Show More
@@ -1,1625 +1,1614 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, print_function |
|
8 | from __future__ import absolute_import, print_function | |
9 |
|
9 | |||
10 | import copy |
|
10 | import copy | |
11 | import itertools |
|
11 | import itertools | |
12 | import os |
|
12 | import os | |
13 | import re |
|
13 | import re | |
14 |
|
14 | |||
15 | from .i18n import _ |
|
15 | from .i18n import _ | |
16 | from .pycompat import open |
|
16 | from .pycompat import open | |
17 | from . import ( |
|
17 | from . import ( | |
18 | encoding, |
|
18 | encoding, | |
19 | error, |
|
19 | error, | |
20 | pathutil, |
|
20 | pathutil, | |
21 | policy, |
|
21 | policy, | |
22 | pycompat, |
|
22 | pycompat, | |
23 | util, |
|
23 | util, | |
24 | ) |
|
24 | ) | |
25 | from .utils import stringutil |
|
25 | from .utils import stringutil | |
26 |
|
26 | |||
27 | rustmod = policy.importrust('filepatterns') |
|
27 | rustmod = policy.importrust('filepatterns') | |
28 |
|
28 | |||
29 | allpatternkinds = ( |
|
29 | allpatternkinds = ( | |
30 | b're', |
|
30 | b're', | |
31 | b'glob', |
|
31 | b'glob', | |
32 | b'path', |
|
32 | b'path', | |
33 | b'relglob', |
|
33 | b'relglob', | |
34 | b'relpath', |
|
34 | b'relpath', | |
35 | b'relre', |
|
35 | b'relre', | |
36 | b'rootglob', |
|
36 | b'rootglob', | |
37 | b'listfile', |
|
37 | b'listfile', | |
38 | b'listfile0', |
|
38 | b'listfile0', | |
39 | b'set', |
|
39 | b'set', | |
40 | b'include', |
|
40 | b'include', | |
41 | b'subinclude', |
|
41 | b'subinclude', | |
42 | b'rootfilesin', |
|
42 | b'rootfilesin', | |
43 | ) |
|
43 | ) | |
44 | cwdrelativepatternkinds = (b'relpath', b'glob') |
|
44 | cwdrelativepatternkinds = (b'relpath', b'glob') | |
45 |
|
45 | |||
46 | propertycache = util.propertycache |
|
46 | propertycache = util.propertycache | |
47 |
|
47 | |||
48 |
|
48 | |||
49 | def _rematcher(regex): |
|
49 | def _rematcher(regex): | |
50 | '''compile the regexp with the best available regexp engine and return a |
|
50 | '''compile the regexp with the best available regexp engine and return a | |
51 | matcher function''' |
|
51 | matcher function''' | |
52 | m = util.re.compile(regex) |
|
52 | m = util.re.compile(regex) | |
53 | try: |
|
53 | try: | |
54 | # slightly faster, provided by facebook's re2 bindings |
|
54 | # slightly faster, provided by facebook's re2 bindings | |
55 | return m.test_match |
|
55 | return m.test_match | |
56 | except AttributeError: |
|
56 | except AttributeError: | |
57 | return m.match |
|
57 | return m.match | |
58 |
|
58 | |||
59 |
|
59 | |||
60 | def _expandsets(kindpats, ctx=None, listsubrepos=False, badfn=None): |
|
60 | def _expandsets(kindpats, ctx=None, listsubrepos=False, badfn=None): | |
61 | '''Returns the kindpats list with the 'set' patterns expanded to matchers''' |
|
61 | '''Returns the kindpats list with the 'set' patterns expanded to matchers''' | |
62 | matchers = [] |
|
62 | matchers = [] | |
63 | other = [] |
|
63 | other = [] | |
64 |
|
64 | |||
65 | for kind, pat, source in kindpats: |
|
65 | for kind, pat, source in kindpats: | |
66 | if kind == b'set': |
|
66 | if kind == b'set': | |
67 | if ctx is None: |
|
67 | if ctx is None: | |
68 | raise error.ProgrammingError( |
|
68 | raise error.ProgrammingError( | |
