Show More
@@ -1,826 +1,833 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.ProgrammingError("fileset expression with no " |
|
41 | raise error.ProgrammingError("fileset expression with no " | |
42 | "context") |
|
42 | "context") | |
43 | s = ctx.getfileset(pat) |
|
43 | s = ctx.getfileset(pat) | |
44 | fset.update(s) |
|
44 | fset.update(s) | |
45 |
|
45 | |||
46 | if listsubrepos: |
|
46 | if listsubrepos: | |
47 | for subpath in ctx.substate: |
|
47 | for subpath in ctx.substate: | |
48 | s = ctx.sub(subpath).getfileset(pat) |
|
48 | s = ctx.sub(subpath).getfileset(pat) | |
49 | fset.update(subpath + '/' + f for f in s) |
|
49 | fset.update(subpath + '/' + f for f in s) | |
50 |
|
50 | |||
51 | continue |
|
51 | continue | |
52 | other.append((kind, pat, source)) |
|
52 | other.append((kind, pat, source)) | |
53 | return fset, other |
|
53 | return fset, other | |
54 |
|
54 | |||
55 | def _expandsubinclude(kindpats, root): |
|
55 | def _expandsubinclude(kindpats, root): | |
56 | '''Returns the list of subinclude matcher args and the kindpats without the |
|
56 | '''Returns the list of subinclude matcher args and the kindpats without the | |
57 | subincludes in it.''' |
|
57 | subincludes in it.''' | |
58 | relmatchers = [] |
|
58 | relmatchers = [] | |
59 | other = [] |
|
59 | other = [] | |
60 |
|
60 | |||
61 | for kind, pat, source in kindpats: |
|
61 | for kind, pat, source in kindpats: | |
62 | if kind == 'subinclude': |
|
62 | if kind == 'subinclude': | |
63 | sourceroot = pathutil.dirname(util.normpath(source)) |
|
63 | sourceroot = pathutil.dirname(util.normpath(source)) | |
64 | pat = util.pconvert(pat) |
|
64 | pat = util.pconvert(pat) | |
65 | path = pathutil.join(sourceroot, pat) |
|
65 | path = pathutil.join(sourceroot, pat) | |
66 |
|
66 | |||
67 | newroot = pathutil.dirname(path) |
|
67 | newroot = pathutil.dirname(path) | |
68 | matcherargs = (newroot, '', [], ['include:%s' % path]) |
|
68 | matcherargs = (newroot, '', [], ['include:%s' % path]) | |
69 |
|
69 | |||
70 | prefix = pathutil.canonpath(root, root, newroot) |
|
70 | prefix = pathutil.canonpath(root, root, newroot) | |
71 | if prefix: |
|
71 | if prefix: | |
72 | prefix += '/' |
|
72 | prefix += '/' | |
73 | relmatchers.append((prefix, matcherargs)) |
|
73 | relmatchers.append((prefix, matcherargs)) | |
74 | else: |
|
74 | else: | |
75 | other.append((kind, pat, source)) |
|
75 | other.append((kind, pat, source)) | |
76 |
|
76 | |||
77 | return relmatchers, other |
|
77 | return relmatchers, other | |
78 |
|
78 | |||
79 | def _kindpatsalwaysmatch(kindpats): |
|
79 | def _kindpatsalwaysmatch(kindpats): | |
80 | """"Checks whether the kindspats match everything, as e.g. |
|
80 | """"Checks whether the kindspats match everything, as e.g. | |
81 | 'relpath:.' does. |
|
81 | 'relpath:.' does. | |
82 | """ |
|
82 | """ | |
83 | for kind, pat, source in kindpats: |
|
83 | for kind, pat, source in kindpats: | |
84 | if pat != '' or kind not in ['relpath', 'glob']: |
|
84 | if pat != '' or kind not in ['relpath', 'glob']: | |
85 | return False |
|
85 | return False | |
86 | return True |
|
86 | return True | |
87 |
|
87 | |||
88 | def match(root, cwd, patterns, include=None, exclude=None, default='glob', |
|
88 | def match(root, cwd, patterns, include=None, exclude=None, default='glob', | |
89 | exact=False, auditor=None, ctx=None, listsubrepos=False, warn=None, |
|
89 | exact=False, auditor=None, ctx=None, listsubrepos=False, warn=None, | |
90 | badfn=None, icasefs=False): |
|
90 | badfn=None, icasefs=False): | |
91 | """build an object to match a set of file patterns |
|
91 | """build an object to match a set of file patterns | |
92 |
|
92 | |||
93 | arguments: |
|
93 | arguments: | |
94 | root - the canonical root of the tree you're matching against |
|
94 | root - the canonical root of the tree you're matching against | |
95 | cwd - the current working directory, if relevant |
|
95 | cwd - the current working directory, if relevant | |
96 | patterns - patterns to find |
|
96 | patterns - patterns to find | |
97 | include - patterns to include (unless they are excluded) |
|
97 | include - patterns to include (unless they are excluded) | |
98 | exclude - patterns to exclude (even if they are included) |
|
98 | exclude - patterns to exclude (even if they are included) | |
99 | default - if a pattern in patterns has no explicit type, assume this one |
|
99 | default - if a pattern in patterns has no explicit type, assume this one | |
100 | exact - patterns are actually filenames (include/exclude still apply) |
|
100 | exact - patterns are actually filenames (include/exclude still apply) | |
101 | warn - optional function used for printing warnings |
|
101 | warn - optional function used for printing warnings | |
102 | badfn - optional bad() callback for this matcher instead of the default |
|
102 | badfn - optional bad() callback for this matcher instead of the default | |
103 | icasefs - make a matcher for wdir on case insensitive filesystems, which |
|
103 | icasefs - make a matcher for wdir on case insensitive filesystems, which | |
104 | normalizes the given patterns to the case in the filesystem |
|
104 | normalizes the given patterns to the case in the filesystem | |
105 |
|
105 | |||
106 | a pattern is one of: |
|
106 | a pattern is one of: | |
107 | 'glob:<glob>' - a glob relative to cwd |
|
107 | 'glob:<glob>' - a glob relative to cwd | |
108 | 're:<regexp>' - a regular expression |
|
108 | 're:<regexp>' - a regular expression | |
109 | 'path:<path>' - a path relative to repository root, which is matched |
|
109 | 'path:<path>' - a path relative to repository root, which is matched | |
110 | recursively |
|
110 | recursively | |
111 | 'rootfilesin:<path>' - a path relative to repository root, which is |
|
111 | 'rootfilesin:<path>' - a path relative to repository root, which is | |
112 | matched non-recursively (will not match subdirectories) |
|
112 | matched non-recursively (will not match subdirectories) | |
113 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
113 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
114 | 'relpath:<path>' - a path relative to cwd |
|
114 | 'relpath:<path>' - a path relative to cwd | |
115 | 'relre:<regexp>' - a regexp that needn't match the start of a name |
|
115 | 'relre:<regexp>' - a regexp that needn't match the start of a name | |
116 | 'set:<fileset>' - a fileset expression |
|
116 | 'set:<fileset>' - a fileset expression | |
117 | 'include:<path>' - a file of patterns to read and include |
|
117 | 'include:<path>' - a file of patterns to read and include | |
118 | 'subinclude:<path>' - a file of patterns to match against files under |
|
118 | 'subinclude:<path>' - a file of patterns to match against files under | |
119 | the same directory |
|
119 | the same directory | |
120 | '<something>' - a pattern of the specified default type |
|
120 | '<something>' - a pattern of the specified default type | |
121 | """ |
|
121 | """ | |
122 | normalize = _donormalize |
|
122 | normalize = _donormalize | |
123 | if icasefs: |
|
123 | if icasefs: | |
124 | if exact: |
|
124 | if exact: | |
125 | raise error.ProgrammingError("a case-insensitive exact matcher " |
|
125 | raise error.ProgrammingError("a case-insensitive exact matcher " | |
