) + text[end:] | |
| 1261 | |
| 1262 | |
| 1263 def parse_udiff(diff, patterns=None, parent='.'): | |
| 1264 """Return a dictionary of matching lines.""" | |
| 1265 # For each file of the diff, the entry key is the filename, | |
| 1266 # and the value is a set of row numbers to consider. | |
| 1267 rv = {} | |
| 1268 path = nrows = None | |
| 1269 for line in diff.splitlines(): | |
| 1270 if nrows: | |
| 1271 if line[:1] != '-': | |
| 1272 nrows -= 1 | |
| 1273 continue | |
| 1274 if line[:3] == '@@ ': | |
| 1275 hunk_match = HUNK_REGEX.match(line) | |
| 1276 row, nrows = [int(g or '1') for g in hunk_match.groups()] | |
| 1277 rv[path].update(range(row, row + nrows)) | |
| 1278 elif line[:3] == '+++': | |
| 1279 path = line[4:].split('\t', 1)[0] | |
| 1280 if path[:2] == 'b/': | |
| 1281 path = path[2:] | |
| 1282 rv[path] = set() | |
| 1283 return dict([(os.path.join(parent, path), rows) | |
| 1284 for (path, rows) in rv.items() | |
| 1285 if rows and filename_match(path, patterns)]) | |
| 1286 | |
| 1287 | |
| 1288 def filename_match(filename, patterns, default=True): | |
| 1289 """ | |
| 1290 Check if patterns contains a pattern that matches filename. | |
| 1291 If patterns is unspecified, this always returns True. | |
| 1292 """ | |
| 1293 if not patterns: | |
| 1294 return default | |
| 1295 return any(fnmatch(filename, pattern) for pattern in patterns) | |
| 1296 | |
| 1297 | |
| 1298 ############################################################################## | |
| 1299 # Framework to run all checks | |
| 1300 ############################################################################## | |
| 1301 | |
| 1302 | |
| 1303 _checks = {'physical_line': {}, 'logical_line': {}, 'tree': {}} | |
| 1304 | |
| 1305 | |
| 1306 def register_check(check, codes=None): | |
| 1307 """ | |
| 1308 Register a new check object. | |
| 1309 """ | |
| 1310 def _add_check(check, kind, codes, args): | |
| 1311 if check in _checks[kind]: | |
| 1312 _checks[kind][check][0].extend(codes or []) | |
| 1313 else: | |
| 1314 _checks[kind][check] = (codes or [''], args) | |
| 1315 if inspect.isfunction(check): | |
| 1316 args = inspect.getargspec(check)[0] | |
| 1317 if args and args[0] in ('physical_line', 'logical_line'): | |
| 1318 if codes is None: | |
| 1319 codes = ERRORCODE_REGEX.findall(check.__doc__ or '') | |
| 1320 _add_check(check, args[0], codes, args) | |
| 1321 elif inspect.isclass(check): | |
| 1322 if inspect.getargspec(check.__init__)[0][:2] == ['self', 'tree']: | |
| 1323 _add_check(check, 'tree', codes, None) | |
| 1324 | |
| 1325 | |
| 1326 def init_checks_registry(): | |
| 1327 """ | |
| 1328 Register all globally visible functions where the first argument name | |
| 1329 is 'physical_line' or 'logical_line'. | |
| 1330 """ | |
| 1331 mod = inspect.getmodule(register_check) | |
| 1332 for (name, function) in inspect.getmembers(mod, inspect.isfunction): | |
| 1333 register_check(function) | |
| 1334 init_checks_registry() | |
| 1335 | |
| 1336 | |
| 1337 class Checker(object): | |
| 1338 """ | |
| 1339 Load a Python source file, tokenize it, check coding style. | |
| 1340 """ | |
| 1341 | |
| 1342 def __init__(self, filename=None, lines=None, | |
| 1343 options=None, report=None, **kwargs): | |
| 1344 if options is None: | |
| 1345 options = StyleGuide(kwargs).options | |
| 1346 else: | |
| 1347 assert not kwargs | |
| 1348 self._io_error = None | |
| 1349 self._physical_checks = options.physical_checks | |
| 1350 self._logical_checks = options.logical_checks | |
| 1351 self._ast_checks = options.ast_checks | |
| 1352 self.max_line_length = options.max_line_length | |
| 1353 self.hang_closing = options.hang_closing | |
| 1354 self.verbose = options.verbose | |
| 1355 self.filename = filename | |
| 1356 if filename is None: | |
| 1357 self.filename = 'stdin' | |
| 1358 self.lines = lines or [] | |
| 1359 elif filename == '-': | |
| 1360 self.filename = 'stdin' | |
| 1361 self.lines = stdin_get_value().splitlines(True) | |
| 1362 elif lines is None: | |
| 1363 try: | |
| 1364 self.lines = readlines(filename) | |
| 1365 except IOError: | |
| 1366 exc_type, exc = sys.exc_info()[:2] | |
| 1367 self._io_error = '%s: %s' % (exc_type.__name__, exc) | |
| 1368 self.lines = [] | |