69 | b"fileset expression with no context" |
|
69 | b"fileset expression with no context" | |
70 | ) |
|
70 | ) | |
71 | matchers.append(ctx.matchfileset(pat, badfn=badfn)) |
|
71 | matchers.append(ctx.matchfileset(pat, badfn=badfn)) | |
72 |
|
72 | |||
73 | if listsubrepos: |
|
73 | if listsubrepos: | |
74 | for subpath in ctx.substate: |
|
74 | for subpath in ctx.substate: | |
75 | sm = ctx.sub(subpath).matchfileset(pat, badfn=badfn) |
|
75 | sm = ctx.sub(subpath).matchfileset(pat, badfn=badfn) | |
76 | pm = prefixdirmatcher(subpath, sm, badfn=badfn) |
|
76 | pm = prefixdirmatcher(subpath, sm, badfn=badfn) | |
77 | matchers.append(pm) |
|
77 | matchers.append(pm) | |
78 |
|
78 | |||
79 | continue |
|
79 | continue | |
80 | other.append((kind, pat, source)) |
|
80 | other.append((kind, pat, source)) | |
81 | return matchers, other |
|
81 | return matchers, other | |
82 |
|
82 | |||
83 |
|
83 | |||
84 | def _expandsubinclude(kindpats, root): |
|
84 | def _expandsubinclude(kindpats, root): | |
85 | '''Returns the list of subinclude matcher args and the kindpats without the |
|
85 | '''Returns the list of subinclude matcher args and the kindpats without the | |
86 | subincludes in it.''' |
|
86 | subincludes in it.''' | |
87 | relmatchers = [] |
|
87 | relmatchers = [] | |
88 | other = [] |
|
88 | other = [] | |
89 |
|
89 | |||
90 | for kind, pat, source in kindpats: |
|
90 | for kind, pat, source in kindpats: | |
91 | if kind == b'subinclude': |
|
91 | if kind == b'subinclude': | |
92 | sourceroot = pathutil.dirname(util.normpath(source)) |
|
92 | sourceroot = pathutil.dirname(util.normpath(source)) | |
93 | pat = util.pconvert(pat) |
|
93 | pat = util.pconvert(pat) | |
94 | path = pathutil.join(sourceroot, pat) |
|
94 | path = pathutil.join(sourceroot, pat) | |
95 |
|
95 | |||
96 | newroot = pathutil.dirname(path) |
|
96 | newroot = pathutil.dirname(path) | |
97 | matcherargs = (newroot, b'', [], [b'include:%s' % path]) |
|
97 | matcherargs = (newroot, b'', [], [b'include:%s' % path]) | |
98 |
|
98 | |||
99 | prefix = pathutil.canonpath(root, root, newroot) |
|
99 | prefix = pathutil.canonpath(root, root, newroot) | |
100 | if prefix: |
|
100 | if prefix: | |
101 | prefix += b'/' |
|
101 | prefix += b'/' | |
102 | relmatchers.append((prefix, matcherargs)) |
|
102 | relmatchers.append((prefix, matcherargs)) | |
103 | else: |
|
103 | else: | |
104 | other.append((kind, pat, source)) |
|
104 | other.append((kind, pat, source)) | |
105 |
|
105 | |||
106 | return relmatchers, other |
|
106 | return relmatchers, other | |
107 |
|
107 | |||
108 |
|
108 | |||
109 | def _kindpatsalwaysmatch(kindpats): |
|
109 | def _kindpatsalwaysmatch(kindpats): | |
110 | """"Checks whether the kindspats match everything, as e.g. |
|
110 | """"Checks whether the kindspats match everything, as e.g. | |
111 | 'relpath:.' does. |
|
111 | 'relpath:.' does. | |
112 | """ |
|
112 | """ | |
113 | for kind, pat, source in kindpats: |
|
113 | for kind, pat, source in kindpats: | |
114 | if pat != b'' or kind not in [b'relpath', b'glob']: |
|
114 | if pat != b'' or kind not in [b'relpath', b'glob']: | |
115 | return False |
|
115 | return False | |
116 | return True |
|
116 | return True | |
117 |
|
117 | |||
118 |
|
118 | |||
119 | def _buildkindpatsmatcher( |