126 | "doesn't make sense") |
|
126 | "doesn't make sense") | |
127 | dirstate = ctx.repo().dirstate |
|
127 | dirstate = ctx.repo().dirstate | |
128 | dsnormalize = dirstate.normalize |
|
128 | dsnormalize = dirstate.normalize | |
129 |
|
129 | |||
130 | def normalize(patterns, default, root, cwd, auditor, warn): |
|
130 | def normalize(patterns, default, root, cwd, auditor, warn): | |
131 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) |
|
131 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) | |
132 | kindpats = [] |
|
132 | kindpats = [] | |
133 | for kind, pats, source in kp: |
|
133 | for kind, pats, source in kp: | |
134 | if kind not in ('re', 'relre'): # regex can't be normalized |
|
134 | if kind not in ('re', 'relre'): # regex can't be normalized | |
135 | p = pats |
|
135 | p = pats | |
136 | pats = dsnormalize(pats) |
|
136 | pats = dsnormalize(pats) | |
137 |
|
137 | |||
138 | # Preserve the original to handle a case only rename. |
|
138 | # Preserve the original to handle a case only rename. | |
139 | if p != pats and p in dirstate: |
|
139 | if p != pats and p in dirstate: | |
140 | kindpats.append((kind, p, source)) |
|
140 | kindpats.append((kind, p, source)) | |
141 |
|
141 | |||
142 | kindpats.append((kind, pats, source)) |
|
142 | kindpats.append((kind, pats, source)) | |
143 | return kindpats |
|
143 | return kindpats | |
144 |
|
144 | |||
145 | return matcher(root, cwd, normalize, patterns, include=include, |
|
145 | return matcher(root, cwd, normalize, patterns, include=include, | |
146 | exclude=exclude, default=default, exact=exact, |
|
146 | exclude=exclude, default=default, exact=exact, | |
147 | auditor=auditor, ctx=ctx, listsubrepos=listsubrepos, |
|
147 | auditor=auditor, ctx=ctx, listsubrepos=listsubrepos, | |
148 | warn=warn, badfn=badfn) |
|
148 | warn=warn, badfn=badfn) | |
149 |
|
149 | |||
150 | def exact(root, cwd, files, badfn=None): |
|
150 | def exact(root, cwd, files, badfn=None): | |
151 | return match(root, cwd, files, exact=True, badfn=badfn) |
|
151 | return match(root, cwd, files, exact=True, badfn=badfn) | |
152 |
|
152 | |||
153 | def always(root, cwd): |
|
153 | def always(root, cwd): | |
154 | return match(root, cwd, []) |
|
154 | return match(root, cwd, []) | |
155 |
|
155 | |||
156 | def badmatch(match, badfn): |
|
156 | def badmatch(match, badfn): | |
157 | """Make a copy of the given matcher, replacing its bad method with the given |
|
157 | """Make a copy of the given matcher, replacing its bad method with the given | |
158 | one. |
|
158 | one. | |
159 | """ |
|
159 | """ | |
160 | m = copy.copy(match) |
|
160 | m = copy.copy(match) | |
161 | m.bad = badfn |
|
161 | m.bad = badfn | |
162 | return m |
|
162 | return m | |
163 |
|
163 | |||
164 | def _donormalize(patterns, default, root, cwd, auditor, warn): |
|
164 | def _donormalize(patterns, default, root, cwd, auditor, warn): | |
165 | '''Convert 'kind:pat' from the patterns list to tuples with kind and |
|
165 | '''Convert 'kind:pat' from the patterns list to tuples with kind and | |
166 | normalized and rooted patterns and with listfiles expanded.''' |
|
166 | normalized and rooted patterns and with listfiles expanded.''' | |
167 | kindpats = [] |
|
167 | kindpats = [] | |
168 | for kind, pat in [_patsplit(p, default) for p in patterns]: |
|
168 | for kind, pat in [_patsplit(p, default) for p in patterns]: | |
169 | if kind in ('glob', 'relpath'): |
|
169 | if kind in ('glob', 'relpath'): | |
170 | pat = pathutil.canonpath(root, cwd, pat, auditor) |
|
170 | pat = pathutil.canonpath(root, cwd, pat, auditor) | |
171 | elif kind in ('relglob', 'path', 'rootfilesin'): |
|
171 | elif kind in ('relglob', 'path', 'rootfilesin'): | |
172 | pat = util.normpath(pat) |
|
172 | pat = util.normpath(pat) | |
173 | elif kind in ('listfile', 'listfile0'): |
|
173 | elif kind in ('listfile', 'listfile0'): | |
174 | try: |
|
174 | try: | |
175 | files = util.readfile(pat) |
|
175 | files = util.readfile(pat) | |
176 | if kind == 'listfile0': |
|
176 | if kind == 'listfile0': | |
177 | files = files.split('\0') |
|
177 | files = files.split('\0') | |
178 | else: |
|
178 | else: | |
179 | files = files.splitlines() |
|
179 | files = files.splitlines() | |
180 | files = [f for f in files if f] |
|
180 | files = [f for f in files if f] | |
181 | except EnvironmentError: |
|
181 | except EnvironmentError: | |
182 | raise error.Abort(_("unable to read file list (%s)") % pat) |
|
182 | raise error.Abort(_("unable to read file list (%s)") % pat) | |
183 | for k, p, source in _donormalize(files, default, root, cwd, |
|
183 | for k, p, source in _donormalize(files, default, root, cwd, | |
184 | auditor, warn): |
|
184 | auditor, warn): | |
185 | kindpats.append((k, p, pat)) |
|
185 | kindpats.append((k, p, pat)) | |
186 | continue |
|
186 | continue | |
187 | elif kind == 'include': |
|
187 | elif kind == 'include': | |
188 | try: |
|
188 | try: | |
189 | fullpath = os.path.join(root, util.localpath(pat)) |
|
189 | fullpath = os.path.join(root, util.localpath(pat)) | |
190 | includepats = readpatternfile(fullpath, warn) |
|
190 | includepats = readpatternfile(fullpath, warn) | |
191 | for k, p, source in _donormalize(includepats, default, |
|
191 | for k, p, source in _donormalize(includepats, default, | |
192 | root, cwd, auditor, warn): |
|
192 | root, cwd, auditor, warn): | |
193 | kindpats.append((k, p, source or pat)) |
|
193 | kindpats.append((k, p, source or pat)) | |
194 | except error.Abort as inst: |
|
194 | except error.Abort as inst: | |
195 | raise error.Abort('%s: %s' % (pat, inst[0])) |
|
195 | raise error.Abort('%s: %s' % (pat, inst[0])) | |
196 | except IOError as inst: |
|
196 | except IOError as inst: | |
197 | if warn: |
|
197 | if warn: | |
198 | warn(_("skipping unreadable pattern file '%s': %s\n") % |
|
198 | warn(_("skipping unreadable pattern file '%s': %s\n") % | |
199 | (pat, inst.strerror)) |
|
199 | (pat, inst.strerror)) | |
200 | continue |
|
200 | continue | |
201 | # else: re or relre - which cannot be normalized |
|
201 | # else: re or relre - which cannot be normalized | |
202 | kindpats.append((kind, pat, '')) |
|
202 | kindpats.append((kind, pat, '')) | |
203 | return kindpats |
|
203 | return kindpats | |
204 |
|
204 | |||
205 | class basematcher(object): |
|
205 | class basematcher(object): | |
206 |
|
206 | |||
207 | def __init__(self, root, cwd, badfn=None): |
|
207 | def __init__(self, root, cwd, badfn=None): | |
208 | self._root = root |
|
208 | self._root = root | |
209 | self._cwd = cwd |
|
209 | self._cwd = cwd | |
210 | if badfn is not None: |
|
210 | if badfn is not None: | |
211 | self.bad = badfn |
|
211 | self.bad = badfn | |
212 | self.matchfn = lambda f: False |
|
212 | self.matchfn = lambda f: False | |
213 |
|
213 | |||
214 | def __call__(self, fn): |
|
214 | def __call__(self, fn): | |
215 | return self.matchfn(fn) |
|
215 | return self.matchfn(fn) | |
216 | def __iter__(self): |
|
216 | def __iter__(self): | |
217 | for f in self._files: |
|
217 | for f in self._files: | |
218 | yield f |
|
218 | yield f | |
219 | # Callbacks related to how the matcher is used by dirstate.walk. |
|
219 | # Callbacks related to how the matcher is used by dirstate.walk. | |
220 | # Subscribers to these events must monkeypatch the matcher object. |
|