| 1369 else: | |
| 1370 self.lines = lines | |
| 1371 if self.lines: | |
| 1372 ord0 = ord(self.lines[0][0]) | |
| 1373 if ord0 in (0xef, 0xfeff): # Strip the UTF-8 BOM | |
| 1374 if ord0 == 0xfeff: | |
| 1375 self.lines[0] = self.lines[0][1:] | |
| 1376 elif self.lines[0][:3] == '\xef\xbb\xbf': | |
| 1377 self.lines[0] = self.lines[0][3:] | |
| 1378 self.report = report or options.report | |
| 1379 self.report_error = self.report.error | |
| 1380 self.report_error_args = self.report.error_args | |
| 1381 | |
| 1382 # added for eric5 integration | |
| 1383 self.options = options | |
| 1384 | |
| 1385 def report_invalid_syntax(self): | |
| 1386 exc_type, exc = sys.exc_info()[:2] | |
| 1387 if len(exc.args) > 1: | |
| 1388 offset = exc.args[1] | |
| 1389 if len(offset) > 2: | |
| 1390 offset = offset[1:3] | |
| 1391 else: | |
| 1392 offset = (1, 0) | |
| 1393 self.report_error_args(offset[0], offset[1] or 0, | |
| 1394 'E901', self.report_invalid_syntax, | |
| 1395 exc_type.__name__, exc.args[0]) | |
| 1396 report_invalid_syntax.__doc__ = " Check if the syntax is valid." | |
| 1397 | |
| 1398 def readline(self): | |
| 1399 """ | |
| 1400 Get the next line from the input buffer. | |
| 1401 """ | |
| 1402 self.line_number += 1 | |
| 1403 if self.line_number > len(self.lines): | |
| 1404 return '' | |
| 1405 return self.lines[self.line_number - 1] | |
| 1406 | |
| 1407 def readline_check_physical(self): | |
| 1408 """ | |
| 1409 Check and return the next physical line. This method can be | |
| 1410 used to feed tokenize.generate_tokens. | |
| 1411 """ | |
| 1412 line = self.readline() | |
| 1413 if line: | |
| 1414 self.check_physical(line) | |
| 1415 return line | |
| 1416 | |
| 1417 def run_check(self, check, argument_names): | |
| 1418 """ | |
| 1419 Run a check plugin. | |
| 1420 """ | |
| 1421 arguments = [] | |
| 1422 for name in argument_names: | |
| 1423 arguments.append(getattr(self, name)) | |
| 1424 return check(*arguments) | |
| 1425 | |
| 1426 def check_physical(self, line): | |
| 1427 """ | |
| 1428 Run all physical checks on a raw input line. | |
| 1429 """ | |
| 1430 self.physical_line = line | |
| 1431 if self.indent_char is None and line[:1] in WHITESPACE: | |
| 1432 self.indent_char = line[0] | |
| 1433 for name, check, argument_names in self._physical_checks: | |
| 1434 result = self.run_check(check, argument_names) | |
| 1435 if result is not None: | |
| 1436 offset, code, *args = result | |
| 1437 self.report_error_args(self.line_number, offset, code, check, | |
| 1438 *args) | |
| 1439 | |
| 1440 def build_tokens_line(self): | |
| 1441 """ | |
| 1442 Build a logical line from tokens. | |
| 1443 """ | |
| 1444 self.mapping = [] | |
| 1445 logical = [] | |
| 1446 comments = [] | |
| 1447 length = 0 | |
| 1448 previous = None | |
| 1449 for token in self.tokens: | |
| 1450 token_type, text = token[0:2] | |
| 1451 if token_type == tokenize.COMMENT: | |
| 1452 comments.append(text) | |
| 1453 continue | |
| 1454 if token_type in SKIP_TOKENS: | |
| 1455 continue | |
| 1456 if token_type == tokenize.STRING: | |
| 1457 text = mute_string(text) | |
| 1458 if previous: | |
| 1459 end_row, end = previous[3] | |
| 1460 start_row, start = token[2] | |
| 1461 if end_row != start_row: # different row | |
| 1462 prev_text = self.lines[end_row - 1][end - 1] | |
| 1463 if prev_text == ',' or (prev_text not in '{[(' | |
| 1464 and text not in '}])'): | |
| 1465 logical.append(' ') | |
| 1466 length += 1 | |
| 1467 elif end != start: # different column | |
| 1468 fill = self.lines[end_row - 1][end:start] | |
| 1469 logical.append(fill) | |
| 1470 length += len(fill) | |
| 1471 self.mapping.append((length, token)) | |
| 1472 logical.append(text) | |
| 1473 length += len(text) | |
| 1474 previous = token | |
| 1475 self.logical_line = ''.join(logical) | |
| 1476 self.noqa = comments and noqa(''.join(comments)) | |
| 1477 # With Python 2, if the line ends with '\r\r\n' the assertion fails | |
| 1478 # assert self.logical_line.strip() == self.logical_line | |
| 1479 | |
| 1480 def check_logical(self): | |
| 1481 """ | |
| 1482 Build a line from tokens and run all logical checks on it. | |
| 1483 """ | |
| 1484 self.build_tokens_line() | |
| 1485 self.report.increment_logical_line() | |
