Show More
@@ -1,808 +1,805 b'' | |||||
1 | # match.py - filename matching |
|
1 | # match.py - filename matching | |
2 | # |
|
2 | # | |
3 | # Copyright 2008, 2009 Matt Mackall <mpm@selenic.com> and others |
|
3 | # Copyright 2008, 2009 Matt Mackall <mpm@selenic.com> and others | |
4 | # |
|
4 | # | |
5 | # This software may be used and distributed according to the terms of the |
|
5 | # This software may be used and distributed according to the terms of the | |
6 | # GNU General Public License version 2 or any later version. |
|
6 | # GNU General Public License version 2 or any later version. | |
7 |
|
7 | |||
8 | from __future__ import absolute_import |
|
8 | from __future__ import absolute_import | |
9 |
|
9 | |||
10 | import copy |
|
10 | import copy | |
11 | import os |
|
11 | import os | |
12 | import re |
|
12 | import re | |
13 |
|
13 | |||
14 | from .i18n import _ |
|
14 | from .i18n import _ | |
15 | from . import ( |
|
15 | from . import ( | |
16 | error, |
|
16 | error, | |
17 | pathutil, |
|
17 | pathutil, | |
18 | util, |
|
18 | util, | |
19 | ) |
|
19 | ) | |
20 |
|
20 | |||
21 | propertycache = util.propertycache |
|
21 | propertycache = util.propertycache | |
22 |
|
22 | |||
23 | def _rematcher(regex): |
|
23 | def _rematcher(regex): | |
24 | '''compile the regexp with the best available regexp engine and return a |
|
24 | '''compile the regexp with the best available regexp engine and return a | |
25 | matcher function''' |
|
25 | matcher function''' | |
26 | m = util.re.compile(regex) |
|
26 | m = util.re.compile(regex) | |
27 | try: |
|
27 | try: | |
28 | # slightly faster, provided by facebook's re2 bindings |
|
28 | # slightly faster, provided by facebook's re2 bindings | |
29 | return m.test_match |
|
29 | return m.test_match | |
30 | except AttributeError: |
|
30 | except AttributeError: | |
31 | return m.match |
|
31 | return m.match | |
32 |
|
32 | |||
33 | def _expandsets(kindpats, ctx, listsubrepos): |
|
33 | def _expandsets(kindpats, ctx, listsubrepos): | |
34 | '''Returns the kindpats list with the 'set' patterns expanded.''' |
|
34 | '''Returns the kindpats list with the 'set' patterns expanded.''' | |
35 | fset = set() |
|
35 | fset = set() | |
36 | other = [] |
|
36 | other = [] | |
37 |
|
37 | |||
38 | for kind, pat, source in kindpats: |
|
38 | for kind, pat, source in kindpats: | |
39 | if kind == 'set': |
|
39 | if kind == 'set': | |
40 | if not ctx: |
|
40 | if not ctx: | |
41 | raise error.Abort(_("fileset expression with no context")) |
|
41 | raise error.Abort(_("fileset expression with no context")) | |
42 | s = ctx.getfileset(pat) |
|
42 | s = ctx.getfileset(pat) | |
43 | fset.update(s) |
|
43 | fset.update(s) | |
44 |
|
44 | |||
45 | if listsubrepos: |
|
45 | if listsubrepos: | |
46 | for subpath in ctx.substate: |
|
46 | for subpath in ctx.substate: | |
47 | s = ctx.sub(subpath).getfileset(pat) |
|
47 | s = ctx.sub(subpath).getfileset(pat) | |
48 | fset.update(subpath + '/' + f for f in s) |
|
48 | fset.update(subpath + '/' + f for f in s) | |
49 |
|
49 | |||
50 | continue |
|
50 | continue | |
51 | other.append((kind, pat, source)) |
|
51 | other.append((kind, pat, source)) | |
52 | return fset, other |
|
52 | return fset, other | |
53 |
|
53 | |||
54 | def _expandsubinclude(kindpats, root): |
|
54 | def _expandsubinclude(kindpats, root): | |
55 | '''Returns the list of subinclude matcher args and the kindpats without the |
|
55 | '''Returns the list of subinclude matcher args and the kindpats without the | |
56 | subincludes in it.''' |
|
56 | subincludes in it.''' | |
57 | relmatchers = [] |
|
57 | relmatchers = [] | |
58 | other = [] |
|
58 | other = [] | |
59 |
|
59 | |||
60 | for kind, pat, source in kindpats: |
|
60 | for kind, pat, source in kindpats: | |
61 | if kind == 'subinclude': |
|
61 | if kind == 'subinclude': | |
62 | sourceroot = pathutil.dirname(util.normpath(source)) |
|
62 | sourceroot = pathutil.dirname(util.normpath(source)) | |
63 | pat = util.pconvert(pat) |
|
63 | pat = util.pconvert(pat) | |
64 | path = pathutil.join(sourceroot, pat) |
|
64 | path = pathutil.join(sourceroot, pat) | |
65 |
|
65 | |||
66 | newroot = pathutil.dirname(path) |
|
66 | newroot = pathutil.dirname(path) | |
67 | matcherargs = (newroot, '', [], ['include:%s' % path]) |
|
67 | matcherargs = (newroot, '', [], ['include:%s' % path]) | |
68 |
|
68 | |||
69 | prefix = pathutil.canonpath(root, root, newroot) |
|
69 | prefix = pathutil.canonpath(root, root, newroot) | |
70 | if prefix: |
|
70 | if prefix: | |
71 | prefix += '/' |
|
71 | prefix += '/' | |
72 | relmatchers.append((prefix, matcherargs)) |
|
72 | relmatchers.append((prefix, matcherargs)) | |
73 | else: |
|
73 | else: | |
74 | other.append((kind, pat, source)) |
|
74 | other.append((kind, pat, source)) | |
75 |
|
75 | |||
76 | return relmatchers, other |
|
76 | return relmatchers, other | |
77 |
|
77 | |||
78 | def _kindpatsalwaysmatch(kindpats): |
|
78 | def _kindpatsalwaysmatch(kindpats): | |
79 | """"Checks whether the kindspats match everything, as e.g. |
|
79 | """"Checks whether the kindspats match everything, as e.g. | |
80 | 'relpath:.' does. |
|
80 | 'relpath:.' does. | |
81 | """ |
|
81 | """ | |
82 | for kind, pat, source in kindpats: |
|
82 | for kind, pat, source in kindpats: | |
83 | if pat != '' or kind not in ['relpath', 'glob']: |
|
83 | if pat != '' or kind not in ['relpath', 'glob']: | |
84 | return False |
|
84 | return False | |
85 | return True |
|
85 | return True | |
86 |
|
86 | |||
87 | def match(root, cwd, patterns, include=None, exclude=None, default='glob', |
|
87 | def match(root, cwd, patterns, include=None, exclude=None, default='glob', | |
88 | exact=False, auditor=None, ctx=None, listsubrepos=False, warn=None, |
|
88 | exact=False, auditor=None, ctx=None, listsubrepos=False, warn=None, | |
89 | badfn=None): |
|
89 | badfn=None): | |
90 | """build an object to match a set of file patterns |
|
90 | """build an object to match a set of file patterns | |
91 |
|
91 | |||
92 | arguments: |
|
92 | arguments: | |
93 | root - the canonical root of the tree you're matching against |
|
93 | root - the canonical root of the tree you're matching against | |
94 | cwd - the current working directory, if relevant |
|
94 | cwd - the current working directory, if relevant | |
95 | patterns - patterns to find |
|
95 | patterns - patterns to find | |
96 | include - patterns to include (unless they are excluded) |
|
96 | include - patterns to include (unless they are excluded) | |
97 | exclude - patterns to exclude (even if they are included) |
|
97 | exclude - patterns to exclude (even if they are included) | |
98 | default - if a pattern in patterns has no explicit type, assume this one |
|
98 | default - if a pattern in patterns has no explicit type, assume this one | |
99 | exact - patterns are actually filenames (include/exclude still apply) |
|
99 | exact - patterns are actually filenames (include/exclude still apply) | |
100 | warn - optional function used for printing warnings |
|
100 | warn - optional function used for printing warnings | |
101 | badfn - optional bad() callback for this matcher instead of the default |
|
101 | badfn - optional bad() callback for this matcher instead of the default | |
102 |
|
102 | |||
103 | a pattern is one of: |
|
103 | a pattern is one of: | |
104 | 'glob:<glob>' - a glob relative to cwd |
|
104 | 'glob:<glob>' - a glob relative to cwd | |
105 | 're:<regexp>' - a regular expression |
|
105 | 're:<regexp>' - a regular expression | |