220 | # Subscribers to these events must monkeypatch the matcher object. | |
221 | def bad(self, f, msg): |
|
221 | def bad(self, f, msg): | |
222 | '''Callback from dirstate.walk for each explicit file that can't be |
|
222 | '''Callback from dirstate.walk for each explicit file that can't be | |
223 | found/accessed, with an error message.''' |
|
223 | found/accessed, with an error message.''' | |
224 | pass |
|
224 | pass | |
225 |
|
225 | |||
226 | # If an explicitdir is set, it will be called when an explicitly listed |
|
226 | # If an explicitdir is set, it will be called when an explicitly listed | |
227 | # directory is visited. |
|
227 | # directory is visited. | |
228 | explicitdir = None |
|
228 | explicitdir = None | |
229 |
|
229 | |||
230 | # If an traversedir is set, it will be called when a directory discovered |
|
230 | # If an traversedir is set, it will be called when a directory discovered | |
231 | # by recursive traversal is visited. |
|
231 | # by recursive traversal is visited. | |
232 | traversedir = None |
|
232 | traversedir = None | |
233 |
|
233 | |||
234 | def abs(self, f): |
|
234 | def abs(self, f): | |
235 | '''Convert a repo path back to path that is relative to the root of the |
|
235 | '''Convert a repo path back to path that is relative to the root of the | |
236 | matcher.''' |
|
236 | matcher.''' | |
237 | return f |
|
237 | return f | |
238 |
|
238 | |||
239 | def rel(self, f): |
|
239 | def rel(self, f): | |
240 | '''Convert repo path back to path that is relative to cwd of matcher.''' |
|
240 | '''Convert repo path back to path that is relative to cwd of matcher.''' | |
241 | return util.pathto(self._root, self._cwd, f) |
|
241 | return util.pathto(self._root, self._cwd, f) | |
242 |
|
242 | |||
243 | def uipath(self, f): |
|
243 | def uipath(self, f): | |
244 | '''Convert repo path to a display path. If patterns or -I/-X were used |
|
244 | '''Convert repo path to a display path. If patterns or -I/-X were used | |
245 | to create this matcher, the display path will be relative to cwd. |
|
245 | to create this matcher, the display path will be relative to cwd. | |
246 | Otherwise it is relative to the root of the repo.''' |
|
246 | Otherwise it is relative to the root of the repo.''' | |
247 | return self.rel(f) |
|
247 | return self.rel(f) | |
248 |
|
248 | |||
249 | @propertycache |
|
249 | @propertycache | |
250 | def _files(self): |
|
250 | def _files(self): | |
251 | return [] |
|
251 | return [] | |
252 |
|
252 | |||
253 | def files(self): |
|
253 | def files(self): | |
254 | '''Explicitly listed files or patterns or roots: |
|
254 | '''Explicitly listed files or patterns or roots: | |
255 | if no patterns or .always(): empty list, |
|
255 | if no patterns or .always(): empty list, | |
256 | if exact: list exact files, |
|
256 | if exact: list exact files, | |
257 | if not .anypats(): list all files and dirs, |
|
257 | if not .anypats(): list all files and dirs, | |
258 | else: optimal roots''' |
|
258 | else: optimal roots''' | |
259 | return self._files |
|
259 | return self._files | |
260 |
|
260 | |||
261 | @propertycache |
|
261 | @propertycache | |
262 | def _fileset(self): |
|
262 | def _fileset(self): | |
263 | return set(self._files) |
|
263 | return set(self._files) | |
264 |
|
264 | |||
265 | def exact(self, f): |
|
265 | def exact(self, f): | |
266 | '''Returns True if f is in .files().''' |
|
266 | '''Returns True if f is in .files().''' | |
267 | return f in self._fileset |
|
267 | return f in self._fileset | |
268 |
|
268 | |||
269 | def visitdir(self, dir): |
|
269 | def visitdir(self, dir): | |
270 | '''Decides whether a directory should be visited based on whether it |
|
270 | '''Decides whether a directory should be visited based on whether it | |
271 | has potential matches in it or one of its subdirectories. This is |
|
271 | has potential matches in it or one of its subdirectories. This is | |
272 | based on the match's primary, included, and excluded patterns. |
|
272 | based on the match's primary, included, and excluded patterns. | |
273 |
|
273 | |||
274 | Returns the string 'all' if the given directory and all subdirectories |
|
274 | Returns the string 'all' if the given directory and all subdirectories | |
275 | should be visited. Otherwise returns True or False indicating whether |
|
275 | should be visited. Otherwise returns True or False indicating whether | |
276 | the given directory should be visited. |
|
276 | the given directory should be visited. | |
277 |
|
277 | |||
278 | This function's behavior is undefined if it has returned False for |
|
278 | This function's behavior is undefined if it has returned False for | |
279 | one of the dir's parent directories. |
|
279 | one of the dir's parent directories. | |
280 | ''' |
|
280 | ''' | |
281 | return False |
|
281 | return False | |
282 |
|
282 | |||
283 | def anypats(self): |
|
283 | def anypats(self): | |
284 | '''Matcher uses patterns or include/exclude.''' |
|
284 | '''Matcher uses patterns or include/exclude.''' | |
285 | return False |
|
285 | return False | |
286 |
|
286 | |||
287 | def always(self): |
|
287 | def always(self): | |
288 | '''Matcher will match everything and .files() will be empty |
|
288 | '''Matcher will match everything and .files() will be empty | |
289 | - optimization might be possible and necessary.''' |
|
289 | - optimization might be possible and necessary.''' | |
290 | return False |
|
290 | return False | |
291 |
|
291 | |||
292 | def isexact(self): |
|
292 | def isexact(self): | |
293 | return False |
|
293 | return False | |
294 |
|
294 | |||
295 | def prefix(self): |
|
295 | def prefix(self): | |
296 | return not self.always() and not self.isexact() and not self.anypats() |
|
296 | return not self.always() and not self.isexact() and not self.anypats() | |
297 |
|
297 | |||
298 | class matcher(basematcher): |
|
298 | class matcher(basematcher): | |
299 |
|
299 | |||
300 | def __init__(self, root, cwd, normalize, patterns, include=None, |
|
300 | def __init__(self, root, cwd, normalize, patterns, include=None, | |
301 | exclude=None, default='glob', exact=False, auditor=None, |
|
301 | exclude=None, default='glob', exact=False, auditor=None, | |
302 | ctx=None, listsubrepos=False, warn=None, badfn=None): |
|
302 | ctx=None, listsubrepos=False, warn=None, badfn=None): | |
303 | super(matcher, self).__init__(root, cwd, badfn) |
|
303 | super(matcher, self).__init__(root, cwd, badfn) | |
304 | if include is None: |
|
304 | if include is None: | |
305 | include = [] |
|
305 | include = [] | |
306 | if exclude is None: |
|
306 | if exclude is None: | |
307 | exclude = [] |
|
307 | exclude = [] | |
308 |
|
308 | |||
309 | self._anypats = bool(include or exclude) |
|
309 | self._anypats = bool(include or exclude) | |
|
310 | self._anyincludepats = False | |||
310 | self._always = False |
|
311 | self._always = False | |
311 | self._pathrestricted = bool(include or exclude or patterns) |
|
312 | self._pathrestricted = bool(include or exclude or patterns) | |
312 | self.patternspat = None |
|
313 | self.patternspat = None | |
313 | self.includepat = None |
|
314 | self.includepat = None | |
314 | self.excludepat = None |
|
315 | self.excludepat = None | |
315 |
|
316 | |||
316 | # roots are directories which are recursively included/excluded. |