| 1486 first_line = self.lines[self.mapping[0][1][2][0] - 1] | |
| 1487 indent = first_line[:self.mapping[0][1][2][1]] | |
| 1488 self.previous_indent_level = self.indent_level | |
| 1489 self.indent_level = expand_indent(indent) | |
| 1490 if self.verbose >= 2: | |
| 1491 print(self.logical_line[:80].rstrip()) | |
| 1492 for name, check, argument_names in self._logical_checks: | |
| 1493 if self.verbose >= 4: | |
| 1494 print(' ' + name) | |
| 1495 for result in self.run_check(check, argument_names): | |
| 1496 offset, code, *args = result | |
| 1497 if isinstance(offset, tuple): | |
| 1498 orig_number, orig_offset = offset | |
| 1499 else: | |
| 1500 for token_offset, token in self.mapping: | |
| 1501 if offset >= token_offset: | |
| 1502 orig_number = token[2][0] | |
| 1503 orig_offset = (token[2][1] + offset - token_offset) | |
| 1504 self.report_error_args(orig_number, orig_offset, code, check, | |
| 1505 *args) | |
| 1506 self.previous_logical = self.logical_line | |
| 1507 | |
| 1508 def check_ast(self): | |
| 1509 try: | |
| 1510 tree = compile(''.join(self.lines), '', 'exec', PyCF_ONLY_AST) | |
| 1511 except (SyntaxError, TypeError): | |
| 1512 return self.report_invalid_syntax() | |
| 1513 for name, cls, _ in self._ast_checks: | |
| 1514 # extended API for eric5 integration | |
| 1515 checker = cls(tree, self.filename, self.options) | |
| 1516 for lineno, offset, code, check, *args in checker.run(): | |
| 1517 if not noqa(self.lines[lineno - 1]): | |
| 1518 self.report_error_args(lineno, offset, code, check, *args) | |
| 1519 | |
| 1520 def generate_tokens(self): | |
| 1521 if self._io_error: | |
| 1522 self.report_error(1, 0, 'E902 %s' % self._io_error, readlines) | |
| 1523 tokengen = tokenize.generate_tokens(self.readline_check_physical) | |
| 1524 try: | |
| 1525 for token in tokengen: | |
| 1526 yield token | |
| 1527 except (SyntaxError, tokenize.TokenError): | |
| 1528 self.report_invalid_syntax() | |
| 1529 | |
| 1530 def check_all(self, expected=None, line_offset=0): | |
| 1531 """ | |
| 1532 Run all checks on the input file. | |
| 1533 """ | |
| 1534 self.report.init_file(self.filename, self.lines, expected, line_offset) | |
| 1535 if self._ast_checks: | |
| 1536 self.check_ast() | |
| 1537 self.line_number = 0 | |
| 1538 self.indent_char = None | |
| 1539 self.indent_level = 0 | |
| 1540 self.previous_logical = '' | |
| 1541 self.tokens = [] | |
| 1542 self.blank_lines = blank_lines_before_comment = 0 | |
| 1543 parens = 0 | |
| 1544 for token in self.generate_tokens(): | |
| 1545 self.tokens.append(token) | |
| 1546 token_type, text = token[0:2] | |
| 1547 if self.verbose >= 3: | |
| 1548 if token[2][0] == token[3][0]: | |
| 1549 pos = '[%s:%s]' % (token[2][1] or '', token[3][1]) | |
| 1550 else: | |
| 1551 pos = 'l.%s' % token[3][0] | |
| 1552 print('l.%s\t%s\t%s\t%r' % | |
| 1553 (token[2][0], pos, tokenize.tok_name[token[0]], text)) | |
| 1554 if token_type == tokenize.OP: | |
| 1555 if text in '([{': | |
| 1556 parens += 1 | |
| 1557 elif text in '}])': | |
| 1558 parens -= 1 | |
| 1559 elif not parens: | |
| 1560 if token_type == tokenize.NEWLINE: | |
| 1561 if self.blank_lines < blank_lines_before_comment: | |
| 1562 self.blank_lines = blank_lines_before_comment | |
| 1563 self.check_logical() | |
| 1564 self.tokens = [] | |
| 1565 self.blank_lines = blank_lines_before_comment = 0 | |
| 1566 elif token_type == tokenize.NL: | |
| 1567 if len(self.tokens) == 1: | |
| 1568 # The physical line contains only this token. | |
| 1569 self.blank_lines += 1 | |
| 1570 self.tokens = [] | |
| 1571 elif token_type == tokenize.COMMENT and len(self.tokens) == 1: | |
| 1572 if blank_lines_before_comment < self.blank_lines: | |
| 1573 blank_lines_before_comment = self.blank_lines | |
| 1574 self.blank_lines = 0 | |
| 1575 if COMMENT_WITH_NL: | |
| 1576 # The comment also ends a physical line | |
| 1577 self.tokens = [] | |
| 1578 return self.report.get_file_results() | |
| 1579 | |
| 1580 | |
| 1581 class BaseReport(object): | |
| 1582 """Collect the results of the checks.""" | |
| 1583 print_filename = False | |
| 1584 | |
| 1585 def __init__(self, options): | |
| 1586 self._benchmark_keys = options.benchmark_keys | |