106 | 'path:<path>' - a path relative to repository root, which is matched |
|
106 | 'path:<path>' - a path relative to repository root, which is matched | |
107 | recursively |
|
107 | recursively | |
108 | 'rootfilesin:<path>' - a path relative to repository root, which is |
|
108 | 'rootfilesin:<path>' - a path relative to repository root, which is | |
109 | matched non-recursively (will not match subdirectories) |
|
109 | matched non-recursively (will not match subdirectories) | |
110 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
110 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
111 | 'relpath:<path>' - a path relative to cwd |
|
111 | 'relpath:<path>' - a path relative to cwd | |
112 | 'relre:<regexp>' - a regexp that needn't match the start of a name |
|
112 | 'relre:<regexp>' - a regexp that needn't match the start of a name | |
113 | 'set:<fileset>' - a fileset expression |
|
113 | 'set:<fileset>' - a fileset expression | |
114 | 'include:<path>' - a file of patterns to read and include |
|
114 | 'include:<path>' - a file of patterns to read and include | |
115 | 'subinclude:<path>' - a file of patterns to match against files under |
|
115 | 'subinclude:<path>' - a file of patterns to match against files under | |
116 | the same directory |
|
116 | the same directory | |
117 | '<something>' - a pattern of the specified default type |
|
117 | '<something>' - a pattern of the specified default type | |
118 | """ |
|
118 | """ | |
119 |
return matcher(root, cwd, patterns, include=include, |
|
119 | return matcher(root, cwd, _donormalize, patterns, include=include, | |
120 |
default=default, exact=exact, |
|
120 | exclude=exclude, default=default, exact=exact, | |
121 |
listsubrepos=listsubrepos, |
|
121 | auditor=auditor, ctx=ctx, listsubrepos=listsubrepos, | |
|
122 | warn=warn, badfn=badfn) | |||
122 |
|
123 | |||
123 | def icasefsmatch(root, cwd, patterns, include=None, exclude=None, |
|
124 | def icasefsmatch(root, cwd, patterns, include=None, exclude=None, | |
124 | default='glob', auditor=None, ctx=None, |
|
125 | default='glob', auditor=None, ctx=None, | |
125 | listsubrepos=False, badfn=None): |
|
126 | listsubrepos=False, badfn=None): | |
126 | """A matcher for wdir on case insensitive filesystems, which normalizes the |
|
127 | """A matcher for wdir on case insensitive filesystems, which normalizes the | |
127 | given patterns to the case in the filesystem. |
|
128 | given patterns to the case in the filesystem. | |
128 | """ |
|
129 | """ | |
129 | return icasefsmatcher(root, cwd, patterns, include=include, exclude=exclude, |
|
130 | dirstate = ctx.repo().dirstate | |
130 | default=default, auditor=auditor, ctx=ctx, |
|
131 | dsnormalize = dirstate.normalize | |
131 | listsubrepos=listsubrepos, badfn=badfn) |
|
132 | ||
|
133 | def normalize(patterns, default, root, cwd, auditor, warn): | |||
|
134 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) | |||
|
135 | kindpats = [] | |||
|
136 | for kind, pats, source in kp: | |||
|
137 | if kind not in ('re', 'relre'): # regex can't be normalized | |||
|
138 | p = pats | |||
|
139 | pats = dsnormalize(pats) | |||
|
140 | ||||
|
141 | # Preserve the original to handle a case only rename. | |||
|
142 | if p != pats and p in dirstate: | |||
|
143 | kindpats.append((kind, p, source)) | |||
|
144 | ||||
|
145 | kindpats.append((kind, pats, source)) | |||
|
146 | return kindpats | |||
|
147 | ||||
|
148 | return icasefsmatcher(root, cwd, normalize, patterns=patterns, | |||
|
149 | include=include, exclude=exclude, default=default, | |||
|
150 | auditor=auditor, ctx=ctx, listsubrepos=listsubrepos, | |||
|
151 | badfn=badfn) | |||
132 |
|
152 | |||
133 | def exact(root, cwd, files, badfn=None): |
|
153 | def exact(root, cwd, files, badfn=None): | |
134 | return match(root, cwd, files, exact=True, badfn=badfn) |
|
154 | return match(root, cwd, files, exact=True, badfn=badfn) | |
135 |
|
155 | |||
136 | def always(root, cwd): |
|
156 | def always(root, cwd): | |
137 | return match(root, cwd, []) |
|
157 | return match(root, cwd, []) | |
138 |
|
158 | |||
139 | def badmatch(match, badfn): |
|
159 | def badmatch(match, badfn): | |
140 | """Make a copy of the given matcher, replacing its bad method with the given |
|
160 | """Make a copy of the given matcher, replacing its bad method with the given | |
141 | one. |
|
161 | one. | |
142 | """ |
|
162 | """ | |
143 | m = copy.copy(match) |
|
163 | m = copy.copy(match) | |
144 | m.bad = badfn |
|
164 | m.bad = badfn | |
145 | return m |
|
165 | return m | |
146 |
|
166 | |||
147 | def _donormalize(patterns, default, root, cwd, auditor, warn): |
|
167 | def _donormalize(patterns, default, root, cwd, auditor, warn): | |
148 | '''Convert 'kind:pat' from the patterns list to tuples with kind and |
|
168 | '''Convert 'kind:pat' from the patterns list to tuples with kind and | |
149 | normalized and rooted patterns and with listfiles expanded.''' |
|
169 | normalized and rooted patterns and with listfiles expanded.''' | |
150 | kindpats = [] |
|
170 | kindpats = [] | |
151 | for kind, pat in [_patsplit(p, default) for p in patterns]: |
|
171 | for kind, pat in [_patsplit(p, default) for p in patterns]: | |
152 | if kind in ('glob', 'relpath'): |
|
172 | if kind in ('glob', 'relpath'): | |
153 | pat = pathutil.canonpath(root, cwd, pat, auditor) |
|
173 | pat = pathutil.canonpath(root, cwd, pat, auditor) | |
154 | elif kind in ('relglob', 'path', 'rootfilesin'): |
|
174 | elif kind in ('relglob', 'path', 'rootfilesin'): | |
155 | pat = util.normpath(pat) |
|
175 | pat = util.normpath(pat) | |
156 | elif kind in ('listfile', 'listfile0'): |
|
176 | elif kind in ('listfile', 'listfile0'): | |
157 | try: |
|
177 | try: | |
158 | files = util.readfile(pat) |
|
178 | files = util.readfile(pat) | |
159 | if kind == 'listfile0': |
|
179 | if kind == 'listfile0': | |
160 | files = files.split('\0') |
|
180 | files = files.split('\0') | |
161 | else: |
|
181 | else: | |
162 | files = files.splitlines() |
|
182 | files = files.splitlines() | |
163 | files = [f for f in files if f] |
|
183 | files = [f for f in files if f] | |
164 | except EnvironmentError: |
|
184 | except EnvironmentError: | |
165 | raise error.Abort(_("unable to read file list (%s)") % pat) |
|
185 | raise error.Abort(_("unable to read file list (%s)") % pat) | |
166 | for k, p, source in _donormalize(files, default, root, cwd, |
|
186 | for k, p, source in _donormalize(files, default, root, cwd, | |
167 | auditor, warn): |
|
187 | auditor, warn): | |
168 | kindpats.append((k, p, pat)) |
|
188 | kindpats.append((k, p, pat)) | |
169 | continue |
|
189 | continue | |
170 | elif kind == 'include': |
|
190 | elif kind == 'include': | |
171 | try: |
|
191 | try: | |
172 | fullpath = os.path.join(root, util.localpath(pat)) |
|
192 | fullpath = os.path.join(root, util.localpath(pat)) | |
173 | includepats = readpatternfile(fullpath, warn) |
|
193 | includepats = readpatternfile(fullpath, warn) | |
174 | for k, p, source in _donormalize(includepats, default, |
|
194 | for k, p, source in _donormalize(includepats, default, | |
175 | root, cwd, auditor, warn): |
|
195 | root, cwd, auditor, warn): | |
176 | kindpats.append((k, p, source or pat)) |
|
196 | kindpats.append((k, p, source or pat)) | |
177 | except error.Abort as inst: |
|
197 | except error.Abort as inst: | |
178 | raise error.Abort('%s: %s' % (pat, inst[0])) |
|
198 | raise error.Abort('%s: %s' % (pat, inst[0])) | |
179 | except IOError as inst: |
|