|
119 | def _buildkindpatsmatcher( | |
120 | matchercls, root, kindpats, ctx=None, listsubrepos=False, badfn=None |
|
120 | matchercls, root, kindpats, ctx=None, listsubrepos=False, badfn=None | |
121 | ): |
|
121 | ): | |
122 | matchers = [] |
|
122 | matchers = [] | |
123 | fms, kindpats = _expandsets( |
|
123 | fms, kindpats = _expandsets( | |
124 | kindpats, ctx=ctx, listsubrepos=listsubrepos, badfn=badfn |
|
124 | kindpats, ctx=ctx, listsubrepos=listsubrepos, badfn=badfn | |
125 | ) |
|
125 | ) | |
126 | if kindpats: |
|
126 | if kindpats: | |
127 | m = matchercls(root, kindpats, badfn=badfn) |
|
127 | m = matchercls(root, kindpats, badfn=badfn) | |
128 | matchers.append(m) |
|
128 | matchers.append(m) | |
129 | if fms: |
|
129 | if fms: | |
130 | matchers.extend(fms) |
|
130 | matchers.extend(fms) | |
131 | if not matchers: |
|
131 | if not matchers: | |
132 | return nevermatcher(badfn=badfn) |
|
132 | return nevermatcher(badfn=badfn) | |
133 | if len(matchers) == 1: |
|
133 | if len(matchers) == 1: | |
134 | return matchers[0] |
|
134 | return matchers[0] | |
135 | return unionmatcher(matchers) |
|
135 | return unionmatcher(matchers) | |
136 |
|
136 | |||
137 |
|
137 | |||
138 | def match( |
|
138 | def match( | |
139 | root, |
|
139 | root, | |
140 | cwd, |
|
140 | cwd, | |
141 | patterns=None, |
|
141 | patterns=None, | |
142 | include=None, |
|
142 | include=None, | |
143 | exclude=None, |
|
143 | exclude=None, | |
144 | default=b'glob', |
|
144 | default=b'glob', | |
145 | auditor=None, |
|
145 | auditor=None, | |
146 | ctx=None, |
|
146 | ctx=None, | |
147 | listsubrepos=False, |
|
147 | listsubrepos=False, | |
148 | warn=None, |
|
148 | warn=None, | |
149 | badfn=None, |
|
149 | badfn=None, | |
150 | icasefs=False, |
|
150 | icasefs=False, | |
151 | ): |
|
151 | ): | |
152 | r"""build an object to match a set of file patterns |
|
152 | r"""build an object to match a set of file patterns | |
153 |
|
153 | |||
154 | arguments: |
|
154 | arguments: | |
155 | root - the canonical root of the tree you're matching against |
|
155 | root - the canonical root of the tree you're matching against | |
156 | cwd - the current working directory, if relevant |
|
156 | cwd - the current working directory, if relevant | |
157 | patterns - patterns to find |
|
157 | patterns - patterns to find | |
158 | include - patterns to include (unless they are excluded) |
|
158 | include - patterns to include (unless they are excluded) | |
159 | exclude - patterns to exclude (even if they are included) |
|
159 | exclude - patterns to exclude (even if they are included) | |
160 | default - if a pattern in patterns has no explicit type, assume this one |
|
160 | default - if a pattern in patterns has no explicit type, assume this one | |
161 | auditor - optional path auditor |
|
161 | auditor - optional path auditor | |
162 | ctx - optional changecontext |
|
162 | ctx - optional changecontext | |
163 | listsubrepos - if True, recurse into subrepositories |
|
163 | listsubrepos - if True, recurse into subrepositories | |
164 | warn - optional function used for printing warnings |
|
164 | warn - optional function used for printing warnings | |
165 | badfn - optional bad() callback for this matcher instead of the default |
|