|
317 | # roots are directories which are recursively included/excluded. | |
317 | self._includeroots = set() |
|
318 | self._includeroots = set() | |
318 | self._excluderoots = set() |
|
319 | self._excluderoots = set() | |
319 | # dirs are directories which are non-recursively included. |
|
320 | # dirs are directories which are non-recursively included. | |
320 | self._includedirs = set() |
|
321 | self._includedirs = set() | |
321 |
|
322 | |||
322 | matchfns = [] |
|
323 | matchfns = [] | |
323 | if include: |
|
324 | if include: | |
324 | kindpats = normalize(include, 'glob', root, cwd, auditor, warn) |
|
325 | kindpats = normalize(include, 'glob', root, cwd, auditor, warn) | |
325 | self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)', |
|
326 | self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)', | |
326 | listsubrepos, root) |
|
327 | listsubrepos, root) | |
|
328 | self._anyincludepats = _anypats(kindpats) | |||
327 | roots, dirs = _rootsanddirs(kindpats) |
|
329 | roots, dirs = _rootsanddirs(kindpats) | |
328 | self._includeroots.update(roots) |
|
330 | self._includeroots.update(roots) | |
329 | self._includedirs.update(dirs) |
|
331 | self._includedirs.update(dirs) | |
330 | matchfns.append(im) |
|
332 | matchfns.append(im) | |
331 | if exclude: |
|
333 | if exclude: | |
332 | kindpats = normalize(exclude, 'glob', root, cwd, auditor, warn) |
|
334 | kindpats = normalize(exclude, 'glob', root, cwd, auditor, warn) | |
333 | self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)', |
|
335 | self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)', | |
334 | listsubrepos, root) |
|
336 | listsubrepos, root) | |
335 | if not _anypats(kindpats): |
|
337 | if not _anypats(kindpats): | |
336 | # Only consider recursive excludes as such - if a non-recursive |
|
338 | # Only consider recursive excludes as such - if a non-recursive | |
337 | # exclude is used, we must still recurse into the excluded |
|
339 | # exclude is used, we must still recurse into the excluded | |
338 | # directory, at least to find subdirectories. In such a case, |
|
340 | # directory, at least to find subdirectories. In such a case, | |
339 | # the regex still won't match the non-recursively-excluded |
|
341 | # the regex still won't match the non-recursively-excluded | |
340 | # files. |
|
342 | # files. | |
341 | self._excluderoots.update(_roots(kindpats)) |
|
343 | self._excluderoots.update(_roots(kindpats)) | |
342 | matchfns.append(lambda f: not em(f)) |
|
344 | matchfns.append(lambda f: not em(f)) | |
343 | if exact: |
|
345 | if exact: | |
344 | if isinstance(patterns, list): |
|
346 | if isinstance(patterns, list): | |
345 | self._files = patterns |
|
347 | self._files = patterns | |
346 | else: |
|
348 | else: | |
347 | self._files = list(patterns) |
|
349 | self._files = list(patterns) | |
348 | matchfns.append(self.exact) |
|
350 | matchfns.append(self.exact) | |
349 | elif patterns: |
|
351 | elif patterns: | |
350 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) |
|
352 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) | |
351 | if not _kindpatsalwaysmatch(kindpats): |
|
353 | if not _kindpatsalwaysmatch(kindpats): | |
352 | self._files = _explicitfiles(kindpats) |
|
354 | self._files = _explicitfiles(kindpats) | |
353 | self._anypats = self._anypats or _anypats(kindpats) |
|
355 | self._anypats = self._anypats or _anypats(kindpats) | |
354 | self.patternspat, pm = _buildmatch(ctx, kindpats, '$', |
|
356 | self.patternspat, pm = _buildmatch(ctx, kindpats, '$', | |
355 | listsubrepos, root) |
|
357 | listsubrepos, root) | |
356 | matchfns.append(pm) |
|
358 | matchfns.append(pm) | |
357 |
|
359 | |||
358 | if not matchfns: |
|
360 | if not matchfns: | |
359 | m = util.always |
|
361 | m = util.always | |
360 | self._always = True |
|
362 | self._always = True | |
361 | elif len(matchfns) == 1: |
|
363 | elif len(matchfns) == 1: | |
362 | m = matchfns[0] |
|
364 | m = matchfns[0] | |
363 | else: |
|
365 | else: | |
364 | def m(f): |
|
366 | def m(f): | |
365 | for matchfn in matchfns: |
|
367 | for matchfn in matchfns: | |
366 | if not matchfn(f): |
|
368 | if not matchfn(f): | |
367 | return False |
|
369 | return False | |
368 | return True |
|
370 | return True | |
369 |
|
371 | |||
370 | self.matchfn = m |
|
372 | self.matchfn = m | |
371 |
|
373 | |||
372 | def uipath(self, f): |
|
374 | def uipath(self, f): | |
373 | return (self._pathrestricted and self.rel(f)) or self.abs(f) |
|
375 | return (self._pathrestricted and self.rel(f)) or self.abs(f) | |
374 |
|
376 | |||
375 | @propertycache |
|
377 | @propertycache | |
376 | def _dirs(self): |
|
378 | def _dirs(self): | |
377 | return set(util.dirs(self._fileset)) | {'.'} |
|
379 | return set(util.dirs(self._fileset)) | {'.'} | |
378 |
|
380 | |||
379 | def visitdir(self, dir): |
|
381 | def visitdir(self, dir): | |
380 | if self.prefix() and dir in self._fileset: |
|
382 | if self.prefix() and dir in self._fileset: | |
381 | return 'all' |
|
383 | return 'all' | |
382 | if dir in self._excluderoots: |
|
384 | if dir in self._excluderoots: | |
383 | return False |
|
385 | return False | |
384 |
if |
|
386 | if self._includeroots or self._includedirs: | |
385 |
|
|
387 | if (not self._anyincludepats and | |
386 |
|
|
388 | not self._excluderoots and | |
387 |
dir |
|
389 | dir in self._includeroots): | |
388 | not any(parent in self._includeroots |
|
390 | # The condition above is essentially self.prefix() for includes | |
389 | for parent in util.finddirs(dir))): |
|
391 | return 'all' | |
390 | return False |
|
392 | if ('.' not in self._includeroots and | |
|
393 | dir not in self._includeroots and | |||
|
394 | dir not in self._includedirs and | |||
|
395 | not any(parent in self._includeroots | |||
|
396 | for parent in util.finddirs(dir))): | |||
|
397 | return False | |||
391 | return (not self._fileset or |
|
398 | return (not self._fileset or | |
392 | '.' in self._fileset or |
|
399 | '.' in self._fileset or | |
393 | dir in self._fileset or |
|
400 | dir in self._fileset or | |
394 | dir in self._dirs or |
|
401 | dir in self._dirs or | |
395 | any(parentdir in self._fileset |
|
402 | any(parentdir in self._fileset | |
396 | for parentdir in util.finddirs(dir))) |
|
403 | for parentdir in util.finddirs(dir))) | |
397 |
|
404 | |||
398 | def anypats(self): |
|
405 | def anypats(self): | |
399 | return self._anypats |
|
406 | return self._anypats | |
400 |
|
407 | |||
401 | def always(self): |
|
408 | def always(self): | |
402 | return self._always |
|
409 | return self._always | |
403 |
|
410 | |||
404 | def isexact(self): |
|
411 | def isexact(self): | |
405 | return self.matchfn == self.exact |
|
412 | return self.matchfn == self.exact | |
406 |
|
413 | |||
407 | def __repr__(self): |
|
414 | def __repr__(self): | |
408 | return ('<matcher files=%r, patterns=%r, includes=%r, excludes=%r>' % |
|
415 | return ('<matcher files=%r, patterns=%r, includes=%r, excludes=%r>' % | |
409 | (self._files, self.patternspat, self.includepat, |
|
416 | (self._files, self.patternspat, self.includepat, | |
410 | self.excludepat)) |
|
417 | self.excludepat)) | |
411 |
|
418 | |||
412 | class subdirmatcher(basematcher): |
|
419 | class subdirmatcher(basematcher): | |
