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
34
class Replacer(object):
35
"""Do a multiple-pattern substitution.
37
The patterns and substitutions are combined into one, so the result of
38
one replacement is never substituted again. Add the patterns and
39
replacements via the add method and then call the object. The patterns
40
must not contain capturing groups.
43
_expand = lazy_regex.lazy_compile(u'\\\\&')
45
def __init__(self, source=None):
48
self._pats = list(source._pats)
49
self._funs = list(source._funs)
54
def add(self, pat, fun):
55
r"""Add a pattern and replacement.
57
The pattern must not contain capturing groups.
58
The replacement might be either a string template in which \& will be
59
replaced with the match, or a function that will get the matching text
60
as argument. It does not get match object, because capturing is
64
self._pats.append(pat)
65
self._funs.append(fun)
67
def add_replacer(self, replacer):
68
r"""Add all patterns from another replacer.
70
All patterns and replacements from replacer are appended to the ones
74
self._pats.extend(replacer._pats)
75
self._funs.extend(replacer._funs)
77
def __call__(self, text):
79
self._pat = lazy_regex.lazy_compile(
80
u'|'.join([u'(%s)' % p for p in self._pats]),
82
return self._pat.sub(self._do_sub, text)
85
fun = self._funs[m.lastindex - 1]
86
if hasattr(fun, '__call__'):
87
return fun(m.group(0))
89
return self._expand.sub(m.group(0), fun)
92
_sub_named = Replacer()
93
_sub_named.add(r'\[:digit:\]', r'\d')
94
_sub_named.add(r'\[:space:\]', r'\s')
95
_sub_named.add(r'\[:alnum:\]', r'\w')
96
_sub_named.add(r'\[:ascii:\]', r'\0-\x7f')
97
_sub_named.add(r'\[:blank:\]', r' \t')
98
_sub_named.add(r'\[:cntrl:\]', r'\0-\x1f\x7f-\x9f')
102
if m[1] in (u'!', u'^'):
103
return u'[^' + _sub_named(m[2:-1]) + u']'
104
return u'[' + _sub_named(m[1:-1]) + u']'
107
def _invalid_regex(repl):
109
warning(u"'%s' not allowed within a regular expression. "
110
"Replacing with '%s'" % (m, repl))
115
def _trailing_backslashes_regex(m):
116
"""Check trailing backslashes.
118
Does a head count on trailing backslashes to ensure there isn't an odd
119
one on the end that would escape the brackets we wrap the RE in.
121
if (len(m) % 2) != 0:
122
warning(u"Regular expressions cannot end with an odd number of '\\'. "
123
"Dropping the final '\\'.")
129
_sub_re.add(u'^RE:', u'')
130
_sub_re.add(u'\\((?!\\?)', u'(?:')
131
_sub_re.add(u'\\(\\?P<.*>', _invalid_regex(u'(?:'))
132
_sub_re.add(u'\\(\\?P=[^)]*\\)', _invalid_regex(u''))
133
_sub_re.add(r'\\+$', _trailing_backslashes_regex)
136
_sub_fullpath = Replacer()
137
_sub_fullpath.add(r'^RE:.*', _sub_re) # RE:<anything> is a regex
138
_sub_fullpath.add(r'\[\^?\]?(?:[^][]|\[:[^]]+:\])+\]',
139
_sub_group) # char group
140
_sub_fullpath.add(r'(?:(?<=/)|^)(?:\.?/)+', u'') # canonicalize path
141
_sub_fullpath.add(r'\\.', r'\&') # keep anything backslashed
142
_sub_fullpath.add(r'[(){}|^$+.]', r'\\&') # escape specials
143
_sub_fullpath.add(r'(?:(?<=/)|^)\*\*+/', r'(?:.*/)?') # **/ after ^ or /
144
_sub_fullpath.add(r'\*+', r'[^/]*') # * elsewhere
145
_sub_fullpath.add(r'\?', r'[^/]') # ? everywhere
148
_sub_basename = Replacer()
149
_sub_basename.add(r'\[\^?\]?(?:[^][]|\[:[^]]+:\])+\]',
150
_sub_group) # char group
151
_sub_basename.add(r'\\.', r'\&') # keep anything backslashed
152
_sub_basename.add(r'[(){}|^$+.]', r'\\&') # escape specials
153
_sub_basename.add(r'\*+', r'.*') # * everywhere
154
_sub_basename.add(r'\?', r'.') # ? everywhere
157
def _sub_extension(pattern):
158
return _sub_basename(pattern[2:])
161
class Globster(object):
162
"""A simple wrapper for a set of glob patterns.
164
Provides the capability to search the patterns to find a match for
165
a given filename (including the full path).
167
Patterns are translated to regular expressions to expidite matching.
169
The regular expressions for multiple patterns are aggregated into
170
a super-regex containing groups of up to 99 patterns.
171
The 99 limitation is due to the grouping limit of the Python re module.
172
The resulting super-regex and associated patterns are stored as a list of
173
(regex,[patterns]) in _regex_patterns.
175
For performance reasons the patterns are categorised as extension patterns
176
(those that match against a file extension), basename patterns
177
(those that match against the basename of the filename),
178
and fullpath patterns (those that match against the full path).