| 1587 self._ignore_code = options.ignore_code | |
| 1588 # Results | |
| 1589 self.elapsed = 0 | |
| 1590 self.total_errors = 0 | |
| 1591 self.counters = dict.fromkeys(self._benchmark_keys, 0) | |
| 1592 self.messages = {} | |
| 1593 | |
| 1594 def start(self): | |
| 1595 """Start the timer.""" | |
| 1596 self._start_time = time.time() | |
| 1597 | |
| 1598 def stop(self): | |
| 1599 """Stop the timer.""" | |
| 1600 self.elapsed = time.time() - self._start_time | |
| 1601 | |
| 1602 def init_file(self, filename, lines, expected, line_offset): | |
| 1603 """Signal a new file.""" | |
| 1604 self.filename = filename | |
| 1605 self.lines = lines | |
| 1606 self.expected = expected or () | |
| 1607 self.line_offset = line_offset | |
| 1608 self.file_errors = 0 | |
| 1609 self.counters['files'] += 1 | |
| 1610 self.counters['physical lines'] += len(lines) | |
| 1611 | |
| 1612 def increment_logical_line(self): | |
| 1613 """Signal a new logical line.""" | |
| 1614 self.counters['logical lines'] += 1 | |
| 1615 | |
| 1616 def error(self, line_number, offset, text, check): | |
| 1617 """Report an error, according to options.""" | |
| 1618 code = text[:4] | |
| 1619 if self._ignore_code(code): | |
| 1620 return | |
| 1621 if code in self.counters: | |
| 1622 self.counters[code] += 1 | |
| 1623 else: | |
| 1624 self.counters[code] = 1 | |
| 1625 self.messages[code] = text[5:] | |
| 1626 # Don't care about expected errors or warnings | |
| 1627 if code in self.expected: | |
| 1628 return | |
| 1629 if self.print_filename and not self.file_errors: | |
| 1630 print(self.filename) | |
| 1631 self.file_errors += 1 | |
| 1632 self.total_errors += 1 | |
| 1633 return code | |
| 1634 | |
| 1635 def error_args(self, line_number, offset, code, check, *args): | |
| 1636 """Report an error, according to options.""" | |
| 1637 if self._ignore_code(code): | |
| 1638 return | |
| 1639 text = getMessage(code, *args) | |
| 1640 if code in self.counters: | |
| 1641 self.counters[code] += 1 | |
| 1642 else: | |
| 1643 self.counters[code] = 1 | |
| 1644 self.messages[code] = text[5:] | |
| 1645 # Don't care about expected errors or warnings | |
| 1646 if code in self.expected: | |
| 1647 return | |
| 1648 if self.print_filename and not self.file_errors: | |
| 1649 print(self.filename) | |
| 1650 self.file_errors += 1 | |
| 1651 self.total_errors += 1 | |
| 1652 return code | |
| 1653 | |
| 1654 def get_file_results(self): | |
| 1655 """Return the count of errors and warnings for this file.""" | |
| 1656 return self.file_errors | |
| 1657 | |
| 1658 def get_count(self, prefix=''): | |
| 1659 """Return the total count of errors and warnings.""" | |
| 1660 return sum([self.counters[key] | |
| 1661 for key in self.messages if key.startswith(prefix)]) | |
| 1662 | |
| 1663 def get_statistics(self, prefix=''): | |
| 1664 """ | |
| 1665 Get statistics for message codes that start with the prefix. | |
| 1666 | |
| 1667 prefix='' matches all errors and warnings | |
| 1668 prefix='E' matches all errors | |
| 1669 prefix='W' matches all warnings | |
| 1670 prefix='E4' matches all errors that have to do with imports | |
| 1671 """ | |
| 1672 return ['%-7s %s %s' % (self.counters[key], key, self.messages[key]) | |
| 1673 for key in sorted(self.messages) if key.startswith(prefix)] | |
| 1674 | |
| 1675 def print_statistics(self, prefix=''): | |
| 1676 """Print overall statistics (number of errors and warnings).""" | |
| 1677 for line in self.get_statistics(prefix): | |
| 1678 print(line) | |
| 1679 | |
| 1680 def print_benchmark(self): | |
| 1681 """Print benchmark numbers.""" | |
| 1682 print('%-7.2f %s' % (self.elapsed, 'seconds elapsed')) | |
| 1683 if self.elapsed: | |
| 1684 for key in self._benchmark_keys: | |
| 1685 print('%-7d %s per second (%d total)' % | |
| 1686 (self.counters[key] / self.elapsed, key, | |
| 1687 self.counters[key])) | |
| 1688 | |
| 1689 | |
| 1690 class FileReport(BaseReport): | |
| 1691 """Collect the results of the checks and print only the filenames.""" | |
| 1692 print_filename = True | |
| 1693 | |
| 1694 | |
| 1695 class StandardReport(BaseReport): | |
| 1696 """Collect and print the results of the checks.""" | |
| 1697 | |
| 1698 def __init__(self, options): | |