165 | badfn - optional bad() callback for this matcher instead of the default | |
166 | icasefs - make a matcher for wdir on case insensitive filesystems, which |
|
166 | icasefs - make a matcher for wdir on case insensitive filesystems, which | |
167 | normalizes the given patterns to the case in the filesystem |
|
167 | normalizes the given patterns to the case in the filesystem | |
168 |
|
168 | |||
169 | a pattern is one of: |
|
169 | a pattern is one of: | |
170 | 'glob:<glob>' - a glob relative to cwd |
|
170 | 'glob:<glob>' - a glob relative to cwd | |
171 | 're:<regexp>' - a regular expression |
|
171 | 're:<regexp>' - a regular expression | |
172 | 'path:<path>' - a path relative to repository root, which is matched |
|
172 | 'path:<path>' - a path relative to repository root, which is matched | |
173 | recursively |
|
173 | recursively | |
174 | 'rootfilesin:<path>' - a path relative to repository root, which is |
|
174 | 'rootfilesin:<path>' - a path relative to repository root, which is | |
175 | matched non-recursively (will not match subdirectories) |
|
175 | matched non-recursively (will not match subdirectories) | |
176 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
176 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
177 | 'relpath:<path>' - a path relative to cwd |
|
177 | 'relpath:<path>' - a path relative to cwd | |
178 | 'relre:<regexp>' - a regexp that needn't match the start of a name |
|
178 | 'relre:<regexp>' - a regexp that needn't match the start of a name | |
179 | 'set:<fileset>' - a fileset expression |
|
179 | 'set:<fileset>' - a fileset expression | |
180 | 'include:<path>' - a file of patterns to read and include |
|
180 | 'include:<path>' - a file of patterns to read and include | |
181 | 'subinclude:<path>' - a file of patterns to match against files under |
|
181 | 'subinclude:<path>' - a file of patterns to match against files under | |
182 | the same directory |
|
182 | the same directory | |
183 | '<something>' - a pattern of the specified default type |
|
183 | '<something>' - a pattern of the specified default type | |
184 |
|
184 | |||
185 | Usually a patternmatcher is returned: |
|
185 | Usually a patternmatcher is returned: | |
186 | >>> match(b'foo', b'.', [b're:.*\.c$', b'path:foo/a', b'*.py']) |
|
186 | >>> match(b'foo', b'.', [b're:.*\.c$', b'path:foo/a', b'*.py']) | |
187 | <patternmatcher patterns='.*\\.c$|foo/a(?:/|$)|[^/]*\\.py$'> |
|
187 | <patternmatcher patterns='.*\\.c$|foo/a(?:/|$)|[^/]*\\.py$'> | |
188 |
|
188 | |||
189 | Combining 'patterns' with 'include' (resp. 'exclude') gives an |
|
189 | Combining 'patterns' with 'include' (resp. 'exclude') gives an | |
190 | intersectionmatcher (resp. a differencematcher): |
|
190 | intersectionmatcher (resp. a differencematcher): | |
191 | >>> type(match(b'foo', b'.', [b're:.*\.c$'], include=[b'path:lib'])) |
|
191 | >>> type(match(b'foo', b'.', [b're:.*\.c$'], include=[b'path:lib'])) | |
192 | <class 'mercurial.match.intersectionmatcher'> |
|
192 | <class 'mercurial.match.intersectionmatcher'> | |
193 | >>> type(match(b'foo', b'.', [b're:.*\.c$'], exclude=[b'path:build'])) |
|
193 | >>> type(match(b'foo', b'.', [b're:.*\.c$'], exclude=[b'path:build'])) | |
194 | <class 'mercurial.match.differencematcher'> |
|