179
The translations used for extensions and basenames are relatively simpler
180
and therefore faster to perform than the fullpath patterns.
182
Also, the extension patterns are more likely to find a match and
183
so are matched first, then the basename patterns, then the fullpath
186
# We want to _add_patterns in a specific order (as per type_list below)
187
# starting with the shortest and going to the longest.
188
# As some Python version don't support ordered dicts the list below is
189
# used to select inputs for _add_pattern in a specific order.
190
pattern_types = ["extension", "basename", "fullpath"]
194
"translator": _sub_extension,
195
"prefix": r'(?:.*/)?(?!.*/)(?:.*\.)'
198
"translator": _sub_basename,
199
"prefix": r'(?:.*/)?(?!.*/)'
202
"translator": _sub_fullpath,
207
def __init__(self, patterns):
208
self._regex_patterns = []
215
pat = normalize_pattern(pat)
216
pattern_lists[Globster.identify(pat)].append(pat)
217
pi = Globster.pattern_info
218
for t in Globster.pattern_types:
219
self._add_patterns(pattern_lists[t], pi[t]["translator"],
222
def _add_patterns(self, patterns, translator, prefix=''):
225
'(%s)' % translator(pat) for pat in patterns[:99]]
226
joined_rule = '%s(?:%s)$' % (prefix, '|'.join(grouped_rules))
227
# Explicitly use lazy_compile here, because we count on its
228
# nicer error reporting.
229
self._regex_patterns.append((
230
lazy_regex.lazy_compile(joined_rule, re.UNICODE),
232
patterns = patterns[99:]
234
def match(self, filename):
235
"""Searches for a pattern that matches the given filename.
237
:return A matching pattern or None if there is no matching pattern.
240
for regex, patterns in self._regex_patterns:
241
match = regex.match(filename)
243
return patterns[match.lastindex - 1]
244
except lazy_regex.InvalidPattern as e:
245
# We can't show the default e.msg to the user as thats for
246
# the combined pattern we sent to regex. Instead we indicate to
247
# the user that an ignore file needs fixing.
248
mutter('Invalid pattern found in regex: %s.', e.msg)
250
"File ~/.config/breezy/ignore or "
251
".bzrignore contains error(s).")
253
for _, patterns in self._regex_patterns:
255
if not Globster.is_pattern_valid(p):
256
bad_patterns += ('\n %s' % p)
257
e.msg += bad_patterns
262
def identify(pattern):
263
"""Returns pattern category.
265
:param pattern: normalized pattern.
266
Identify if a pattern is fullpath, basename or extension
267
and returns the appropriate type.
269
if pattern.startswith(u'RE:') or u'/' in pattern:
271
elif pattern.startswith(u'*.'):
277
def is_pattern_valid(pattern):
278
"""Returns True if pattern is valid.
280
:param pattern: Normalized pattern.
281
is_pattern_valid() assumes pattern to be normalized.
282
see: globbing.normalize_pattern
285
translator = Globster.pattern_info[Globster.identify(
286
pattern)]["translator"]
287
tpattern = '(%s)' % translator(pattern)
289
re_obj = lazy_regex.lazy_compile(tpattern, re.UNICODE)
290
re_obj.search("") # force compile
291
except lazy_regex.InvalidPattern:
296
class ExceptionGlobster(object):
297
"""A Globster that supports exception patterns.
299
Exceptions are ignore patterns prefixed with '!'. Exception
300
patterns take precedence over regular patterns and cause a
301
matching filename to return None from the match() function.
302
Patterns using a '!!' prefix are highest precedence, and act
303
as regular ignores. '!!' patterns are useful to establish ignores
304
that apply under paths specified by '!' exception patterns.
307
def __init__(self, patterns):
308
ignores = [[], [], []]
310
if p.startswith(u'!!'):
311
ignores[2].append(p[2:])
312
elif p.startswith(u'!'):
313
ignores[1].append(p[1:])
316
self._ignores = [Globster(i) for i in ignores]
318
def match(self, filename):
319
"""Searches for a pattern that matches the given filename.
321
:return A matching pattern or None if there is no matching pattern.
323
double_neg = self._ignores[2].match(filename)
325
return "!!%s" % double_neg
326
elif self._ignores[1].match(filename):
329
return self._ignores[0].match(filename)
332
class _OrderedGlobster(Globster):
333
"""A Globster that keeps pattern order."""
335
def __init__(self, patterns):
338
:param patterns: sequence of glob patterns
340
# Note: This could be smarter by running like sequences together
341
self._regex_patterns = []
343
pat = normalize_pattern(pat)
344
t = Globster.identify(pat)
345
self._add_patterns([pat], Globster.pattern_info[t]["translator"],
346
Globster.pattern_info[t]["prefix"])
349
_slashes = lazy_regex.lazy_compile(r'[\\/]+')
352
def normalize_pattern(pattern):
353
"""Converts backslashes in path patterns to forward slashes.
355
Doesn't normalize regular expressions - they may contain escapes.
357
if not (pattern.startswith('RE:') or pattern.startswith('!RE:')):
358
pattern = _slashes.sub('/', pattern)
360
pattern = pattern.rstrip('/')