| 1699 super(StandardReport, self).__init__(options) | |
| 1700 self._fmt = REPORT_FORMAT.get(options.format.lower(), | |
| 1701 options.format) | |
| 1702 self._repeat = options.repeat | |
| 1703 self._show_source = options.show_source | |
| 1704 self._show_pep8 = options.show_pep8 | |
| 1705 | |
| 1706 def init_file(self, filename, lines, expected, line_offset): | |
| 1707 """Signal a new file.""" | |
| 1708 self._deferred_print = [] | |
| 1709 return super(StandardReport, self).init_file( | |
| 1710 filename, lines, expected, line_offset) | |
| 1711 | |
| 1712 def error(self, line_number, offset, text, check): | |
| 1713 """Report an error, according to options.""" | |
| 1714 code = super(StandardReport, self).error(line_number, offset, | |
| 1715 text, check) | |
| 1716 if code and (self.counters[code] == 1 or self._repeat): | |
| 1717 self._deferred_print.append( | |
| 1718 (line_number, offset, code, text[5:], check.__doc__)) | |
| 1719 return code | |
| 1720 | |
| 1721 def error_args(self, line_number, offset, code, check, *args): | |
| 1722 """Report an error, according to options.""" | |
| 1723 code = super(StandardReport, self).error_args(line_number, offset, | |
| 1724 code, check, *args) | |
| 1725 if code and (self.counters[code] == 1 or self._repeat): | |
| 1726 text = getMessage(code, *args) | |
| 1727 self._deferred_print.append( | |
| 1728 (line_number, offset, code, text[5:], check.__doc__)) | |
| 1729 return code | |
| 1730 | |
| 1731 def get_file_results(self): | |
| 1732 """Print the result and return the overall count for this file.""" | |
| 1733 self._deferred_print.sort() | |
| 1734 for line_number, offset, code, text, doc in self._deferred_print: | |
| 1735 print(self._fmt % { | |
| 1736 'path': self.filename, | |
| 1737 'row': self.line_offset + line_number, 'col': offset + 1, | |
| 1738 'code': code, 'text': text, | |
| 1739 }) | |
| 1740 if self._show_source: | |
| 1741 if line_number > len(self.lines): | |
| 1742 line = '' | |
| 1743 else: | |
| 1744 line = self.lines[line_number - 1] | |
| 1745 print(line.rstrip()) | |
| 1746 print(' ' * offset + '^') | |
| 1747 if self._show_pep8 and doc: | |
| 1748 print(doc.lstrip('\n').rstrip()) | |
| 1749 return self.file_errors | |
| 1750 | |
| 1751 | |
| 1752 class DiffReport(StandardReport): | |
| 1753 """Collect and print the results for the changed lines only.""" | |
| 1754 | |
| 1755 def __init__(self, options): | |
| 1756 super(DiffReport, self).__init__(options) | |
| 1757 self._selected = options.selected_lines | |
| 1758 | |
| 1759 def error(self, line_number, offset, text, check): | |
| 1760 if line_number not in self._selected[self.filename]: | |
| 1761 return | |
| 1762 return super(DiffReport, self).error(line_number, offset, text, check) | |
| 1763 | |
| 1764 | |
| 1765 class StyleGuide(object): | |
| 1766 """Initialize a PEP-8 instance with few options.""" | |
| 1767 | |
| 1768 def __init__(self, *args, **kwargs): | |
| 1769 # build options from the command line | |
| 1770 self.checker_class = kwargs.pop('checker_class', Checker) | |
| 1771 parse_argv = kwargs.pop('parse_argv', False) | |
| 1772 config_file = kwargs.pop('config_file', None) | |
| 1773 parser = kwargs.pop('parser', None) | |
| 1774 options, self.paths = process_options( | |
| 1775 parse_argv=parse_argv, config_file=config_file, parser=parser) | |
| 1776 if args or kwargs: | |
| 1777 # build options from dict | |
| 1778 options_dict = dict(*args, **kwargs) | |
| 1779 options.__dict__.update(options_dict) | |
| 1780 if 'paths' in options_dict: | |
| 1781 self.paths = options_dict['paths'] | |
| 1782 | |
| 1783 self.runner = self.input_file | |
| 1784 self.options = options | |
| 1785 | |
| 1786 if not options.reporter: | |
| 1787 options.reporter = BaseReport if options.quiet else StandardReport | |
| 1788 | |
| 1789 for index, value in enumerate(options.exclude): | |
| 1790 options.exclude[index] = value.rstrip('/') | |
| 1791 options.select = tuple(options.select or ()) | |
| 1792 if not (options.select or options.ignore or | |
| 1793 options.testsuite or options.doctest) and DEFAULT_IGNORE: | |
| 1794 # The default choice: ignore controversial checks | |
| 1795 options.ignore = tuple(DEFAULT_IGNORE.split(',')) | |
| 1796 else: | |