199 | except IOError as inst: | |
180 | if warn: |
|
200 | if warn: | |
181 | warn(_("skipping unreadable pattern file '%s': %s\n") % |
|
201 | warn(_("skipping unreadable pattern file '%s': %s\n") % | |
182 | (pat, inst.strerror)) |
|
202 | (pat, inst.strerror)) | |
183 | continue |
|
203 | continue | |
184 | # else: re or relre - which cannot be normalized |
|
204 | # else: re or relre - which cannot be normalized | |
185 | kindpats.append((kind, pat, '')) |
|
205 | kindpats.append((kind, pat, '')) | |
186 | return kindpats |
|
206 | return kindpats | |
187 |
|
207 | |||
188 | class matcher(object): |
|
208 | class matcher(object): | |
189 |
|
209 | |||
190 |
def __init__(self, root, cwd, patterns, include=None, |
|
210 | def __init__(self, root, cwd, normalize, patterns, include=None, | |
191 |
default='glob', exact=False, auditor=None, |
|
211 | exclude=None, default='glob', exact=False, auditor=None, | |
192 | listsubrepos=False, warn=None, badfn=None): |
|
212 | ctx=None, listsubrepos=False, warn=None, badfn=None): | |
193 | if include is None: |
|
213 | if include is None: | |
194 | include = [] |
|
214 | include = [] | |
195 | if exclude is None: |
|
215 | if exclude is None: | |
196 | exclude = [] |
|
216 | exclude = [] | |
197 |
|
217 | |||
198 | self._root = root |
|
218 | self._root = root | |
199 | self._cwd = cwd |
|
219 | self._cwd = cwd | |
200 | self._files = [] # exact files and roots of patterns |
|
220 | self._files = [] # exact files and roots of patterns | |
201 | self._anypats = bool(include or exclude) |
|
221 | self._anypats = bool(include or exclude) | |
202 | self._always = False |
|
222 | self._always = False | |
203 | self._pathrestricted = bool(include or exclude or patterns) |
|
223 | self._pathrestricted = bool(include or exclude or patterns) | |
204 |
|
224 | |||
205 | # roots are directories which are recursively included/excluded. |
|
225 | # roots are directories which are recursively included/excluded. | |
206 | self._includeroots = set() |
|
226 | self._includeroots = set() | |
207 | self._excluderoots = set() |
|
227 | self._excluderoots = set() | |
208 | # dirs are directories which are non-recursively included. |
|
228 | # dirs are directories which are non-recursively included. | |
209 | self._includedirs = set() |
|
229 | self._includedirs = set() | |
210 |
|
230 | |||
211 | if badfn is not None: |
|
231 | if badfn is not None: | |
212 | self.bad = badfn |
|
232 | self.bad = badfn | |
213 |
|
233 | |||
214 | matchfns = [] |
|
234 | matchfns = [] | |
215 | if include: |
|
235 | if include: | |
216 |
kindpats = |
|
236 | kindpats = normalize(include, 'glob', root, cwd, auditor, warn) | |
217 | warn) |
|
|||
218 | self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)', |
|
237 | self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)', | |
219 | listsubrepos, root) |
|
238 | listsubrepos, root) | |
220 | roots, dirs = _rootsanddirs(kindpats) |
|
239 | roots, dirs = _rootsanddirs(kindpats) | |
221 | self._includeroots.update(roots) |
|
240 | self._includeroots.update(roots) | |
222 | self._includedirs.update(dirs) |
|
241 | self._includedirs.update(dirs) | |
223 | matchfns.append(im) |
|
242 | matchfns.append(im) | |
224 | if exclude: |
|
243 | if exclude: | |
225 |
kindpats = |
|
244 | kindpats = normalize(exclude, 'glob', root, cwd, auditor, warn) | |
226 | warn) |
|
|||
227 | self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)', |
|
245 | self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)', | |
228 | listsubrepos, root) |
|
246 | listsubrepos, root) | |
229 | if not _anypats(kindpats): |
|
247 | if not _anypats(kindpats): | |
230 | # Only consider recursive excludes as such - if a non-recursive |
|
248 | # Only consider recursive excludes as such - if a non-recursive | |
231 | # exclude is used, we must still recurse into the excluded |
|
249 | # exclude is used, we must still recurse into the excluded | |
232 | # directory, at least to find subdirectories. In such a case, |
|
250 | # directory, at least to find subdirectories. In such a case, | |
233 | # the regex still won't match the non-recursively-excluded |
|
251 | # the regex still won't match the non-recursively-excluded | |
234 | # files. |
|
252 | # files. | |
235 | self._excluderoots.update(_roots(kindpats)) |
|
253 | self._excluderoots.update(_roots(kindpats)) | |
236 | matchfns.append(lambda f: not em(f)) |
|
254 | matchfns.append(lambda f: not em(f)) | |
237 | if exact: |
|
255 | if exact: | |
238 | if isinstance(patterns, list): |
|
256 | if isinstance(patterns, list): | |
239 | self._files = patterns |
|
257 | self._files = patterns | |
240 | else: |
|
258 | else: | |
241 | self._files = list(patterns) |
|
259 | self._files = list(patterns) | |
242 | matchfns.append(self.exact) |
|
260 | matchfns.append(self.exact) | |
243 | elif patterns: |
|
261 | elif patterns: | |
244 |
kindpats = |
|
262 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) | |
245 | warn) |
|
|||
246 | if not _kindpatsalwaysmatch(kindpats): |
|
263 | if not _kindpatsalwaysmatch(kindpats): | |
247 | self._files = _explicitfiles(kindpats) |
|
264 | self._files = _explicitfiles(kindpats) | |
248 | self._anypats = self._anypats or _anypats(kindpats) |
|
265 | self._anypats = self._anypats or _anypats(kindpats) | |
249 | self.patternspat, pm = _buildmatch(ctx, kindpats, '$', |
|
266 | self.patternspat, pm = _buildmatch(ctx, kindpats, '$', | |
250 | listsubrepos, root) |
|
267 | listsubrepos, root) | |
251 | matchfns.append(pm) |
|
268 | matchfns.append(pm) | |
252 |
|
269 | |||
253 | if not matchfns: |
|
270 | if not matchfns: | |
254 | m = util.always |
|
271 | m = util.always | |
255 | self._always = True |
|
272 | self._always = True | |
256 | elif len(matchfns) == 1: |
|
273 | elif len(matchfns) == 1: | |
257 | m = matchfns[0] |
|
274 | m = matchfns[0] | |
258 | else: |
|
275 | else: | |
259 | def m(f): |
|
276 | def m(f): | |
260 | for matchfn in matchfns: |
|
277 | for matchfn in matchfns: | |
261 | if not matchfn(f): |
|
278 | if not matchfn(f): | |
262 | return False |
|
279 | return False | |
263 | return True |
|
280 | return True | |
264 |
|
281 | |||
265 | self.matchfn = m |
|
282 | self.matchfn = m | |
266 |
|
283 | |||
267 | def __call__(self, fn): |
|
284 | def __call__(self, fn): | |
268 | return self.matchfn(fn) |
|
285 | return self.matchfn(fn) | |
269 | def __iter__(self): |
|
286 | def __iter__(self): | |
270 | for f in self._files: |
|
287 | for f in self._files: | |
271 | yield f |
|
288 | yield f | |
272 |
|
289 | |||
273 | # Callbacks related to how the matcher is used by dirstate.walk. |
|
290 | # Callbacks related to how the matcher is used by dirstate.walk. | |
274 | # Subscribers to these events must monkeypatch the matcher object. |
|
291 | # Subscribers to these events must monkeypatch the matcher object. | |
275 | def bad(self, f, msg): |
|
292 | def bad(self, f, msg): | |
276 | '''Callback from dirstate.walk for each explicit file that can't be |
|
293 | '''Callback from dirstate.walk for each explicit file that can't be | |
277 | found/accessed, with an error message.''' |
|
294 | found/accessed, with an error message.''' | |
278 | pass |
|
295 | pass | |
279 |
|
296 | |||
280 | # If an explicitdir is set, it will be called when an explicitly listed |
|
297 | # If an explicitdir is set, it will be called when an explicitly listed | |
281 | # directory is visited. |
|
298 | # directory is visited. | |
282 | explicitdir = None |
|
299 | explicitdir = None | |