413 | """Adapt a matcher to work on a subdirectory only. |
|
420 | """Adapt a matcher to work on a subdirectory only. | |
414 |
|
421 | |||
415 | The paths are remapped to remove/insert the path as needed: |
|
422 | The paths are remapped to remove/insert the path as needed: | |
416 |
|
423 | |||
417 | >>> m1 = match('root', '', ['a.txt', 'sub/b.txt']) |
|
424 | >>> m1 = match('root', '', ['a.txt', 'sub/b.txt']) | |
418 | >>> m2 = subdirmatcher('sub', m1) |
|
425 | >>> m2 = subdirmatcher('sub', m1) | |
419 | >>> bool(m2('a.txt')) |
|
426 | >>> bool(m2('a.txt')) | |
420 | False |
|
427 | False | |
421 | >>> bool(m2('b.txt')) |
|
428 | >>> bool(m2('b.txt')) | |
422 | True |
|
429 | True | |
423 | >>> bool(m2.matchfn('a.txt')) |
|
430 | >>> bool(m2.matchfn('a.txt')) | |
424 | False |
|
431 | False | |
425 | >>> bool(m2.matchfn('b.txt')) |
|
432 | >>> bool(m2.matchfn('b.txt')) | |
426 | True |
|
433 | True | |
427 | >>> m2.files() |
|
434 | >>> m2.files() | |
428 | ['b.txt'] |
|
435 | ['b.txt'] | |
429 | >>> m2.exact('b.txt') |
|
436 | >>> m2.exact('b.txt') | |
430 | True |
|
437 | True | |
431 | >>> util.pconvert(m2.rel('b.txt')) |
|
438 | >>> util.pconvert(m2.rel('b.txt')) | |
432 | 'sub/b.txt' |
|
439 | 'sub/b.txt' | |
433 | >>> def bad(f, msg): |
|
440 | >>> def bad(f, msg): | |
434 | ... print "%s: %s" % (f, msg) |
|
441 | ... print "%s: %s" % (f, msg) | |
435 | >>> m1.bad = bad |
|
442 | >>> m1.bad = bad | |
436 | >>> m2.bad('x.txt', 'No such file') |
|
443 | >>> m2.bad('x.txt', 'No such file') | |
437 | sub/x.txt: No such file |
|
444 | sub/x.txt: No such file | |
438 | >>> m2.abs('c.txt') |
|
445 | >>> m2.abs('c.txt') | |
439 | 'sub/c.txt' |
|
446 | 'sub/c.txt' | |
440 | """ |
|
447 | """ | |
441 |
|
448 | |||
442 | def __init__(self, path, matcher): |
|
449 | def __init__(self, path, matcher): | |
443 | super(subdirmatcher, self).__init__(matcher._root, matcher._cwd) |
|
450 | super(subdirmatcher, self).__init__(matcher._root, matcher._cwd) | |
444 | self._path = path |
|
451 | self._path = path | |
445 | self._matcher = matcher |
|
452 | self._matcher = matcher | |
446 | self._always = matcher.always() |
|
453 | self._always = matcher.always() | |
447 |
|
454 | |||
448 | self._files = [f[len(path) + 1:] for f in matcher._files |
|
455 | self._files = [f[len(path) + 1:] for f in matcher._files | |
449 | if f.startswith(path + "/")] |
|
456 | if f.startswith(path + "/")] | |
450 |
|
457 | |||
451 | # If the parent repo had a path to this subrepo and the matcher is |
|
458 | # If the parent repo had a path to this subrepo and the matcher is | |
452 | # a prefix matcher, this submatcher always matches. |
|
459 | # a prefix matcher, this submatcher always matches. | |
453 | if matcher.prefix(): |
|
460 | if matcher.prefix(): | |
454 | self._always = any(f == path for f in matcher._files) |
|
461 | self._always = any(f == path for f in matcher._files) | |
455 |
|
462 | |||
456 | # Some information is lost in the superclass's constructor, so we |
|
463 | # Some information is lost in the superclass's constructor, so we | |
457 | # can not accurately create the matching function for the subdirectory |
|
464 | # can not accurately create the matching function for the subdirectory | |
458 | # from the inputs. Instead, we override matchfn() and visitdir() to |
|
465 | # from the inputs. Instead, we override matchfn() and visitdir() to | |
459 | # call the original matcher with the subdirectory path prepended. |
|
466 | # call the original matcher with the subdirectory path prepended. | |
460 | self.matchfn = lambda fn: matcher.matchfn(self._path + "/" + fn) |
|
467 | self.matchfn = lambda fn: matcher.matchfn(self._path + "/" + fn) | |
461 |
|
468 | |||
462 | def bad(self, f, msg): |
|
469 | def bad(self, f, msg): | |
463 | self._matcher.bad(self._path + "/" + f, msg) |
|
470 | self._matcher.bad(self._path + "/" + f, msg) | |
464 |
|
471 | |||
465 | def abs(self, f): |
|
472 | def abs(self, f): | |
466 | return self._matcher.abs(self._path + "/" + f) |
|
473 | return self._matcher.abs(self._path + "/" + f) | |
467 |
|
474 | |||
468 | def rel(self, f): |
|
475 | def rel(self, f): | |
469 | return self._matcher.rel(self._path + "/" + f) |
|
476 | return self._matcher.rel(self._path + "/" + f) | |
470 |
|
477 | |||
471 | def uipath(self, f): |
|
478 | def uipath(self, f): | |
472 | return self._matcher.uipath(self._path + "/" + f) |
|
479 | return self._matcher.uipath(self._path + "/" + f) | |
473 |
|
480 | |||
474 | def visitdir(self, dir): |
|
481 | def visitdir(self, dir): | |
475 | if dir == '.': |
|
482 | if dir == '.': | |
476 | dir = self._path |
|
483 | dir = self._path | |
477 | else: |
|
484 | else: | |
478 | dir = self._path + "/" + dir |
|
485 | dir = self._path + "/" + dir | |
479 | return self._matcher.visitdir(dir) |
|
486 | return self._matcher.visitdir(dir) | |
480 |
|
487 | |||
481 | def always(self): |
|
488 | def always(self): | |
482 | return self._always |
|
489 | return self._always | |
483 |
|
490 | |||
484 | def anypats(self): |
|
491 | def anypats(self): | |
485 | return self._matcher.anypats() |
|
492 | return self._matcher.anypats() | |
486 |
|
493 | |||
487 | def patkind(pattern, default=None): |
|
494 | def patkind(pattern, default=None): | |
488 | '''If pattern is 'kind:pat' with a known kind, return kind.''' |
|
495 | '''If pattern is 'kind:pat' with a known kind, return kind.''' | |
489 | return _patsplit(pattern, default)[0] |
|
496 | return _patsplit(pattern, default)[0] | |
490 |
|
497 | |||
491 | def _patsplit(pattern, default): |
|
498 | def _patsplit(pattern, default): | |
492 | """Split a string into the optional pattern kind prefix and the actual |
|
499 | """Split a string into the optional pattern kind prefix and the actual | |
493 | pattern.""" |
|
500 | pattern.""" | |
494 | if ':' in pattern: |
|
501 | if ':' in pattern: | |
495 | kind, pat = pattern.split(':', 1) |
|
502 | kind, pat = pattern.split(':', 1) | |
496 | if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre', |
|
503 | if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre', | |
497 | 'listfile', 'listfile0', 'set', 'include', 'subinclude', |
|
504 | 'listfile', 'listfile0', 'set', 'include', 'subinclude', | |
498 | 'rootfilesin'): |
|
505 | 'rootfilesin'): | |
499 | return kind, pat |
|
506 | return kind, pat | |
500 | return default, pattern |
|
507 | return default, pattern | |
501 |
|
508 | |||
502 | def _globre(pat): |
|
509 | def _globre(pat): | |
503 | r'''Convert an extended glob string to a regexp string. |
|
510 | r'''Convert an extended glob string to a regexp string. | |
504 |
|
511 | |||
505 | >>> print _globre(r'?') |
|
512 | >>> print _globre(r'?') | |
506 | . |
|
513 | . | |
507 | >>> print _globre(r'*') |
|
514 | >>> print _globre(r'*') | |
508 | [^/]* |
|
515 | [^/]* | |
509 | >>> print _globre(r'**') |
|
516 | >>> print _globre(r'**') | |
510 | .* |
|
517 | .* | |
511 | >>> print _globre(r'**/a') |
|
518 | >>> print _globre(r'**/a') | |
512 | (?:.*/)?a |
|
519 | (?:.*/)?a | |
513 | >>> print _globre(r'a/**/b') |
|
520 | >>> print _globre(r'a/**/b') | |
514 | a\/(?:.*/)?b |
|
521 | a\/(?:.*/)?b | |
515 | >>> print _globre(r'[a*?!^][^b][!c]') |
|
522 | >>> print _globre(r'[a*?!^][^b][!c]') | |
516 | [a*?!^][\^b][^c] |
|
523 | [a*?!^][\^b][^c] | |
517 | >>> print _globre(r'{a,b}') |