194 | <class 'mercurial.match.differencematcher'> | |
195 |
|
195 | |||
196 | Notice that, if 'patterns' is empty, an alwaysmatcher is returned: |
|
196 | Notice that, if 'patterns' is empty, an alwaysmatcher is returned: | |
197 | >>> match(b'foo', b'.', []) |
|
197 | >>> match(b'foo', b'.', []) | |
198 | <alwaysmatcher> |
|
198 | <alwaysmatcher> | |
199 |
|
199 | |||
200 | The 'default' argument determines which kind of pattern is assumed if a |
|
200 | The 'default' argument determines which kind of pattern is assumed if a | |
201 | pattern has no prefix: |
|
201 | pattern has no prefix: | |
202 | >>> match(b'foo', b'.', [b'.*\.c$'], default=b're') |
|
202 | >>> match(b'foo', b'.', [b'.*\.c$'], default=b're') | |
203 | <patternmatcher patterns='.*\\.c$'> |
|
203 | <patternmatcher patterns='.*\\.c$'> | |
204 | >>> match(b'foo', b'.', [b'main.py'], default=b'relpath') |
|
204 | >>> match(b'foo', b'.', [b'main.py'], default=b'relpath') | |
205 | <patternmatcher patterns='main\\.py(?:/|$)'> |
|
205 | <patternmatcher patterns='main\\.py(?:/|$)'> | |
206 | >>> match(b'foo', b'.', [b'main.py'], default=b're') |
|
206 | >>> match(b'foo', b'.', [b'main.py'], default=b're') | |
207 | <patternmatcher patterns='main.py'> |
|
207 | <patternmatcher patterns='main.py'> | |
208 |
|
208 | |||
209 | The primary use of matchers is to check whether a value (usually a file |
|
209 | The primary use of matchers is to check whether a value (usually a file | |
210 | name) matches againset one of the patterns given at initialization. There |
|
210 | name) matches againset one of the patterns given at initialization. There | |
211 | are two ways of doing this check. |
|
211 | are two ways of doing this check. | |
212 |
|
212 | |||
213 | >>> m = match(b'foo', b'', [b're:.*\.c$', b'relpath:a']) |
|
213 | >>> m = match(b'foo', b'', [b're:.*\.c$', b'relpath:a']) | |
214 |
|
214 | |||
215 | 1. Calling the matcher with a file name returns True if any pattern |
|
215 | 1. Calling the matcher with a file name returns True if any pattern | |
216 | matches that file name: |
|
216 | matches that file name: | |
217 | >>> m(b'a') |
|
217 | >>> m(b'a') | |
218 | True |
|
218 | True | |
219 | >>> m(b'main.c') |
|
219 | >>> m(b'main.c') | |
220 | True |
|
220 | True | |
221 | >>> m(b'test.py') |
|
221 | >>> m(b'test.py') | |
222 | False |
|
222 | False | |
223 |
|
223 | |||
224 | 2. Using the exact() method only returns True if the file name matches one |
|
224 | 2. Using the exact() method only returns True if the file name matches one | |
225 | of the exact patterns (i.e. not re: or glob: patterns): |
|
225 | of the exact patterns (i.e. not re: or glob: patterns): | |
226 | >>> m.exact(b'a') |
|
226 | >>> m.exact(b'a') | |
227 | True |
|
227 | True | |
228 | >>> m.exact(b'main.c') |
|
228 | >>> m.exact(b'main.c') | |
229 | False |
|
229 | False | |
230 | """ |
|
230 | """ | |
231 | normalize = _donormalize |
|
231 | normalize = _donormalize | |
232 | if icasefs: |
|
232 | if icasefs: | |
233 | dirstate = ctx.repo().dirstate |
|
233 | dirstate = ctx.repo().dirstate | |
234 | dsnormalize = dirstate.normalize |
|
234 | dsnormalize = dirstate.normalize | |
235 |
|
235 | |||
236 | def normalize(patterns, default, root, cwd, auditor, warn): |
|