283 |
|
300 | |||
284 | # If an traversedir is set, it will be called when a directory discovered |
|
301 | # If an traversedir is set, it will be called when a directory discovered | |
285 | # by recursive traversal is visited. |
|
302 | # by recursive traversal is visited. | |
286 | traversedir = None |
|
303 | traversedir = None | |
287 |
|
304 | |||
288 | def abs(self, f): |
|
305 | def abs(self, f): | |
289 | '''Convert a repo path back to path that is relative to the root of the |
|
306 | '''Convert a repo path back to path that is relative to the root of the | |
290 | matcher.''' |
|
307 | matcher.''' | |
291 | return f |
|
308 | return f | |
292 |
|
309 | |||
293 | def rel(self, f): |
|
310 | def rel(self, f): | |
294 | '''Convert repo path back to path that is relative to cwd of matcher.''' |
|
311 | '''Convert repo path back to path that is relative to cwd of matcher.''' | |
295 | return util.pathto(self._root, self._cwd, f) |
|
312 | return util.pathto(self._root, self._cwd, f) | |
296 |
|
313 | |||
297 | def uipath(self, f): |
|
314 | def uipath(self, f): | |
298 | '''Convert repo path to a display path. If patterns or -I/-X were used |
|
315 | '''Convert repo path to a display path. If patterns or -I/-X were used | |
299 | to create this matcher, the display path will be relative to cwd. |
|
316 | to create this matcher, the display path will be relative to cwd. | |
300 | Otherwise it is relative to the root of the repo.''' |
|
317 | Otherwise it is relative to the root of the repo.''' | |
301 | return (self._pathrestricted and self.rel(f)) or self.abs(f) |
|
318 | return (self._pathrestricted and self.rel(f)) or self.abs(f) | |
302 |
|
319 | |||
303 | def files(self): |
|
320 | def files(self): | |
304 | '''Explicitly listed files or patterns or roots: |
|
321 | '''Explicitly listed files or patterns or roots: | |
305 | if no patterns or .always(): empty list, |
|
322 | if no patterns or .always(): empty list, | |
306 | if exact: list exact files, |
|
323 | if exact: list exact files, | |
307 | if not .anypats(): list all files and dirs, |
|
324 | if not .anypats(): list all files and dirs, | |
308 | else: optimal roots''' |
|
325 | else: optimal roots''' | |
309 | return self._files |
|
326 | return self._files | |
310 |
|
327 | |||
311 | @propertycache |
|
328 | @propertycache | |
312 | def _fileset(self): |
|
329 | def _fileset(self): | |
313 | return set(self._files) |
|
330 | return set(self._files) | |
314 |
|
331 | |||
315 | @propertycache |
|
332 | @propertycache | |
316 | def _dirs(self): |
|
333 | def _dirs(self): | |
317 | return set(util.dirs(self._fileset)) | {'.'} |
|
334 | return set(util.dirs(self._fileset)) | {'.'} | |
318 |
|
335 | |||
319 | def visitdir(self, dir): |
|
336 | def visitdir(self, dir): | |
320 | '''Decides whether a directory should be visited based on whether it |
|
337 | '''Decides whether a directory should be visited based on whether it | |
321 | has potential matches in it or one of its subdirectories. This is |
|
338 | has potential matches in it or one of its subdirectories. This is | |
322 | based on the match's primary, included, and excluded patterns. |
|
339 | based on the match's primary, included, and excluded patterns. | |
323 |
|
340 | |||
324 | Returns the string 'all' if the given directory and all subdirectories |
|
341 | Returns the string 'all' if the given directory and all subdirectories | |
325 | should be visited. Otherwise returns True or False indicating whether |
|
342 | should be visited. Otherwise returns True or False indicating whether | |
326 | the given directory should be visited. |
|
343 | the given directory should be visited. | |
327 |
|
344 | |||
328 | This function's behavior is undefined if it has returned False for |
|
345 | This function's behavior is undefined if it has returned False for | |
329 | one of the dir's parent directories. |
|
346 | one of the dir's parent directories. | |
330 | ''' |
|
347 | ''' | |
331 | if self.prefix() and dir in self._fileset: |
|
348 | if self.prefix() and dir in self._fileset: | |
332 | return 'all' |
|
349 | return 'all' | |
333 | if dir in self._excluderoots: |
|
350 | if dir in self._excluderoots: | |
334 | return False |
|
351 | return False | |
335 | if ((self._includeroots or self._includedirs) and |
|
352 | if ((self._includeroots or self._includedirs) and | |
336 | '.' not in self._includeroots and |
|
353 | '.' not in self._includeroots and | |
337 | dir not in self._includeroots and |
|
354 | dir not in self._includeroots and | |
338 | dir not in self._includedirs and |
|
355 | dir not in self._includedirs and | |
339 | not any(parent in self._includeroots |
|
356 | not any(parent in self._includeroots | |
340 | for parent in util.finddirs(dir))): |
|
357 | for parent in util.finddirs(dir))): | |
341 | return False |
|
358 | return False | |
342 | return (not self._fileset or |
|
359 | return (not self._fileset or | |
343 | '.' in self._fileset or |
|
360 | '.' in self._fileset or | |
344 | dir in self._fileset or |
|
361 | dir in self._fileset or | |
345 | dir in self._dirs or |
|
362 | dir in self._dirs or | |
346 | any(parentdir in self._fileset |
|
363 | any(parentdir in self._fileset | |
347 | for parentdir in util.finddirs(dir))) |
|
364 | for parentdir in util.finddirs(dir))) | |
348 |
|
365 | |||
349 | def exact(self, f): |
|
366 | def exact(self, f): | |
350 | '''Returns True if f is in .files().''' |
|
367 | '''Returns True if f is in .files().''' | |
351 | return f in self._fileset |
|
368 | return f in self._fileset | |
352 |
|
369 | |||
353 | def anypats(self): |
|
370 | def anypats(self): | |
354 | '''Matcher uses patterns or include/exclude.''' |
|
371 | '''Matcher uses patterns or include/exclude.''' | |
355 | return self._anypats |
|
372 | return self._anypats | |
356 |
|
373 | |||
357 | def always(self): |
|
374 | def always(self): | |
358 | '''Matcher will match everything and .files() will be empty |
|
375 | '''Matcher will match everything and .files() will be empty | |
359 | - optimization might be possible and necessary.''' |
|
376 | - optimization might be possible and necessary.''' | |
360 | return self._always |
|
377 | return self._always | |
361 |
|
378 | |||
362 | def isexact(self): |
|
379 | def isexact(self): | |
363 | return self.matchfn == self.exact |
|
380 | return self.matchfn == self.exact | |
364 |
|
381 | |||
365 | def prefix(self): |
|
382 | def prefix(self): | |
366 | return not self.always() and not self.isexact() and not self.anypats() |
|
383 | return not self.always() and not self.isexact() and not self.anypats() | |
367 |
|
384 | |||
368 | def _normalize(self, patterns, default, root, cwd, auditor, warn): |
|
|||
369 | return _donormalize(patterns, default, root, cwd, auditor, warn) |
|
|||
370 |
|
||||
371 | class subdirmatcher(matcher): |
|
385 | class subdirmatcher(matcher): | |
372 | """Adapt a matcher to work on a subdirectory only. |
|
386 | """Adapt a matcher to work on a subdirectory only. | |
373 |
|
387 | |||
374 | The paths are remapped to remove/insert the path as needed: |
|
388 | The paths are remapped to remove/insert the path as needed: | |
375 |
|
389 | |||
376 | >>> m1 = match('root', '', ['a.txt', 'sub/b.txt']) |
|
390 | >>> m1 = match('root', '', ['a.txt', 'sub/b.txt']) | |
377 | >>> m2 = subdirmatcher('sub', m1) |
|
391 | >>> m2 = subdirmatcher('sub', m1) | |
378 | >>> bool(m2('a.txt')) |
|
392 | >>> bool(m2('a.txt')) | |
379 | False |
|
393 | False | |
380 | >>> bool(m2('b.txt')) |