|
524 | >>> print _globre(r'{a,b}') | |
518 | (?:a|b) |
|
525 | (?:a|b) | |
519 | >>> print _globre(r'.\*\?') |
|
526 | >>> print _globre(r'.\*\?') | |
520 | \.\*\? |
|
527 | \.\*\? | |
521 | ''' |
|
528 | ''' | |
522 | i, n = 0, len(pat) |
|
529 | i, n = 0, len(pat) | |
523 | res = '' |
|
530 | res = '' | |
524 | group = 0 |
|
531 | group = 0 | |
525 | escape = util.re.escape |
|
532 | escape = util.re.escape | |
526 | def peek(): |
|
533 | def peek(): | |
527 | return i < n and pat[i:i + 1] |
|
534 | return i < n and pat[i:i + 1] | |
528 | while i < n: |
|
535 | while i < n: | |
529 | c = pat[i:i + 1] |
|
536 | c = pat[i:i + 1] | |
530 | i += 1 |
|
537 | i += 1 | |
531 | if c not in '*?[{},\\': |
|
538 | if c not in '*?[{},\\': | |
532 | res += escape(c) |
|
539 | res += escape(c) | |
533 | elif c == '*': |
|
540 | elif c == '*': | |
534 | if peek() == '*': |
|
541 | if peek() == '*': | |
535 | i += 1 |
|
542 | i += 1 | |
536 | if peek() == '/': |
|
543 | if peek() == '/': | |
537 | i += 1 |
|
544 | i += 1 | |
538 | res += '(?:.*/)?' |
|
545 | res += '(?:.*/)?' | |
539 | else: |
|
546 | else: | |
540 | res += '.*' |
|
547 | res += '.*' | |
541 | else: |
|
548 | else: | |
542 | res += '[^/]*' |
|
549 | res += '[^/]*' | |
543 | elif c == '?': |
|
550 | elif c == '?': | |
544 | res += '.' |
|
551 | res += '.' | |
545 | elif c == '[': |
|
552 | elif c == '[': | |
546 | j = i |
|
553 | j = i | |
547 | if j < n and pat[j:j + 1] in '!]': |
|
554 | if j < n and pat[j:j + 1] in '!]': | |
548 | j += 1 |
|
555 | j += 1 | |
549 | while j < n and pat[j:j + 1] != ']': |
|
556 | while j < n and pat[j:j + 1] != ']': | |
550 | j += 1 |
|
557 | j += 1 | |
551 | if j >= n: |
|
558 | if j >= n: | |
552 | res += '\\[' |
|
559 | res += '\\[' | |
553 | else: |
|
560 | else: | |
554 | stuff = pat[i:j].replace('\\','\\\\') |
|
561 | stuff = pat[i:j].replace('\\','\\\\') | |
555 | i = j + 1 |
|
562 | i = j + 1 | |
556 | if stuff[0:1] == '!': |
|
563 | if stuff[0:1] == '!': | |
557 | stuff = '^' + stuff[1:] |
|
564 | stuff = '^' + stuff[1:] | |
558 | elif stuff[0:1] == '^': |
|
565 | elif stuff[0:1] == '^': | |
559 | stuff = '\\' + stuff |
|
566 | stuff = '\\' + stuff | |
560 | res = '%s[%s]' % (res, stuff) |
|
567 | res = '%s[%s]' % (res, stuff) | |
561 | elif c == '{': |
|
568 | elif c == '{': | |
562 | group += 1 |
|
569 | group += 1 | |
563 | res += '(?:' |
|
570 | res += '(?:' | |
564 | elif c == '}' and group: |
|
571 | elif c == '}' and group: | |
565 | res += ')' |
|
572 | res += ')' | |
566 | group -= 1 |
|
573 | group -= 1 | |
567 | elif c == ',' and group: |
|
574 | elif c == ',' and group: | |
568 | res += '|' |
|
575 | res += '|' | |
569 | elif c == '\\': |
|
576 | elif c == '\\': | |
570 | p = peek() |
|
577 | p = peek() | |
571 | if p: |
|
578 | if p: | |
572 | i += 1 |
|
579 | i += 1 | |
573 | res += escape(p) |
|
580 | res += escape(p) | |
574 | else: |
|
581 | else: | |
575 | res += escape(c) |
|
582 | res += escape(c) | |
576 | else: |
|
583 | else: | |
577 | res += escape(c) |
|
584 | res += escape(c) | |
578 | return res |
|
585 | return res | |
579 |
|
586 | |||
580 | def _regex(kind, pat, globsuffix): |
|
587 | def _regex(kind, pat, globsuffix): | |
581 | '''Convert a (normalized) pattern of any kind into a regular expression. |
|
588 | '''Convert a (normalized) pattern of any kind into a regular expression. | |
582 | globsuffix is appended to the regexp of globs.''' |
|
589 | globsuffix is appended to the regexp of globs.''' | |
583 | if not pat: |
|
590 | if not pat: | |
584 | return '' |
|
591 | return '' | |
585 | if kind == 're': |
|
592 | if kind == 're': | |
586 | return pat |
|
593 | return pat | |
587 | if kind == 'path': |
|
594 | if kind == 'path': | |
588 | if pat == '.': |
|
595 | if pat == '.': | |
589 | return '' |
|
596 | return '' | |
590 | return '^' + util.re.escape(pat) + '(?:/|$)' |
|
597 | return '^' + util.re.escape(pat) + '(?:/|$)' | |
591 | if kind == 'rootfilesin': |
|
598 | if kind == 'rootfilesin': | |
592 | if pat == '.': |
|
599 | if pat == '.': | |
593 | escaped = '' |
|
600 | escaped = '' | |
594 | else: |
|
601 | else: | |
595 | # Pattern is a directory name. |
|
602 | # Pattern is a directory name. | |
596 | escaped = util.re.escape(pat) + '/' |
|
603 | escaped = util.re.escape(pat) + '/' | |
597 | # Anything after the pattern must be a non-directory. |
|
604 | # Anything after the pattern must be a non-directory. | |
598 | return '^' + escaped + '[^/]+$' |
|
605 | return '^' + escaped + '[^/]+$' | |
599 | if kind == 'relglob': |
|
606 | if kind == 'relglob': | |
600 | return '(?:|.*/)' + _globre(pat) + globsuffix |
|
607 | return '(?:|.*/)' + _globre(pat) + globsuffix | |
601 | if kind == 'relpath': |
|
608 | if kind == 'relpath': | |
602 | return util.re.escape(pat) + '(?:/|$)' |
|
609 | return util.re.escape(pat) + '(?:/|$)' | |
603 | if kind == 'relre': |
|
610 | if kind == 'relre': | |
604 | if pat.startswith('^'): |
|
611 | if pat.startswith('^'): | |
605 | return pat |
|
612 | return pat | |
606 | return '.*' + pat |
|
613 | return '.*' + pat | |
607 | return _globre(pat) + globsuffix |
|
614 | return _globre(pat) + globsuffix | |
608 |
|
615 | |||
609 | def _buildmatch(ctx, kindpats, globsuffix, listsubrepos, root): |
|
616 | def _buildmatch(ctx, kindpats, globsuffix, listsubrepos, root): | |
610 | '''Return regexp string and a matcher function for kindpats. |
|
617 | '''Return regexp string and a matcher function for kindpats. | |
611 | globsuffix is appended to the regexp of globs.''' |
|
618 | globsuffix is appended to the regexp of globs.''' | |
612 | matchfuncs = [] |
|
619 | matchfuncs = [] | |
613 |
|
620 | |||
614 | subincludes, kindpats = _expandsubinclude(kindpats, root) |
|
621 | subincludes, kindpats = _expandsubinclude(kindpats, root) | |
615 | if subincludes: |
|
622 | if subincludes: | |
616 | submatchers = {} |
|
623 | submatchers = {} | |
617 | def matchsubinclude(f): |
|
624 | def matchsubinclude(f): | |
618 | for prefix, matcherargs in subincludes: |
|
625 | for prefix, matcherargs in subincludes: | |
619 | if f.startswith(prefix): |
|
626 | if f.startswith(prefix): | |
620 | mf = submatchers.get(prefix) |
|
627 | mf = submatchers.get(prefix) | |
621 | if mf is None: |
|
628 | if mf is None: | |
622 | mf = match(*matcherargs) |
|
629 | mf = match(*matcherargs) | |
623 | submatchers[prefix] = mf |
|
630 | submatchers[prefix] = mf | |
624 |
|
631 | |||
625 | if mf(f[len(prefix):]): |
|
632 | if mf(f[len(prefix):]): | |
626 | return True |
|
633 | return True | |
627 | return False |
|
634 | return False | |
628 | matchfuncs.append(matchsubinclude) |
|
635 | matchfuncs.append(matchsubinclude) | |
629 |
|
636 | |||
630 | fset, kindpats = _expandsets(kindpats, ctx, listsubrepos) |
|
637 | fset, kindpats = _expandsets(kindpats, ctx, listsubrepos) | |
631 | if fset: |
|
638 | if fset: | |
632 | matchfuncs.append(fset.__contains__) |
|
639 | matchfuncs.append(fset.__contains__) | |
633 |
|
640 | |||
634 | regex = '' |
|
641 | regex = '' | |
635 | if kindpats: |
|
642 | if kindpats: | |
636 | regex, mf = _buildregexmatch(kindpats, globsuffix) |
|
643 | regex, mf = _buildregexmatch(kindpats, globsuffix) | |
637 | matchfuncs.append(mf) |
|
644 | matchfuncs.append(mf) | |
638 |
|
645 | |||
639 | if len(matchfuncs) == 1: |
