1
# Copyright (C) 2006-2011 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
"""Tools for converting globs to regular expressions.
19
This module provides functions for converting shell-like globs to regular
23
from __future__ import absolute_import
36
class Replacer(object):
37
"""Do a multiple-pattern substitution.
39
The patterns and substitutions are combined into one, so the result of
40
one replacement is never substituted again. Add the patterns and
41
replacements via the add method and then call the object. The patterns
42
must not contain capturing groups.
45
_expand = lazy_regex.lazy_compile(u'\\\\&')
47
def __init__(self, source=None):
50
self._pats = list(source._pats)
51
self._funs = list(source._funs)
56
def add(self, pat, fun):
57
r"""Add a pattern and replacement.
59
The pattern must not contain capturing groups.
60
The replacement might be either a string template in which \& will be
61
replaced with the match, or a function that will get the matching text
62
as argument. It does not get match object, because capturing is
66
self._pats.append(pat)
67
self._funs.append(fun)
69
def add_replacer(self, replacer):
70
r"""Add all patterns from another replacer.
72
All patterns and replacements from replacer are appended to the ones
76
self._pats.extend(replacer._pats)
77
self._funs.extend(replacer._funs)
79
def __call__(self, text):
81
self._pat = lazy_regex.lazy_compile(
82
u'|'.join([u'(%s)' % p for p in self._pats]),
84
return self._pat.sub(self._do_sub, text)
87
fun = self._funs[m.lastindex - 1]
88
if hasattr(fun, '__call__'):
89
return fun(m.group(0))
91
return self._expand.sub(m.group(0), fun)
94
_sub_named = Replacer()
95
_sub_named.add(r'\[:digit:\]', r'\d')
96
_sub_named.add(r'\[:space:\]', r'\s')
97
_sub_named.add(r'\[:alnum:\]', r'\w')
98
_sub_named.add(r'\[:ascii:\]', r'\0-\x7f')
99
_sub_named.add(r'\[:blank:\]', r' \t')
100
_sub_named.add(r'\[:cntrl:\]', r'\0-\x1f\x7f-\x9f')
104
if m[1] in (u'!', u'^'):
105
return u'[^' + _sub_named(m[2:-1]) + u']'
106
return u'[' + _sub_named(m[1:-1]) + u']'
109
def _invalid_regex(repl):
111
warning(u"'%s' not allowed within a regular expression. "
112
"Replacing with '%s'" % (m, repl))
117
def _trailing_backslashes_regex(m):
118
"""Check trailing backslashes.
120
Does a head count on trailing backslashes to ensure there isn't an odd
121
one on the end that would escape the brackets we wrap the RE in.
123
if (len(m) % 2) != 0:
124
warning(u"Regular expressions cannot end with an odd number of '\\'. "
125
"Dropping the final '\\'.")
131
_sub_re.add(u'^RE:', u'')
132
_sub_re.add(u'\\((?!\\?)', u'(?:')
133
_sub_re.add(u'\\(\\?P<.*>', _invalid_regex(u'(?:'))
134
_sub_re.add(u'\\(\\?P=[^)]*\\)', _invalid_regex(u''))
135
_sub_re.add(r'\\+$', _trailing_backslashes_regex)
138
_sub_fullpath = Replacer()
139
_sub_fullpath.add(r'^RE:.*', _sub_re) # RE:<anything> is a regex
140
_sub_fullpath.add(r'\[\^?\]?(?:[^][]|\[:[^]]+:\])+\]',
141
_sub_group) # char group
142
_sub_fullpath.add(r'(?:(?<=/)|^)(?:\.?/)+', u'') # canonicalize path
143
_sub_fullpath.add(r'\\.', r'\&') # keep anything backslashed
144
_sub_fullpath.add(r'[(){}|^$+.]', r'\\&') # escape specials
145
_sub_fullpath.add(r'(?:(?<=/)|^)\*\*+/', r'(?:.*/)?') # **/ after ^ or /
146
_sub_fullpath.add(r'\*+', r'[^/]*') # * elsewhere
147
_sub_fullpath.add(r'\?', r'[^/]') # ? everywhere
150
_sub_basename = Replacer()
151
_sub_basename.add(r'\[\^?\]?(?:[^][]|\[:[^]]+:\])+\]',
152
_sub_group) # char group
153
_sub_basename.add(r'\\.', r'\&') # keep anything backslashed
154
_sub_basename.add(r'[(){}|^$+.]', r'\\&') # escape specials
155
_sub_basename.add(r'\*+', r'.*') # * everywhere
156
_sub_basename.add(r'\?', r'.') # ? everywhere
159
def _sub_extension(pattern):
160
return _sub_basename(pattern[2:])
163
class Globster(object):
164
"""A simple wrapper for a set of glob patterns.
166
Provides the capability to search the patterns to find a match for
167
a given filename (including the full path).
169
Patterns are translated to regular expressions to expidite matching.
171
The regular expressions for multiple patterns are aggregated into
172
a super-regex containing groups of up to 99 patterns.
173
The 99 limitation is due to the grouping limit of the Python re module.
174
The resulting super-regex and associated patterns are stored as a list of
175
(regex,[patterns]) in _regex_patterns.