|
394 | >>> bool(m2('b.txt')) | |
381 | True |
|
395 | True | |
382 | >>> bool(m2.matchfn('a.txt')) |
|
396 | >>> bool(m2.matchfn('a.txt')) | |
383 | False |
|
397 | False | |
384 | >>> bool(m2.matchfn('b.txt')) |
|
398 | >>> bool(m2.matchfn('b.txt')) | |
385 | True |
|
399 | True | |
386 | >>> m2.files() |
|
400 | >>> m2.files() | |
387 | ['b.txt'] |
|
401 | ['b.txt'] | |
388 | >>> m2.exact('b.txt') |
|
402 | >>> m2.exact('b.txt') | |
389 | True |
|
403 | True | |
390 | >>> util.pconvert(m2.rel('b.txt')) |
|
404 | >>> util.pconvert(m2.rel('b.txt')) | |
391 | 'sub/b.txt' |
|
405 | 'sub/b.txt' | |
392 | >>> def bad(f, msg): |
|
406 | >>> def bad(f, msg): | |
393 | ... print "%s: %s" % (f, msg) |
|
407 | ... print "%s: %s" % (f, msg) | |
394 | >>> m1.bad = bad |
|
408 | >>> m1.bad = bad | |
395 | >>> m2.bad('x.txt', 'No such file') |
|
409 | >>> m2.bad('x.txt', 'No such file') | |
396 | sub/x.txt: No such file |
|
410 | sub/x.txt: No such file | |
397 | >>> m2.abs('c.txt') |
|
411 | >>> m2.abs('c.txt') | |
398 | 'sub/c.txt' |
|
412 | 'sub/c.txt' | |
399 | """ |
|
413 | """ | |
400 |
|
414 | |||
401 | def __init__(self, path, matcher): |
|
415 | def __init__(self, path, matcher): | |
402 | self._root = matcher._root |
|
416 | self._root = matcher._root | |
403 | self._cwd = matcher._cwd |
|
417 | self._cwd = matcher._cwd | |
404 | self._path = path |
|
418 | self._path = path | |
405 | self._matcher = matcher |
|
419 | self._matcher = matcher | |
406 | self._always = matcher._always |
|
420 | self._always = matcher._always | |
407 |
|
421 | |||
408 | self._files = [f[len(path) + 1:] for f in matcher._files |
|
422 | self._files = [f[len(path) + 1:] for f in matcher._files | |
409 | if f.startswith(path + "/")] |
|
423 | if f.startswith(path + "/")] | |
410 |
|
424 | |||
411 | # If the parent repo had a path to this subrepo and the matcher is |
|
425 | # If the parent repo had a path to this subrepo and the matcher is | |
412 | # a prefix matcher, this submatcher always matches. |
|
426 | # a prefix matcher, this submatcher always matches. | |
413 | if matcher.prefix(): |
|
427 | if matcher.prefix(): | |
414 | self._always = any(f == path for f in matcher._files) |
|
428 | self._always = any(f == path for f in matcher._files) | |
415 |
|
429 | |||
416 | self._anypats = matcher._anypats |
|
430 | self._anypats = matcher._anypats | |
417 | # Some information is lost in the superclass's constructor, so we |
|
431 | # Some information is lost in the superclass's constructor, so we | |
418 | # can not accurately create the matching function for the subdirectory |
|
432 | # can not accurately create the matching function for the subdirectory | |
419 | # from the inputs. Instead, we override matchfn() and visitdir() to |
|
433 | # from the inputs. Instead, we override matchfn() and visitdir() to | |
420 | # call the original matcher with the subdirectory path prepended. |
|
434 | # call the original matcher with the subdirectory path prepended. | |
421 | self.matchfn = lambda fn: matcher.matchfn(self._path + "/" + fn) |
|
435 | self.matchfn = lambda fn: matcher.matchfn(self._path + "/" + fn) | |
422 |
|
436 | |||
423 | def bad(self, f, msg): |
|
437 | def bad(self, f, msg): | |
424 | self._matcher.bad(self._path + "/" + f, msg) |
|
438 | self._matcher.bad(self._path + "/" + f, msg) | |
425 |
|
439 | |||
426 | def abs(self, f): |
|
440 | def abs(self, f): | |
427 | return self._matcher.abs(self._path + "/" + f) |
|
441 | return self._matcher.abs(self._path + "/" + f) | |
428 |
|
442 | |||
429 | def rel(self, f): |
|
443 | def rel(self, f): | |
430 | return self._matcher.rel(self._path + "/" + f) |
|
444 | return self._matcher.rel(self._path + "/" + f) | |
431 |
|
445 | |||
432 | def uipath(self, f): |
|
446 | def uipath(self, f): | |
433 | return self._matcher.uipath(self._path + "/" + f) |
|
447 | return self._matcher.uipath(self._path + "/" + f) | |
434 |
|
448 | |||
435 | def visitdir(self, dir): |
|
449 | def visitdir(self, dir): | |
436 | if dir == '.': |
|
450 | if dir == '.': | |
437 | dir = self._path |
|
451 | dir = self._path | |
438 | else: |
|
452 | else: | |
439 | dir = self._path + "/" + dir |
|
453 | dir = self._path + "/" + dir | |
440 | return self._matcher.visitdir(dir) |
|
454 | return self._matcher.visitdir(dir) | |
441 |
|
455 | |||
442 | class icasefsmatcher(matcher): |
|
456 | class icasefsmatcher(matcher): | |
443 |
|
457 | |||
444 |
def __init__(self, root, cwd, patterns, include, exclude, |
|
458 | def __init__(self, root, cwd, normalize, patterns, include, exclude, | |
445 | ctx, listsubrepos=False, badfn=None): |
|
459 | default, auditor, ctx, listsubrepos=False, badfn=None): | |
446 | init = super(icasefsmatcher, self).__init__ |
|
460 | init = super(icasefsmatcher, self).__init__ | |
447 | self._dirstate = ctx.repo().dirstate |
|
|||
448 | self._dsnormalize = self._dirstate.normalize |
|
|||
449 |
|
||||
450 | init(root, cwd, patterns, include, exclude, default, auditor=auditor, |
|
|||
451 | ctx=ctx, listsubrepos=listsubrepos, badfn=badfn) |
|
|||
452 |
|
461 | |||
453 | def _normalize(self, patterns, default, root, cwd, auditor, warn): |
|
462 | init(root, cwd, normalize=normalize, patterns=patterns, include=include, | |
454 | kp = super(icasefsmatcher, self)._normalize(patterns, default, root, |
|
463 | exclude=exclude, default=default, auditor=auditor, | |
455 | cwd, auditor, warn) |
|
464 | ctx=ctx, listsubrepos=listsubrepos, badfn=badfn) | |
456 | kindpats = [] |
|
|||
457 | for kind, pats, source in kp: |
|
|||
458 | if kind not in ('re', 'relre'): # regex can't be normalized |
|
|||
459 | p = pats |
|
|||
460 | pats = self._dsnormalize(pats) |
|
|||
461 |
|
||||
462 | # Preserve the original to handle a case only rename. |
|
|||
463 | if p != pats and p in self._dirstate: |
|
|||
464 | kindpats.append((kind, p, source)) |
|
|||
465 |
|
||||
466 | kindpats.append((kind, pats, source)) |
|
|||
467 | return kindpats |
|
|||
468 |
|
465 | |||
469 | def patkind(pattern, default=None): |
|
466 | def patkind(pattern, default=None): | |
470 | '''If pattern is 'kind:pat' with a known kind, return kind.''' |
|
467 | '''If pattern is 'kind:pat' with a known kind, return kind.''' | |
471 | return _patsplit(pattern, default)[0] |
|
468 | return _patsplit(pattern, default)[0] | |
472 |
|
469 | |||
473 | def _patsplit(pattern, default): |
|
470 | def _patsplit(pattern, default): | |
474 | """Split a string into the optional pattern kind prefix and the actual |
|
471 | """Split a string into the optional pattern kind prefix and the actual | |
475 | pattern.""" |
|
472 | pattern.""" | |
476 | if ':' in pattern: |
|
473 | if ':' in pattern: | |
477 | kind, pat = pattern.split(':', 1) |
|
474 | kind, pat = pattern.split(':', 1) | |
478 | if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre', |
|
475 | if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre', | |
479 | 'listfile', 'listfile0', 'set', 'include', 'subinclude', |
|
476 | 'listfile', 'listfile0', 'set', 'include', 'subinclude', | |
480 | 'rootfilesin'): |
|
477 | 'rootfilesin'): | |
481 | return kind, pat |
|
478 | return kind, pat | |
482 | return default, pattern |
|
479 | return default, pattern | |
483 |
|
480 | |||
484 | def _globre(pat): |
|
481 | def _globre(pat): | |
485 | r'''Convert an extended glob string to a regexp string. |
|
482 | r'''Convert an extended glob string to a regexp string. | |
486 |
|
483 | |||
487 | >>> print _globre(r'?') |
|
484 | >>> print _globre(r'?') | |