|
646 | if len(matchfuncs) == 1: | |
640 | return regex, matchfuncs[0] |
|
647 | return regex, matchfuncs[0] | |
641 | else: |
|
648 | else: | |
642 | return regex, lambda f: any(mf(f) for mf in matchfuncs) |
|
649 | return regex, lambda f: any(mf(f) for mf in matchfuncs) | |
643 |
|
650 | |||
644 | def _buildregexmatch(kindpats, globsuffix): |
|
651 | def _buildregexmatch(kindpats, globsuffix): | |
645 | """Build a match function from a list of kinds and kindpats, |
|
652 | """Build a match function from a list of kinds and kindpats, | |
646 | return regexp string and a matcher function.""" |
|
653 | return regexp string and a matcher function.""" | |
647 | try: |
|
654 | try: | |
648 | regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix) |
|
655 | regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix) | |
649 | for (k, p, s) in kindpats]) |
|
656 | for (k, p, s) in kindpats]) | |
650 | if len(regex) > 20000: |
|
657 | if len(regex) > 20000: | |
651 | raise OverflowError |
|
658 | raise OverflowError | |
652 | return regex, _rematcher(regex) |
|
659 | return regex, _rematcher(regex) | |
653 | except OverflowError: |
|
660 | except OverflowError: | |
654 | # We're using a Python with a tiny regex engine and we |
|
661 | # We're using a Python with a tiny regex engine and we | |
655 | # made it explode, so we'll divide the pattern list in two |
|
662 | # made it explode, so we'll divide the pattern list in two | |
656 | # until it works |
|
663 | # until it works | |
657 | l = len(kindpats) |
|
664 | l = len(kindpats) | |
658 | if l < 2: |
|
665 | if l < 2: | |
659 | raise |
|
666 | raise | |
660 | regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix) |
|
667 | regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix) | |
661 | regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix) |
|
668 | regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix) | |
662 | return regex, lambda s: a(s) or b(s) |
|
669 | return regex, lambda s: a(s) or b(s) | |
663 | except re.error: |
|
670 | except re.error: | |
664 | for k, p, s in kindpats: |
|
671 | for k, p, s in kindpats: | |
665 | try: |
|
672 | try: | |
666 | _rematcher('(?:%s)' % _regex(k, p, globsuffix)) |
|
673 | _rematcher('(?:%s)' % _regex(k, p, globsuffix)) | |
667 | except re.error: |
|
674 | except re.error: | |
668 | if s: |
|
675 | if s: | |
669 | raise error.Abort(_("%s: invalid pattern (%s): %s") % |
|
676 | raise error.Abort(_("%s: invalid pattern (%s): %s") % | |
670 | (s, k, p)) |
|
677 | (s, k, p)) | |
671 | else: |
|
678 | else: | |
672 | raise error.Abort(_("invalid pattern (%s): %s") % (k, p)) |
|
679 | raise error.Abort(_("invalid pattern (%s): %s") % (k, p)) | |
673 | raise error.Abort(_("invalid pattern")) |
|
680 | raise error.Abort(_("invalid pattern")) | |
674 |
|
681 | |||
675 | def _patternrootsanddirs(kindpats): |
|
682 | def _patternrootsanddirs(kindpats): | |
676 | '''Returns roots and directories corresponding to each pattern. |
|
683 | '''Returns roots and directories corresponding to each pattern. | |
677 |
|
684 | |||
678 | This calculates the roots and directories exactly matching the patterns and |
|
685 | This calculates the roots and directories exactly matching the patterns and | |
679 | returns a tuple of (roots, dirs) for each. It does not return other |
|
686 | returns a tuple of (roots, dirs) for each. It does not return other | |
680 | directories which may also need to be considered, like the parent |
|
687 | directories which may also need to be considered, like the parent | |
681 | directories. |
|
688 | directories. | |
682 | ''' |
|
689 | ''' | |
683 | r = [] |
|
690 | r = [] | |
684 | d = [] |
|
691 | d = [] | |
685 | for kind, pat, source in kindpats: |
|
692 | for kind, pat, source in kindpats: | |
686 | if kind == 'glob': # find the non-glob prefix |
|
693 | if kind == 'glob': # find the non-glob prefix | |
687 | root = [] |
|
694 | root = [] | |
688 | for p in pat.split('/'): |
|
695 | for p in pat.split('/'): | |
689 | if '[' in p or '{' in p or '*' in p or '?' in p: |
|
696 | if '[' in p or '{' in p or '*' in p or '?' in p: | |
690 | break |
|
697 | break | |
691 | root.append(p) |
|
698 | root.append(p) | |
692 | r.append('/'.join(root) or '.') |
|
699 | r.append('/'.join(root) or '.') | |
693 | elif kind in ('relpath', 'path'): |
|
700 | elif kind in ('relpath', 'path'): | |
694 | r.append(pat or '.') |
|
701 | r.append(pat or '.') | |
695 | elif kind in ('rootfilesin',): |
|
702 | elif kind in ('rootfilesin',): | |
696 | d.append(pat or '.') |
|
703 | d.append(pat or '.') | |
697 | else: # relglob, re, relre |
|
704 | else: # relglob, re, relre | |
698 | r.append('.') |
|
705 | r.append('.') | |
699 | return r, d |
|
706 | return r, d | |
700 |
|
707 | |||
701 | def _roots(kindpats): |
|
708 | def _roots(kindpats): | |
702 | '''Returns root directories to match recursively from the given patterns.''' |
|
709 | '''Returns root directories to match recursively from the given patterns.''' | |
703 | roots, dirs = _patternrootsanddirs(kindpats) |
|
710 | roots, dirs = _patternrootsanddirs(kindpats) | |
704 | return roots |
|
711 | return roots | |
705 |
|
712 | |||
706 | def _rootsanddirs(kindpats): |
|
713 | def _rootsanddirs(kindpats): | |
707 | '''Returns roots and exact directories from patterns. |
|
714 | '''Returns roots and exact directories from patterns. | |
708 |
|
715 | |||
709 | roots are directories to match recursively, whereas exact directories should |
|
716 | roots are directories to match recursively, whereas exact directories should | |
710 | be matched non-recursively. The returned (roots, dirs) tuple will also |
|
717 | be matched non-recursively. The returned (roots, dirs) tuple will also | |
711 | include directories that need to be implicitly considered as either, such as |
|
718 | include directories that need to be implicitly considered as either, such as | |
712 | parent directories. |
|
719 | parent directories. | |
713 |
|
720 | |||
714 | >>> _rootsanddirs(\ |
|
721 | >>> _rootsanddirs(\ | |
715 | [('glob', 'g/h/*', ''), ('glob', 'g/h', ''), ('glob', 'g*', '')]) |
|
722 | [('glob', 'g/h/*', ''), ('glob', 'g/h', ''), ('glob', 'g*', '')]) | |
716 | (['g/h', 'g/h', '.'], ['g', '.']) |
|
723 | (['g/h', 'g/h', '.'], ['g', '.']) | |
717 | >>> _rootsanddirs(\ |
|
724 | >>> _rootsanddirs(\ | |
718 | [('rootfilesin', 'g/h', ''), ('rootfilesin', '', '')]) |
|
725 | [('rootfilesin', 'g/h', ''), ('rootfilesin', '', '')]) | |
719 | ([], ['g/h', '.', 'g', '.']) |
|
726 | ([], ['g/h', '.', 'g', '.']) | |
720 | >>> _rootsanddirs(\ |
|
727 | >>> _rootsanddirs(\ | |
721 | [('relpath', 'r', ''), ('path', 'p/p', ''), ('path', '', '')]) |
|
728 | [('relpath', 'r', ''), ('path', 'p/p', ''), ('path', '', '')]) | |
722 | (['r', 'p/p', '.'], ['p', '.']) |
|
729 | (['r', 'p/p', '.'], ['p', '.']) | |
723 | >>> _rootsanddirs(\ |
|
730 | >>> _rootsanddirs(\ | |
724 | [('relglob', 'rg*', ''), ('re', 're/', ''), ('relre', 'rr', '')]) |
|
731 | [('relglob', 'rg*', ''), ('re', 're/', ''), ('relre', 'rr', '')]) | |
725 | (['.', '.', '.'], ['.']) |
|
732 | (['.', '.', '.'], ['.']) | |
726 | ''' |
|
733 | ''' | |
727 | r, d = _patternrootsanddirs(kindpats) |
|
734 | r, d = _patternrootsanddirs(kindpats) | |
728 |
|
735 | |||