| 1797 # Ignore all checks which are not explicitly selected | |
| 1798 options.ignore = ('',) if options.select else tuple(options.ignore) | |
| 1799 options.benchmark_keys = BENCHMARK_KEYS[:] | |
| 1800 options.ignore_code = self.ignore_code | |
| 1801 options.physical_checks = self.get_checks('physical_line') | |
| 1802 options.logical_checks = self.get_checks('logical_line') | |
| 1803 options.ast_checks = self.get_checks('tree') | |
| 1804 self.init_report() | |
| 1805 | |
| 1806 def init_report(self, reporter=None): | |
| 1807 """Initialize the report instance.""" | |
| 1808 self.options.report = (reporter or self.options.reporter)(self.options) | |
| 1809 return self.options.report | |
| 1810 | |
| 1811 def check_files(self, paths=None): | |
| 1812 """Run all checks on the paths.""" | |
| 1813 if paths is None: | |
| 1814 paths = self.paths | |
| 1815 report = self.options.report | |
| 1816 runner = self.runner | |
| 1817 report.start() | |
| 1818 try: | |
| 1819 for path in paths: | |
| 1820 if os.path.isdir(path): | |
| 1821 self.input_dir(path) | |
| 1822 elif not self.excluded(path): | |
| 1823 runner(path) | |
| 1824 except KeyboardInterrupt: | |
| 1825 print('... stopped') | |
| 1826 report.stop() | |
| 1827 return report | |
| 1828 | |
| 1829 def input_file(self, filename, lines=None, expected=None, line_offset=0): | |
| 1830 """Run all checks on a Python source file.""" | |
| 1831 if self.options.verbose: | |
| 1832 print('checking %s' % filename) | |
| 1833 fchecker = self.checker_class( | |
| 1834 filename, lines=lines, options=self.options) | |
| 1835 return fchecker.check_all(expected=expected, line_offset=line_offset) | |
| 1836 | |
| 1837 def input_dir(self, dirname): | |
| 1838 """Check all files in this directory and all subdirectories.""" | |
| 1839 dirname = dirname.rstrip('/') | |
| 1840 if self.excluded(dirname): | |
| 1841 return 0 | |
| 1842 counters = self.options.report.counters | |
| 1843 verbose = self.options.verbose | |
| 1844 filepatterns = self.options.filename | |
| 1845 runner = self.runner | |
| 1846 for root, dirs, files in os.walk(dirname): | |
| 1847 if verbose: | |
| 1848 print('directory ' + root) | |
| 1849 counters['directories'] += 1 | |
| 1850 for subdir in sorted(dirs): | |
| 1851 if self.excluded(subdir, root): | |
| 1852 dirs.remove(subdir) | |
| 1853 for filename in sorted(files): | |
| 1854 # contain a pattern that matches? | |
| 1855 if ((filename_match(filename, filepatterns) and | |
| 1856 not self.excluded(filename, root))): | |
| 1857 runner(os.path.join(root, filename)) | |
| 1858 | |
| 1859 def excluded(self, filename, parent=None): | |
| 1860 """ | |
| 1861 Check if options.exclude contains a pattern that matches filename. | |
| 1862 """ | |
| 1863 if not self.options.exclude: | |
| 1864 return False | |
| 1865 basename = os.path.basename(filename) | |
| 1866 if filename_match(basename, self.options.exclude): | |
| 1867 return True | |
| 1868 if parent: | |
| 1869 filename = os.path.join(parent, filename) | |
| 1870 return filename_match(filename, self.options.exclude) | |
| 1871 | |
| 1872 def ignore_code(self, code): | |
| 1873 """ | |
| 1874 Check if the error code should be ignored. | |
| 1875 | |
| 1876 If 'options.select' contains a prefix of the error code, | |
| 1877 return False. Else, if 'options.ignore' contains a prefix of | |
| 1878 the error code, return True. | |
| 1879 """ | |
| 1880 return (code.startswith(self.options.ignore) and | |
| 1881 not code.startswith(self.options.select)) | |
| 1882 | |
| 1883 def get_checks(self, argument_name): | |
| 1884 """ | |
| 1885 Find all globally visible functions where the first argument name | |
| 1886 starts with argument_name and which contain selected tests. | |
| 1887 """ | |
| 1888 checks = [] | |
| 1889 for check, attrs in _checks[argument_name].items(): | |
| 1890 (codes, args) = attrs | |
| 1891 if any(not (code and self.ignore_code(code)) for code in codes): | |
| 1892 checks.append((check.__name__, check, args)) | |
| 1893 return sorted(checks) | |
| 1894 | |
| 1895 | |
| 1896 def get_parser(prog='pep8', version=__version__): | |
| 1897 parser = OptionParser(prog=prog, version=version, | |
| 1898 usage="%prog [options] input ...") | |