488 | . |
|
485 | . | |
489 | >>> print _globre(r'*') |
|
486 | >>> print _globre(r'*') | |
490 | [^/]* |
|
487 | [^/]* | |
491 | >>> print _globre(r'**') |
|
488 | >>> print _globre(r'**') | |
492 | .* |
|
489 | .* | |
493 | >>> print _globre(r'**/a') |
|
490 | >>> print _globre(r'**/a') | |
494 | (?:.*/)?a |
|
491 | (?:.*/)?a | |
495 | >>> print _globre(r'a/**/b') |
|
492 | >>> print _globre(r'a/**/b') | |
496 | a\/(?:.*/)?b |
|
493 | a\/(?:.*/)?b | |
497 | >>> print _globre(r'[a*?!^][^b][!c]') |
|
494 | >>> print _globre(r'[a*?!^][^b][!c]') | |
498 | [a*?!^][\^b][^c] |
|
495 | [a*?!^][\^b][^c] | |
499 | >>> print _globre(r'{a,b}') |
|
496 | >>> print _globre(r'{a,b}') | |
500 | (?:a|b) |
|
497 | (?:a|b) | |
501 | >>> print _globre(r'.\*\?') |
|
498 | >>> print _globre(r'.\*\?') | |
502 | \.\*\? |
|
499 | \.\*\? | |
503 | ''' |
|
500 | ''' | |
504 | i, n = 0, len(pat) |
|
501 | i, n = 0, len(pat) | |
505 | res = '' |
|
502 | res = '' | |
506 | group = 0 |
|
503 | group = 0 | |
507 | escape = util.re.escape |
|
504 | escape = util.re.escape | |
508 | def peek(): |
|
505 | def peek(): | |
509 | return i < n and pat[i:i + 1] |
|
506 | return i < n and pat[i:i + 1] | |
510 | while i < n: |
|
507 | while i < n: | |
511 | c = pat[i:i + 1] |
|
508 | c = pat[i:i + 1] | |
512 | i += 1 |
|
509 | i += 1 | |
513 | if c not in '*?[{},\\': |
|
510 | if c not in '*?[{},\\': | |
514 | res += escape(c) |
|
511 | res += escape(c) | |
515 | elif c == '*': |
|
512 | elif c == '*': | |
516 | if peek() == '*': |
|
513 | if peek() == '*': | |
517 | i += 1 |
|
514 | i += 1 | |
518 | if peek() == '/': |
|
515 | if peek() == '/': | |
519 | i += 1 |
|
516 | i += 1 | |
520 | res += '(?:.*/)?' |
|
517 | res += '(?:.*/)?' | |
521 | else: |
|
518 | else: | |
522 | res += '.*' |
|
519 | res += '.*' | |
523 | else: |
|
520 | else: | |
524 | res += '[^/]*' |
|
521 | res += '[^/]*' | |
525 | elif c == '?': |
|
522 | elif c == '?': | |
526 | res += '.' |
|
523 | res += '.' | |
527 | elif c == '[': |
|
524 | elif c == '[': | |
528 | j = i |
|
525 | j = i | |
529 | if j < n and pat[j:j + 1] in '!]': |
|
526 | if j < n and pat[j:j + 1] in '!]': | |
530 | j += 1 |
|
527 | j += 1 | |
531 | while j < n and pat[j:j + 1] != ']': |
|
528 | while j < n and pat[j:j + 1] != ']': | |
532 | j += 1 |
|
529 | j += 1 | |
533 | if j >= n: |
|
530 | if j >= n: | |
534 | res += '\\[' |
|
531 | res += '\\[' | |
535 | else: |
|
532 | else: | |
536 | stuff = pat[i:j].replace('\\','\\\\') |
|
533 | stuff = pat[i:j].replace('\\','\\\\') | |
537 | i = j + 1 |
|
534 | i = j + 1 | |
538 | if stuff[0:1] == '!': |
|
535 | if stuff[0:1] == '!': | |
539 | stuff = '^' + stuff[1:] |
|
536 | stuff = '^' + stuff[1:] | |
540 | elif stuff[0:1] == '^': |
|
537 | elif stuff[0:1] == '^': | |
541 | stuff = '\\' + stuff |
|
538 | stuff = '\\' + stuff | |
542 | res = '%s[%s]' % (res, stuff) |
|
539 | res = '%s[%s]' % (res, stuff) | |
543 | elif c == '{': |
|
540 | elif c == '{': | |
544 | group += 1 |
|
541 | group += 1 | |
545 | res += '(?:' |
|
542 | res += '(?:' | |
546 | elif c == '}' and group: |
|
543 | elif c == '}' and group: | |
547 | res += ')' |
|
544 | res += ')' | |
548 | group -= 1 |
|
545 | group -= 1 | |
549 | elif c == ',' and group: |
|
546 | elif c == ',' and group: | |
550 | res += '|' |
|
547 | res += '|' | |
551 | elif c == '\\': |
|
548 | elif c == '\\': | |
552 | p = peek() |
|
549 | p = peek() | |
553 | if p: |
|
550 | if p: | |
554 | i += 1 |
|
551 | i += 1 | |
555 | res += escape(p) |
|
552 | res += escape(p) | |
556 | else: |
|
553 | else: | |
557 | res += escape(c) |
|
554 | res += escape(c) | |
558 | else: |
|
555 | else: | |
559 | res += escape(c) |
|
556 | res += escape(c) | |
560 | return res |
|
557 | return res | |
561 |
|
558 | |||
562 | def _regex(kind, pat, globsuffix): |
|
559 | def _regex(kind, pat, globsuffix): | |
563 | '''Convert a (normalized) pattern of any kind into a regular expression. |
|
560 | '''Convert a (normalized) pattern of any kind into a regular expression. | |
564 | globsuffix is appended to the regexp of globs.''' |
|
561 | globsuffix is appended to the regexp of globs.''' | |
565 | if not pat: |
|
562 | if not pat: | |
566 | return '' |
|
563 | return '' | |
567 | if kind == 're': |
|
564 | if kind == 're': | |
568 | return pat |
|
565 | return pat | |
569 | if kind == 'path': |
|
566 | if kind == 'path': | |
570 | if pat == '.': |
|
567 | if pat == '.': | |
571 | return '' |
|
568 | return '' | |
572 | return '^' + util.re.escape(pat) + '(?:/|$)' |
|
569 | return '^' + util.re.escape(pat) + '(?:/|$)' | |
573 | if kind == 'rootfilesin': |
|
570 | if kind == 'rootfilesin': | |
574 | if pat == '.': |
|
571 | if pat == '.': | |
575 | escaped = '' |
|
572 | escaped = '' | |
576 | else: |
|
573 | else: | |
577 | # Pattern is a directory name. |
|
574 | # Pattern is a directory name. | |
578 | escaped = util.re.escape(pat) + '/' |
|
575 | escaped = util.re.escape(pat) + '/' | |
579 | # Anything after the pattern must be a non-directory. |
|
576 | # Anything after the pattern must be a non-directory. | |
580 | return '^' + escaped + '[^/]+$' |
|
577 | return '^' + escaped + '[^/]+$' | |
581 | if kind == 'relglob': |
|
578 | if kind == 'relglob': | |
582 | return '(?:|.*/)' + _globre(pat) + globsuffix |
|
579 | return '(?:|.*/)' + _globre(pat) + globsuffix | |
583 | if kind == 'relpath': |
|
580 | if kind == 'relpath': | |
584 | return util.re.escape(pat) + '(?:/|$)' |
|
581 | return util.re.escape(pat) + '(?:/|$)' | |
585 | if kind == 'relre': |
|
582 | if kind == 'relre': | |
586 | if pat.startswith('^'): |
|
583 | if pat.startswith('^'): | |
587 | return pat |
|
584 | return pat | |
588 | return '.*' + pat |
|
585 | return '.*' + pat | |
589 | return _globre(pat) + globsuffix |
|
586 | return _globre(pat) + globsuffix | |
590 |
|
587 | |||
591 | def _buildmatch(ctx, kindpats, globsuffix, listsubrepos, root): |
|
588 | def _buildmatch(ctx, kindpats, globsuffix, listsubrepos, root): | |
592 | '''Return regexp string and a matcher function for kindpats. |
|
589 | '''Return regexp string and a matcher function for kindpats. | |
593 | globsuffix is appended to the regexp of globs.''' |
|
590 | globsuffix is appended to the regexp of globs.''' | |
594 | matchfuncs = [] |
|
591 | matchfuncs = [] | |
595 |
|
592 | |||
596 | subincludes, kindpats = _expandsubinclude(kindpats, root) |
|
593 | subincludes, kindpats = _expandsubinclude(kindpats, root) | |
597 | if subincludes: |
|
594 | if subincludes: | |
598 | submatchers = {} |
|
595 | submatchers = {} | |
599 | def matchsubinclude(f): |
|
596 | def matchsubinclude(f): | |
600 | for prefix, matcherargs in subincludes: |
|
597 | for prefix, matcherargs in subincludes: | |
601 | if f.startswith(prefix): |
|
598 | if f.startswith(prefix): | |
602 | mf = submatchers.get(prefix) |
|
599 | mf = submatchers.get(prefix) | |
603 | if mf is None: |
|
600 | if mf is None: | |
604 | mf = match(*matcherargs) |
|
601 | mf = match(*matcherargs) | |
605 | submatchers[prefix] = mf |
|
602 | submatchers[prefix] = mf | |
606 |
|
603 | |||
607 | if mf(f[len(prefix):]): |
|
604 | if mf(f[len(prefix):]): | |
608 | return True |
|
605 | return True | |
609 | return False |
|
606 | return False | |
610 | matchfuncs.append(matchsubinclude) |
|
607 | matchfuncs.append(matchsubinclude) | |
611 |
|
608 | |||
612 | fset, kindpats = _expandsets(kindpats, ctx, listsubrepos) |
|
609 | fset, kindpats = _expandsets(kindpats, ctx, listsubrepos) | |
613 | if fset: |
|
610 | if fset: | |