177
For performance reasons the patterns are categorised as extension patterns
178
(those that match against a file extension), basename patterns
179
(those that match against the basename of the filename),
180
and fullpath patterns (those that match against the full path).
181
The translations used for extensions and basenames are relatively simpler
182
and therefore faster to perform than the fullpath patterns.
184
Also, the extension patterns are more likely to find a match and
185
so are matched first, then the basename patterns, then the fullpath
188
# We want to _add_patterns in a specific order (as per type_list below)
189
# starting with the shortest and going to the longest.
190
# As some Python version don't support ordered dicts the list below is
191
# used to select inputs for _add_pattern in a specific order.
192
pattern_types = ["extension", "basename", "fullpath"]
196
"translator": _sub_extension,
197
"prefix": r'(?:.*/)?(?!.*/)(?:.*\.)'
200
"translator": _sub_basename,
201
"prefix": r'(?:.*/)?(?!.*/)'
204
"translator": _sub_fullpath,
209
def __init__(self, patterns):
210
self._regex_patterns = []
217
pat = normalize_pattern(pat)
218
pattern_lists[Globster.identify(pat)].append(pat)
219
pi = Globster.pattern_info
220
for t in Globster.pattern_types:
221
self._add_patterns(pattern_lists[t], pi[t]["translator"],
224
def _add_patterns(self, patterns, translator, prefix=''):
227
'(%s)' % translator(pat) for pat in patterns[:99]]
228
joined_rule = '%s(?:%s)$' % (prefix, '|'.join(grouped_rules))
229
# Explicitly use lazy_compile here, because we count on its
230
# nicer error reporting.
231
self._regex_patterns.append((
232
lazy_regex.lazy_compile(joined_rule, re.UNICODE),
234
patterns = patterns[99:]
236
def match(self, filename):
237
"""Searches for a pattern that matches the given filename.
239
:return A matching pattern or None if there is no matching pattern.
242
for regex, patterns in self._regex_patterns:
243
match = regex.match(filename)
245
return patterns[match.lastindex - 1]
246
except lazy_regex.InvalidPattern as e:
247
# We can't show the default e.msg to the user as thats for
248
# the combined pattern we sent to regex. Instead we indicate to
249
# the user that an ignore file needs fixing.
250
mutter('Invalid pattern found in regex: %s.', e.msg)
252
"File ~/.config/breezy/ignore or "
253
".bzrignore contains error(s).")
255
for _, patterns in self._regex_patterns:
257
if not Globster.is_pattern_valid(p):
258
bad_patterns += ('\n %s' % p)
259
e.msg += bad_patterns
264
def identify(pattern):
265
"""Returns pattern category.
267
:param pattern: normalized pattern.
268
Identify if a pattern is fullpath, basename or extension
269
and returns the appropriate type.
271
if pattern.startswith(u'RE:') or u'/' in pattern:
273
elif pattern.startswith(u'*.'):
279
def is_pattern_valid(pattern):
280
"""Returns True if pattern is valid.
282
:param pattern: Normalized pattern.
283
is_pattern_valid() assumes pattern to be normalized.
284
see: globbing.normalize_pattern
287
translator = Globster.pattern_info[Globster.identify(
288
pattern)]["translator"]
289
tpattern = '(%s)' % translator(pattern)
291
re_obj = lazy_regex.lazy_compile(tpattern, re.UNICODE)
292
re_obj.search("") # force compile
293
except lazy_regex.InvalidPattern:
298
class ExceptionGlobster(object):
299
"""A Globster that supports exception patterns.
301
Exceptions are ignore patterns prefixed with '!'. Exception
302
patterns take precedence over regular patterns and cause a
303
matching filename to return None from the match() function.
304
Patterns using a '!!' prefix are highest precedence, and act
305
as regular ignores. '!!' patterns are useful to establish ignores
306
that apply under paths specified by '!' exception patterns.
309
def __init__(self, patterns):
310
ignores = [[], [], []]
312
if p.startswith(u'!!'):
313
ignores[2].append(p[2:])
314
elif p.startswith(u'!'):
315
ignores[1].append(p[1:])
318
self._ignores = [Globster(i) for i in ignores]
320
def match(self, filename):
321
"""Searches for a pattern that matches the given filename.
323
:return A matching pattern or None if there is no matching pattern.
325
double_neg = self._ignores[2].match(filename)
327
return "!!%s" % double_neg
328
elif self._ignores[1].match(filename):
331
return self._ignores[0].match(filename)
334
class _OrderedGlobster(Globster):
335
"""A Globster that keeps pattern order."""
337
def __init__(self, patterns):
340
:param patterns: sequence of glob patterns
342
# Note: This could be smarter by running like sequences together
343
self._regex_patterns = []
345
pat = normalize_pattern(pat)
346
t = Globster.identify(pat)
347
self._add_patterns([pat], Globster.pattern_info[t]["translator"],
348
Globster.pattern_info[t]["prefix"])
351
_slashes = lazy_regex.lazy_compile(r'[\\/]+')
354
def normalize_pattern(pattern):
355
"""Converts backslashes in path patterns to forward slashes.
357
Doesn't normalize regular expressions - they may contain escapes.
359
if not (pattern.startswith('RE:') or pattern.startswith('!RE:')):
360
pattern = _slashes.sub('/', pattern)
362
pattern = pattern.rstrip('/')