| 1899 parser.config_options = [ | |
| 1900 'exclude', 'filename', 'select', 'ignore', 'max-line-length', | |
| 1901 'hang-closing', 'count', 'format', 'quiet', 'show-pep8', | |
| 1902 'show-source', 'statistics', 'verbose'] | |
| 1903 parser.add_option('-v', '--verbose', default=0, action='count', | |
| 1904 help="print status messages, or debug with -vv") | |
| 1905 parser.add_option('-q', '--quiet', default=0, action='count', | |
| 1906 help="report only file names, or nothing with -qq") | |
| 1907 parser.add_option('-r', '--repeat', default=True, action='store_true', | |
| 1908 help="(obsolete) show all occurrences of the same error") | |
| 1909 parser.add_option('--first', action='store_false', dest='repeat', | |
| 1910 help="show first occurrence of each error") | |
| 1911 parser.add_option('--exclude', metavar='patterns', default=DEFAULT_EXCLUDE, | |
| 1912 help="exclude files or directories which match these " | |
| 1913 "comma separated patterns (default: %default)") | |
| 1914 parser.add_option('--filename', metavar='patterns', default='*.py', | |
| 1915 help="when parsing directories, only check filenames " | |
| 1916 "matching these comma separated patterns " | |
| 1917 "(default: %default)") | |
| 1918 parser.add_option('--select', metavar='errors', default='', | |
| 1919 help="select errors and warnings (e.g. E,W6)") | |
| 1920 parser.add_option('--ignore', metavar='errors', default='', | |
| 1921 help="skip errors and warnings (e.g. E4,W)") | |
| 1922 parser.add_option('--show-source', action='store_true', | |
| 1923 help="show source code for each error") | |
| 1924 parser.add_option('--show-pep8', action='store_true', | |
| 1925 help="show text of PEP 8 for each error " | |
| 1926 "(implies --first)") | |
| 1927 parser.add_option('--statistics', action='store_true', | |
| 1928 help="count errors and warnings") | |
| 1929 parser.add_option('--count', action='store_true', | |
| 1930 help="print total number of errors and warnings " | |
| 1931 "to standard error and set exit code to 1 if " | |
| 1932 "total is not null") | |
| 1933 parser.add_option('--max-line-length', type='int', metavar='n', | |
| 1934 default=MAX_LINE_LENGTH, | |
| 1935 help="set maximum allowed line length " | |
| 1936 "(default: %default)") | |
| 1937 parser.add_option('--hang-closing', action='store_true', | |
| 1938 help="hang closing bracket instead of matching " | |
| 1939 "indentation of opening bracket's line") | |
| 1940 parser.add_option('--format', metavar='format', default='default', | |
| 1941 help="set the error format [default|pylint|<custom>]") | |
| 1942 parser.add_option('--diff', action='store_true', | |
| 1943 help="report only lines changed according to the " | |
| 1944 "unified diff received on STDIN") | |
| 1945 group = parser.add_option_group("Testing Options") | |
| 1946 if os.path.exists(TESTSUITE_PATH): | |
| 1947 group.add_option('--testsuite', metavar='dir', | |
| 1948 help="run regression tests from dir") | |
| 1949 group.add_option('--doctest', action='store_true', | |
| 1950 help="run doctest on myself") | |
| 1951 group.add_option('--benchmark', action='store_true', | |
| 1952 help="measure processing speed") | |
| 1953 return parser | |
| 1954 | |
| 1955 | |
| 1956 def read_config(options, args, arglist, parser): | |
| 1957 """Read both user configuration and local configuration.""" | |
| 1958 config = RawConfigParser() | |
| 1959 | |
| 1960 user_conf = options.config | |
| 1961 if user_conf and os.path.isfile(user_conf): | |
| 1962 if options.verbose: | |
| 1963 print('user configuration: %s' % user_conf) | |
| 1964 config.read(user_conf) | |
| 1965 | |
| 1966 parent = tail = args and os.path.abspath(os.path.commonprefix(args)) | |
| 1967 while tail: | |
| 1968 if config.read([os.path.join(parent, fn) for fn in PROJECT_CONFIG]): | |
| 1969 if options.verbose: | |
| 1970 print('local configuration: in %s' % parent) | |
| 1971 break | |
| 1972 parent, tail = os.path.split(parent) | |
| 1973 | |
| 1974 pep8_section = parser.prog | |
| 1975 if config.has_section(pep8_section): | |
| 1976 option_list = dict([(o.dest, o.type or o.action) | |
| 1977 for o in parser.option_list]) | |
| 1978 | |
| 1979 # First, read the default values | |