614 | matchfuncs.append(fset.__contains__) |
|
611 | matchfuncs.append(fset.__contains__) | |
615 |
|
612 | |||
616 | regex = '' |
|
613 | regex = '' | |
617 | if kindpats: |
|
614 | if kindpats: | |
618 | regex, mf = _buildregexmatch(kindpats, globsuffix) |
|
615 | regex, mf = _buildregexmatch(kindpats, globsuffix) | |
619 | matchfuncs.append(mf) |
|
616 | matchfuncs.append(mf) | |
620 |
|
617 | |||
621 | if len(matchfuncs) == 1: |
|
618 | if len(matchfuncs) == 1: | |
622 | return regex, matchfuncs[0] |
|
619 | return regex, matchfuncs[0] | |
623 | else: |
|
620 | else: | |
624 | return regex, lambda f: any(mf(f) for mf in matchfuncs) |
|
621 | return regex, lambda f: any(mf(f) for mf in matchfuncs) | |
625 |
|
622 | |||
626 | def _buildregexmatch(kindpats, globsuffix): |
|
623 | def _buildregexmatch(kindpats, globsuffix): | |
627 | """Build a match function from a list of kinds and kindpats, |
|
624 | """Build a match function from a list of kinds and kindpats, | |
628 | return regexp string and a matcher function.""" |
|
625 | return regexp string and a matcher function.""" | |
629 | try: |
|
626 | try: | |
630 | regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix) |
|
627 | regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix) | |
631 | for (k, p, s) in kindpats]) |
|
628 | for (k, p, s) in kindpats]) | |
632 | if len(regex) > 20000: |
|
629 | if len(regex) > 20000: | |
633 | raise OverflowError |
|
630 | raise OverflowError | |
634 | return regex, _rematcher(regex) |
|
631 | return regex, _rematcher(regex) | |
635 | except OverflowError: |
|
632 | except OverflowError: | |
636 | # We're using a Python with a tiny regex engine and we |
|
633 | # We're using a Python with a tiny regex engine and we | |
637 | # made it explode, so we'll divide the pattern list in two |
|
634 | # made it explode, so we'll divide the pattern list in two | |
638 | # until it works |
|
635 | # until it works | |
639 | l = len(kindpats) |
|
636 | l = len(kindpats) | |
640 | if l < 2: |
|
637 | if l < 2: | |
641 | raise |
|
638 | raise | |
642 | regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix) |
|
639 | regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix) | |
643 | regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix) |
|
640 | regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix) | |
644 | return regex, lambda s: a(s) or b(s) |
|
641 | return regex, lambda s: a(s) or b(s) | |
645 | except re.error: |
|
642 | except re.error: | |
646 | for k, p, s in kindpats: |
|
643 | for k, p, s in kindpats: | |
647 | try: |
|
644 | try: | |
648 | _rematcher('(?:%s)' % _regex(k, p, globsuffix)) |
|
645 | _rematcher('(?:%s)' % _regex(k, p, globsuffix)) | |
649 | except re.error: |
|
646 | except re.error: | |
650 | if s: |
|
647 | if s: | |
651 | raise error.Abort(_("%s: invalid pattern (%s): %s") % |
|
648 | raise error.Abort(_("%s: invalid pattern (%s): %s") % | |
652 | (s, k, p)) |
|
649 | (s, k, p)) | |
653 | else: |
|
650 | else: | |
654 | raise error.Abort(_("invalid pattern (%s): %s") % (k, p)) |
|
651 | raise error.Abort(_("invalid pattern (%s): %s") % (k, p)) | |
655 | raise error.Abort(_("invalid pattern")) |
|
652 | raise error.Abort(_("invalid pattern")) | |
656 |
|
653 | |||
657 | def _patternrootsanddirs(kindpats): |
|
654 | def _patternrootsanddirs(kindpats): | |
658 | '''Returns roots and directories corresponding to each pattern. |
|
655 | '''Returns roots and directories corresponding to each pattern. | |
659 |
|
656 | |||
660 | This calculates the roots and directories exactly matching the patterns and |
|
657 | This calculates the roots and directories exactly matching the patterns and | |
661 | returns a tuple of (roots, dirs) for each. It does not return other |
|
658 | returns a tuple of (roots, dirs) for each. It does not return other | |
662 | directories which may also need to be considered, like the parent |
|
659 | directories which may also need to be considered, like the parent | |
663 | directories. |
|
660 | directories. | |
664 | ''' |
|
661 | ''' | |
665 | r = [] |
|
662 | r = [] | |
666 | d = [] |
|
663 | d = [] | |
667 | for kind, pat, source in kindpats: |
|
664 | for kind, pat, source in kindpats: | |
668 | if kind == 'glob': # find the non-glob prefix |
|
665 | if kind == 'glob': # find the non-glob prefix | |
669 | root = [] |
|
666 | root = [] | |
670 | for p in pat.split('/'): |
|
667 | for p in pat.split('/'): | |
671 | if '[' in p or '{' in p or '*' in p or '?' in p: |
|
668 | if '[' in p or '{' in p or '*' in p or '?' in p: | |
672 | break |
|
669 | break | |
673 | root.append(p) |
|
670 | root.append(p) | |
674 | r.append('/'.join(root) or '.') |
|
671 | r.append('/'.join(root) or '.') | |
675 | elif kind in ('relpath', 'path'): |
|
672 | elif kind in ('relpath', 'path'): | |
676 | r.append(pat or '.') |
|
673 | r.append(pat or '.') | |
677 | elif kind in ('rootfilesin',): |
|
674 | elif kind in ('rootfilesin',): | |
678 | d.append(pat or '.') |
|
675 | d.append(pat or '.') | |
679 | else: # relglob, re, relre |
|
676 | else: # relglob, re, relre | |
680 | r.append('.') |
|
677 | r.append('.') | |
681 | return r, d |
|
678 | return r, d | |
682 |
|
679 | |||
683 | def _roots(kindpats): |
|
680 | def _roots(kindpats): | |
684 | '''Returns root directories to match recursively from the given patterns.''' |
|
681 | '''Returns root directories to match recursively from the given patterns.''' | |
685 | roots, dirs = _patternrootsanddirs(kindpats) |
|
682 | roots, dirs = _patternrootsanddirs(kindpats) | |
686 | return roots |
|
683 | return roots | |
687 |
|
684 | |||
688 | def _rootsanddirs(kindpats): |
|
685 | def _rootsanddirs(kindpats): | |
689 | '''Returns roots and exact directories from patterns. |
|
686 | '''Returns roots and exact directories from patterns. | |
690 |
|
687 | |||
691 | roots are directories to match recursively, whereas exact directories should |
|
688 | roots are directories to match recursively, whereas exact directories should | |
692 | be matched non-recursively. The returned (roots, dirs) tuple will also |
|
689 | be matched non-recursively. The returned (roots, dirs) tuple will also | |
693 | include directories that need to be implicitly considered as either, such as |
|
690 | include directories that need to be implicitly considered as either, such as | |
694 | parent directories. |
|
691 | parent directories. | |
695 |
|
692 | |||
696 | >>> _rootsanddirs(\ |
|
693 | >>> _rootsanddirs(\ | |
697 | [('glob', 'g/h/*', ''), ('glob', 'g/h', ''), ('glob', 'g*', '')]) |
|
694 | [('glob', 'g/h/*', ''), ('glob', 'g/h', ''), ('glob', 'g*', '')]) | |
698 | (['g/h', 'g/h', '.'], ['g', '.']) |
|
695 | (['g/h', 'g/h', '.'], ['g', '.']) | |
699 | >>> _rootsanddirs(\ |
|
696 | >>> _rootsanddirs(\ | |
700 | [('rootfilesin', 'g/h', ''), ('rootfilesin', '', '')]) |
|
697 | [('rootfilesin', 'g/h', ''), ('rootfilesin', '', '')]) | |
701 | ([], ['g/h', '.', 'g', '.']) |
|
698 | ([], ['g/h', '.', 'g', '.']) | |
702 | >>> _rootsanddirs(\ |
|
699 | >>> _rootsanddirs(\ | |
703 | [('relpath', 'r', ''), ('path', 'p/p', ''), ('path', '', '')]) |
|
700 | [('relpath', 'r', ''), ('path', 'p/p', ''), ('path', '', '')]) | |
704 | (['r', 'p/p', '.'], ['p', '.']) |
|
701 | (['r', 'p/p', '.'], ['p', '.']) | |
705 | >>> _rootsanddirs(\ |
|
702 | >>> _rootsanddirs(\ | |
706 | [('relglob', 'rg*', ''), ('re', 're/', ''), ('relre', 'rr', '')]) |
|
703 | [('relglob', 'rg*', ''), ('re', 're/', ''), ('relre', 'rr', '')]) | |
707 | (['.', '.', '.'], ['.']) |