236 | def normalize(patterns, default, root, cwd, auditor, warn): | |
237 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) |
|
237 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) | |
238 | kindpats = [] |
|
238 | kindpats = [] | |
239 | for kind, pats, source in kp: |
|
239 | for kind, pats, source in kp: | |
240 | if kind not in (b're', b'relre'): # regex can't be normalized |
|
240 | if kind not in (b're', b'relre'): # regex can't be normalized | |
241 | p = pats |
|
241 | p = pats | |
242 | pats = dsnormalize(pats) |
|
242 | pats = dsnormalize(pats) | |
243 |
|
243 | |||
244 | # Preserve the original to handle a case only rename. |
|
244 | # Preserve the original to handle a case only rename. | |
245 | if p != pats and p in dirstate: |
|
245 | if p != pats and p in dirstate: | |
246 | kindpats.append((kind, p, source)) |
|
246 | kindpats.append((kind, p, source)) | |
247 |
|
247 | |||
248 | kindpats.append((kind, pats, source)) |
|
248 | kindpats.append((kind, pats, source)) | |
249 | return kindpats |
|
249 | return kindpats | |
250 |
|
250 | |||
251 | if patterns: |
|
251 | if patterns: | |
252 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) |
|
252 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) | |
253 | if _kindpatsalwaysmatch(kindpats): |
|
253 | if _kindpatsalwaysmatch(kindpats): | |
254 | m = alwaysmatcher(badfn) |
|
254 | m = alwaysmatcher(badfn) | |
255 | else: |
|
255 | else: | |
256 | m = _buildkindpatsmatcher( |
|
256 | m = _buildkindpatsmatcher( | |
257 | patternmatcher, |
|
257 | patternmatcher, | |
258 | root, |
|
258 | root, | |
259 | kindpats, |
|
259 | kindpats, | |
260 | ctx=ctx, |
|
260 | ctx=ctx, | |
261 | listsubrepos=listsubrepos, |
|
261 | listsubrepos=listsubrepos, | |
262 | badfn=badfn, |
|
262 | badfn=badfn, | |
263 | ) |
|
263 | ) | |
264 | else: |
|
264 | else: | |
265 | # It's a little strange that no patterns means to match everything. |
|
265 | # It's a little strange that no patterns means to match everything. | |
266 | # Consider changing this to match nothing (probably using nevermatcher). |
|
266 | # Consider changing this to match nothing (probably using nevermatcher). | |
267 | m = alwaysmatcher(badfn) |
|
267 | m = alwaysmatcher(badfn) | |
268 |
|
268 | |||
269 | if include: |
|
269 | if include: | |
270 | kindpats = normalize(include, b'glob', root, cwd, auditor, warn) |
|
270 | kindpats = normalize(include, b'glob', root, cwd, auditor, warn) | |
271 | im = _buildkindpatsmatcher( |
|
271 | im = _buildkindpatsmatcher( | |
272 | includematcher, |
|
272 | includematcher, | |
273 | root, |
|
273 | root, | |
274 | kindpats, |
|
274 | kindpats, | |
275 | ctx=ctx, |
|
275 | ctx=ctx, | |
276 | listsubrepos=listsubrepos, |
|
276 | listsubrepos=listsubrepos, | |
277 | badfn=None, |
|
277 | badfn=None, | |
278 | ) |
|
278 | ) | |
279 | m = intersectmatchers(m, im) |
|
279 | m = intersectmatchers(m, im) | |
280 | if exclude: |
|
280 | if exclude: | |
281 | kindpats = normalize(exclude, b'glob', root, cwd, auditor, warn) |
|
281 | kindpats = normalize(exclude, b'glob', root, cwd, auditor, warn) | |
282 | em = _buildkindpatsmatcher( |
|
282 | em = _buildkindpatsmatcher( | |
283 | includematcher, |
|
283 | includematcher, | |
284 | root, |
|
284 | root, | |
285 | kindpats, |
|