729 | # Append the parents as non-recursive/exact directories, since they must be |
|
736 | # Append the parents as non-recursive/exact directories, since they must be | |
730 | # scanned to get to either the roots or the other exact directories. |
|
737 | # scanned to get to either the roots or the other exact directories. | |
731 | d.extend(util.dirs(d)) |
|
738 | d.extend(util.dirs(d)) | |
732 | d.extend(util.dirs(r)) |
|
739 | d.extend(util.dirs(r)) | |
733 | # util.dirs() does not include the root directory, so add it manually |
|
740 | # util.dirs() does not include the root directory, so add it manually | |
734 | d.append('.') |
|
741 | d.append('.') | |
735 |
|
742 | |||
736 | return r, d |
|
743 | return r, d | |
737 |
|
744 | |||
738 | def _explicitfiles(kindpats): |
|
745 | def _explicitfiles(kindpats): | |
739 | '''Returns the potential explicit filenames from the patterns. |
|
746 | '''Returns the potential explicit filenames from the patterns. | |
740 |
|
747 | |||
741 | >>> _explicitfiles([('path', 'foo/bar', '')]) |
|
748 | >>> _explicitfiles([('path', 'foo/bar', '')]) | |
742 | ['foo/bar'] |
|
749 | ['foo/bar'] | |
743 | >>> _explicitfiles([('rootfilesin', 'foo/bar', '')]) |
|
750 | >>> _explicitfiles([('rootfilesin', 'foo/bar', '')]) | |
744 | [] |
|
751 | [] | |
745 | ''' |
|
752 | ''' | |
746 | # Keep only the pattern kinds where one can specify filenames (vs only |
|
753 | # Keep only the pattern kinds where one can specify filenames (vs only | |
747 | # directory names). |
|
754 | # directory names). | |
748 | filable = [kp for kp in kindpats if kp[0] not in ('rootfilesin',)] |
|
755 | filable = [kp for kp in kindpats if kp[0] not in ('rootfilesin',)] | |
749 | return _roots(filable) |
|
756 | return _roots(filable) | |
750 |
|
757 | |||
751 | def _anypats(kindpats): |
|
758 | def _anypats(kindpats): | |
752 | for kind, pat, source in kindpats: |
|
759 | for kind, pat, source in kindpats: | |
753 | if kind in ('glob', 're', 'relglob', 'relre', 'set', 'rootfilesin'): |
|
760 | if kind in ('glob', 're', 'relglob', 'relre', 'set', 'rootfilesin'): | |
754 | return True |
|
761 | return True | |
755 |
|
762 | |||
756 | _commentre = None |
|
763 | _commentre = None | |
757 |
|
764 | |||
758 | def readpatternfile(filepath, warn, sourceinfo=False): |
|
765 | def readpatternfile(filepath, warn, sourceinfo=False): | |
759 | '''parse a pattern file, returning a list of |
|
766 | '''parse a pattern file, returning a list of | |
760 | patterns. These patterns should be given to compile() |
|
767 | patterns. These patterns should be given to compile() | |
761 | to be validated and converted into a match function. |
|
768 | to be validated and converted into a match function. | |
762 |
|
769 | |||
763 | trailing white space is dropped. |
|
770 | trailing white space is dropped. | |
764 | the escape character is backslash. |
|
771 | the escape character is backslash. | |
765 | comments start with #. |
|
772 | comments start with #. | |
766 | empty lines are skipped. |
|
773 | empty lines are skipped. | |
767 |
|
774 | |||
768 | lines can be of the following formats: |
|
775 | lines can be of the following formats: | |
769 |
|
776 | |||
770 | syntax: regexp # defaults following lines to non-rooted regexps |
|
777 | syntax: regexp # defaults following lines to non-rooted regexps | |
771 | syntax: glob # defaults following lines to non-rooted globs |
|
778 | syntax: glob # defaults following lines to non-rooted globs | |
772 | re:pattern # non-rooted regular expression |
|
779 | re:pattern # non-rooted regular expression | |
773 | glob:pattern # non-rooted glob |
|
780 | glob:pattern # non-rooted glob | |
774 | pattern # pattern of the current default type |
|
781 | pattern # pattern of the current default type | |
775 |
|
782 | |||
776 | if sourceinfo is set, returns a list of tuples: |
|
783 | if sourceinfo is set, returns a list of tuples: | |
777 | (pattern, lineno, originalline). This is useful to debug ignore patterns. |
|
784 | (pattern, lineno, originalline). This is useful to debug ignore patterns. | |
778 | ''' |
|
785 | ''' | |
779 |
|
786 | |||
780 | syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:', |
|
787 | syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:', | |
781 | 'include': 'include', 'subinclude': 'subinclude'} |
|
788 | 'include': 'include', 'subinclude': 'subinclude'} | |
782 | syntax = 'relre:' |
|
789 | syntax = 'relre:' | |
783 | patterns = [] |
|
790 | patterns = [] | |
784 |
|
791 | |||
785 | fp = open(filepath, 'rb') |
|
792 | fp = open(filepath, 'rb') | |
786 | for lineno, line in enumerate(util.iterfile(fp), start=1): |
|
793 | for lineno, line in enumerate(util.iterfile(fp), start=1): | |
787 | if "#" in line: |
|
794 | if "#" in line: | |
788 | global _commentre |
|
795 | global _commentre | |
789 | if not _commentre: |
|
796 | if not _commentre: | |
790 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') |
|
797 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') | |
791 | # remove comments prefixed by an even number of escapes |
|
798 | # remove comments prefixed by an even number of escapes | |
792 | m = _commentre.search(line) |
|
799 | m = _commentre.search(line) | |
793 | if m: |
|
800 | if m: | |
794 | line = line[:m.end(1)] |
|
801 | line = line[:m.end(1)] | |
795 | # fixup properly escaped comments that survived the above |
|
802 | # fixup properly escaped comments that survived the above | |
796 | line = line.replace("\\#", "#") |
|
803 | line = line.replace("\\#", "#") | |
797 | line = line.rstrip() |
|
804 | line = line.rstrip() | |
798 | if not line: |
|
805 | if not line: | |
799 | continue |
|
806 | continue | |
800 |
|
807 | |||
801 | if line.startswith('syntax:'): |
|
808 | if line.startswith('syntax:'): | |
802 | s = line[7:].strip() |
|
809 | s = line[7:].strip() | |
803 | try: |
|
810 | try: | |
804 | syntax = syntaxes[s] |
|
811 | syntax = syntaxes[s] | |
805 | except KeyError: |
|
812 | except KeyError: | |
806 | if warn: |
|
813 | if warn: | |
807 | warn(_("%s: ignoring invalid syntax '%s'\n") % |
|
814 | warn(_("%s: ignoring invalid syntax '%s'\n") % | |
808 | (filepath, s)) |
|
815 | (filepath, s)) | |
809 | continue |
|
816 | continue | |
810 |
|
817 | |||
811 | linesyntax = syntax |
|
818 | linesyntax = syntax | |
812 | for s, rels in syntaxes.iteritems(): |
|
819 | for s, rels in syntaxes.iteritems(): | |
813 | if line.startswith(rels): |
|
820 | if line.startswith(rels): | |
814 | linesyntax = rels |
|
821 | linesyntax = rels | |
815 | line = line[len(rels):] |
|
822 | line = line[len(rels):] | |
816 | break |
|
823 | break | |
817 | elif line.startswith(s+':'): |
|
824 | elif line.startswith(s+':'): | |
818 | linesyntax = rels |
|
825 | linesyntax = rels | |
819 | line = line[len(s) + 1:] |
|
826 | line = line[len(s) + 1:] | |
820 | break |
|
827 | break | |
821 | if sourceinfo: |
|
828 | if sourceinfo: | |
822 | patterns.append((linesyntax + line, lineno, line)) |
|
829 | patterns.append((linesyntax + line, lineno, line)) | |
823 | else: |
|
830 | else: | |
824 | patterns.append(linesyntax + line) |
|
831 | patterns.append(linesyntax + line) | |
825 | fp.close() |
|
832 | fp.close() | |
826 | return patterns |
|
833 | return patterns |
General Comments 0
You need to be logged in to leave comments.
Login now