| 1980 new_options, _ = parser.parse_args([]) | |
| 1981 | |
| 1982 # Second, parse the configuration | |
| 1983 for opt in config.options(pep8_section): | |
| 1984 if options.verbose > 1: | |
| 1985 print(" %s = %s" % (opt, config.get(pep8_section, opt))) | |
| 1986 if opt.replace('_', '-') not in parser.config_options: | |
| 1987 print("Unknown option: '%s'\n not in [%s]" % | |
| 1988 (opt, ' '.join(parser.config_options))) | |
| 1989 sys.exit(1) | |
| 1990 normalized_opt = opt.replace('-', '_') | |
| 1991 opt_type = option_list[normalized_opt] | |
| 1992 if opt_type in ('int', 'count'): | |
| 1993 value = config.getint(pep8_section, opt) | |
| 1994 elif opt_type == 'string': | |
| 1995 value = config.get(pep8_section, opt) | |
| 1996 else: | |
| 1997 assert opt_type in ('store_true', 'store_false') | |
| 1998 value = config.getboolean(pep8_section, opt) | |
| 1999 setattr(new_options, normalized_opt, value) | |
| 2000 | |
| 2001 # Third, overwrite with the command-line options | |
| 2002 options, _ = parser.parse_args(arglist, values=new_options) | |
| 2003 options.doctest = options.testsuite = False | |
| 2004 return options | |
| 2005 | |
| 2006 | |
| 2007 def process_options(arglist=None, parse_argv=False, config_file=None, | |
| 2008 parser=None): | |
| 2009 """Process options passed either via arglist or via command line args.""" | |
| 2010 if not arglist and not parse_argv: | |
| 2011 # Don't read the command line if the module is used as a library. | |
| 2012 arglist = [] | |
| 2013 if not parser: | |
| 2014 parser = get_parser() | |
| 2015 if not parser.has_option('--config'): | |
| 2016 if config_file is True: | |
| 2017 config_file = DEFAULT_CONFIG | |
| 2018 group = parser.add_option_group("Configuration", description=( | |
| 2019 "The project options are read from the [%s] section of the " | |
| 2020 "tox.ini file or the setup.cfg file located in any parent folder " | |
| 2021 "of the path(s) being processed. Allowed options are: %s." % | |
| 2022 (parser.prog, ', '.join(parser.config_options)))) | |
| 2023 group.add_option('--config', metavar='path', default=config_file, | |
| 2024 help="user config file location (default: %default)") | |
| 2025 options, args = parser.parse_args(arglist) | |
| 2026 options.reporter = None | |
| 2027 | |
| 2028 if options.ensure_value('testsuite', False): | |
| 2029 args.append(options.testsuite) | |
| 2030 elif not options.ensure_value('doctest', False): | |
| 2031 if parse_argv and not args: | |
| 2032 if options.diff or any(os.path.exists(name) | |
| 2033 for name in PROJECT_CONFIG): | |
| 2034 args = ['.'] | |
| 2035 else: | |
| 2036 parser.error('input not specified') | |
| 2037 options = read_config(options, args, arglist, parser) | |
| 2038 options.reporter = parse_argv and options.quiet == 1 and FileReport | |
| 2039 | |
| 2040 options.filename = options.filename and options.filename.split(',') | |
| 2041 options.exclude = options.exclude.split(',') | |
| 2042 options.select = options.select and options.select.split(',') | |
| 2043 options.ignore = options.ignore and options.ignore.split(',') | |
| 2044 | |
| 2045 if options.diff: | |
| 2046 options.reporter = DiffReport | |
| 2047 stdin = stdin_get_value() | |
| 2048 options.selected_lines = parse_udiff(stdin, options.filename, args[0]) | |
| 2049 args = sorted(options.selected_lines) | |
| 2050 | |
| 2051 return options, args | |
| 2052 | |
| 2053 | |
| 2054 def _main(): | |
| 2055 """Parse options and run checks on Python source.""" | |
| 2056 pep8style = StyleGuide(parse_argv=True, config_file=True) | |
| 2057 options = pep8style.options | |
| 2058 if options.doctest or options.testsuite: | |
| 2059 from testsuite.support import run_tests | |
| 2060 report = run_tests(pep8style) | |
| 2061 else: | |
| 2062 report = pep8style.check_files() | |
| 2063 if options.statistics: | |
| 2064 report.print_statistics() | |
| 2065 if options.benchmark: | |
| 2066 report.print_benchmark() | |
| 2067 if options.testsuite and not options.quiet: | |
| 2068 report.print_results() | |
| 2069 if report.total_errors: | |
| 2070 if options.count: | |
| 2071 sys.stderr.write(str(report.total_errors) + '\n') | |
| 2072 sys.exit(1) | |
| 2073 | |
| 2074 if __name__ == '__main__': | |
| 2075 _main() |