|
704 | (['.', '.', '.'], ['.']) | |
708 | ''' |
|
705 | ''' | |
709 | r, d = _patternrootsanddirs(kindpats) |
|
706 | r, d = _patternrootsanddirs(kindpats) | |
710 |
|
707 | |||
711 | # Append the parents as non-recursive/exact directories, since they must be |
|
708 | # Append the parents as non-recursive/exact directories, since they must be | |
712 | # scanned to get to either the roots or the other exact directories. |
|
709 | # scanned to get to either the roots or the other exact directories. | |
713 | d.extend(util.dirs(d)) |
|
710 | d.extend(util.dirs(d)) | |
714 | d.extend(util.dirs(r)) |
|
711 | d.extend(util.dirs(r)) | |
715 | # util.dirs() does not include the root directory, so add it manually |
|
712 | # util.dirs() does not include the root directory, so add it manually | |
716 | d.append('.') |
|
713 | d.append('.') | |
717 |
|
714 | |||
718 | return r, d |
|
715 | return r, d | |
719 |
|
716 | |||
720 | def _explicitfiles(kindpats): |
|
717 | def _explicitfiles(kindpats): | |
721 | '''Returns the potential explicit filenames from the patterns. |
|
718 | '''Returns the potential explicit filenames from the patterns. | |
722 |
|
719 | |||
723 | >>> _explicitfiles([('path', 'foo/bar', '')]) |
|
720 | >>> _explicitfiles([('path', 'foo/bar', '')]) | |
724 | ['foo/bar'] |
|
721 | ['foo/bar'] | |
725 | >>> _explicitfiles([('rootfilesin', 'foo/bar', '')]) |
|
722 | >>> _explicitfiles([('rootfilesin', 'foo/bar', '')]) | |
726 | [] |
|
723 | [] | |
727 | ''' |
|
724 | ''' | |
728 | # Keep only the pattern kinds where one can specify filenames (vs only |
|
725 | # Keep only the pattern kinds where one can specify filenames (vs only | |
729 | # directory names). |
|
726 | # directory names). | |
730 | filable = [kp for kp in kindpats if kp[0] not in ('rootfilesin',)] |
|
727 | filable = [kp for kp in kindpats if kp[0] not in ('rootfilesin',)] | |
731 | return _roots(filable) |
|
728 | return _roots(filable) | |
732 |
|
729 | |||
733 | def _anypats(kindpats): |
|
730 | def _anypats(kindpats): | |
734 | for kind, pat, source in kindpats: |
|
731 | for kind, pat, source in kindpats: | |
735 | if kind in ('glob', 're', 'relglob', 'relre', 'set', 'rootfilesin'): |
|
732 | if kind in ('glob', 're', 'relglob', 'relre', 'set', 'rootfilesin'): | |
736 | return True |
|
733 | return True | |
737 |
|
734 | |||
738 | _commentre = None |
|
735 | _commentre = None | |
739 |
|
736 | |||
740 | def readpatternfile(filepath, warn, sourceinfo=False): |
|
737 | def readpatternfile(filepath, warn, sourceinfo=False): | |
741 | '''parse a pattern file, returning a list of |
|
738 | '''parse a pattern file, returning a list of | |
742 | patterns. These patterns should be given to compile() |
|
739 | patterns. These patterns should be given to compile() | |
743 | to be validated and converted into a match function. |
|
740 | to be validated and converted into a match function. | |
744 |
|
741 | |||
745 | trailing white space is dropped. |
|
742 | trailing white space is dropped. | |
746 | the escape character is backslash. |
|
743 | the escape character is backslash. | |
747 | comments start with #. |
|
744 | comments start with #. | |
748 | empty lines are skipped. |
|
745 | empty lines are skipped. | |
749 |
|
746 | |||
750 | lines can be of the following formats: |
|
747 | lines can be of the following formats: | |
751 |
|
748 | |||
752 | syntax: regexp # defaults following lines to non-rooted regexps |
|
749 | syntax: regexp # defaults following lines to non-rooted regexps | |
753 | syntax: glob # defaults following lines to non-rooted globs |
|
750 | syntax: glob # defaults following lines to non-rooted globs | |
754 | re:pattern # non-rooted regular expression |
|
751 | re:pattern # non-rooted regular expression | |
755 | glob:pattern # non-rooted glob |
|
752 | glob:pattern # non-rooted glob | |
756 | pattern # pattern of the current default type |
|
753 | pattern # pattern of the current default type | |
757 |
|
754 | |||
758 | if sourceinfo is set, returns a list of tuples: |
|
755 | if sourceinfo is set, returns a list of tuples: | |
759 | (pattern, lineno, originalline). This is useful to debug ignore patterns. |
|
756 | (pattern, lineno, originalline). This is useful to debug ignore patterns. | |
760 | ''' |
|
757 | ''' | |
761 |
|
758 | |||
762 | syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:', |
|
759 | syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:', | |
763 | 'include': 'include', 'subinclude': 'subinclude'} |
|
760 | 'include': 'include', 'subinclude': 'subinclude'} | |
764 | syntax = 'relre:' |
|
761 | syntax = 'relre:' | |
765 | patterns = [] |
|
762 | patterns = [] | |
766 |
|
763 | |||
767 | fp = open(filepath, 'rb') |
|
764 | fp = open(filepath, 'rb') | |
768 | for lineno, line in enumerate(util.iterfile(fp), start=1): |
|
765 | for lineno, line in enumerate(util.iterfile(fp), start=1): | |
769 | if "#" in line: |
|
766 | if "#" in line: | |
770 | global _commentre |
|
767 | global _commentre | |
771 | if not _commentre: |
|
768 | if not _commentre: | |
772 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') |
|
769 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') | |
773 | # remove comments prefixed by an even number of escapes |
|
770 | # remove comments prefixed by an even number of escapes | |
774 | m = _commentre.search(line) |
|
771 | m = _commentre.search(line) | |
775 | if m: |
|
772 | if m: | |
776 | line = line[:m.end(1)] |
|
773 | line = line[:m.end(1)] | |
777 | # fixup properly escaped comments that survived the above |
|
774 | # fixup properly escaped comments that survived the above | |
778 | line = line.replace("\\#", "#") |
|
775 | line = line.replace("\\#", "#") | |
779 | line = line.rstrip() |
|
776 | line = line.rstrip() | |
780 | if not line: |
|
777 | if not line: | |
781 | continue |
|
778 | continue | |
782 |
|
779 | |||
783 | if line.startswith('syntax:'): |
|
780 | if line.startswith('syntax:'): | |
784 | s = line[7:].strip() |
|
781 | s = line[7:].strip() | |
785 | try: |
|
782 | try: | |
786 | syntax = syntaxes[s] |
|
783 | syntax = syntaxes[s] | |
787 | except KeyError: |
|
784 | except KeyError: | |
788 | if warn: |
|
785 | if warn: | |
789 | warn(_("%s: ignoring invalid syntax '%s'\n") % |
|
786 | warn(_("%s: ignoring invalid syntax '%s'\n") % | |
790 | (filepath, s)) |
|
787 | (filepath, s)) | |
791 | continue |
|
788 | continue | |
792 |
|
789 | |||
793 | linesyntax = syntax |
|
790 | linesyntax = syntax | |
794 | for s, rels in syntaxes.iteritems(): |
|
791 | for s, rels in syntaxes.iteritems(): | |
795 | if line.startswith(rels): |
|
792 | if line.startswith(rels): | |
796 | linesyntax = rels |
|
793 | linesyntax = rels | |
797 | line = line[len(rels):] |
|
794 | line = line[len(rels):] | |
798 | break |
|
795 | break | |
799 | elif line.startswith(s+':'): |
|
796 | elif line.startswith(s+':'): | |
800 | linesyntax = rels |
|
797 | linesyntax = rels | |
801 | line = line[len(s) + 1:] |
|
798 | line = line[len(s) + 1:] | |
802 | break |
|
799 | break | |
803 | if sourceinfo: |
|
800 | if sourceinfo: | |
804 | patterns.append((linesyntax + line, lineno, line)) |
|
801 | patterns.append((linesyntax + line, lineno, line)) | |
805 | else: |
|
802 | else: | |
806 | patterns.append(linesyntax + line) |
|
803 | patterns.append(linesyntax + line) | |
807 | fp.close() |
|
804 | fp.close() | |
808 | return patterns |
|
805 | return patterns |
General Comments 0
You need to be logged in to leave comments.
Login now