285 | kindpats, | |
286 | ctx=ctx, |
|
286 | ctx=ctx, | |
287 | listsubrepos=listsubrepos, |
|
287 | listsubrepos=listsubrepos, | |
288 | badfn=None, |
|
288 | badfn=None, | |
289 | ) |
|
289 | ) | |
290 | m = differencematcher(m, em) |
|
290 | m = differencematcher(m, em) | |
291 | return m |
|
291 | return m | |
292 |
|
292 | |||
293 |
|
293 | |||
294 | def exact(files, badfn=None): |
|
294 | def exact(files, badfn=None): | |
295 | return exactmatcher(files, badfn=badfn) |
|
295 | return exactmatcher(files, badfn=badfn) | |
296 |
|
296 | |||
297 |
|
297 | |||
298 | def always(badfn=None): |
|
298 | def always(badfn=None): | |
299 | return alwaysmatcher(badfn) |
|
299 | return alwaysmatcher(badfn) | |
300 |
|
300 | |||
301 |
|
301 | |||
302 | def never(badfn=None): |
|
302 | def never(badfn=None): | |
303 | return nevermatcher(badfn) |
|
303 | return nevermatcher(badfn) | |
304 |
|
304 | |||
305 |
|
305 | |||
306 | def badmatch(match, badfn): |
|
306 | def badmatch(match, badfn): | |
307 | """Make a copy of the given matcher, replacing its bad method with the given |
|
307 | """Make a copy of the given matcher, replacing its bad method with the given | |
308 | one. |
|
308 | one. | |
309 | """ |
|
309 | """ | |
310 | m = copy.copy(match) |
|
310 | m = copy.copy(match) | |
311 | m.bad = badfn |
|
311 | m.bad = badfn | |
312 | return m |
|
312 | return m | |
313 |
|
313 | |||
314 |
|
314 | |||
315 | def _donormalize(patterns, default, root, cwd, auditor=None, warn=None): |
|
315 | def _donormalize(patterns, default, root, cwd, auditor=None, warn=None): | |
316 | '''Convert 'kind:pat' from the patterns list to tuples with kind and |
|
316 | '''Convert 'kind:pat' from the patterns list to tuples with kind and | |
317 | normalized and rooted patterns and with listfiles expanded.''' |
|
317 | normalized and rooted patterns and with listfiles expanded.''' | |
318 | kindpats = [] |
|
318 | kindpats = [] | |
319 | for kind, pat in [_patsplit(p, default) for p in patterns]: |
|
319 | for kind, pat in [_patsplit(p, default) for p in patterns]: | |
320 | if kind in cwdrelativepatternkinds: |
|
320 | if kind in cwdrelativepatternkinds: | |
321 | pat = pathutil.canonpath(root, cwd, pat, auditor=auditor) |
|
321 | pat = pathutil.canonpath(root, cwd, pat, auditor=auditor) | |
322 | elif kind in (b'relglob', b'path', b'rootfilesin', b'rootglob'): |
|
322 | elif kind in (b'relglob', b'path', b'rootfilesin', b'rootglob'): | |
323 | pat = util.normpath(pat) |
|
323 | pat = util.normpath(pat) | |
324 | elif kind in (b'listfile', b'listfile0'): |
|
324 | elif kind in (b'listfile', b'listfile0'): | |
325 | try: |
|
325 | try: | |
326 | files = util.readfile(pat) |
|
326 | files = util.readfile(pat) | |
327 | if kind == b'listfile0': |
|
327 | if kind == b'listfile0': | |
328 | files = files.split(b'\0') |
|
328 | files = files.split(b'\0') | |
329 | else: |
|
329 | else: | |
330 | files = files.splitlines() |
|
330 | files = files.splitlines() | |
331 | files = [f for f in files if f] |
|
331 | files = [f for f in files if f] | |
332 | except EnvironmentError: |
|
332 | except EnvironmentError: | |
333 | raise error.Abort(_(b"unable to read file list (%s)") % pat) |
|
333 | raise error.Abort(_(b"unable to read file list (%s)") % pat) | |
334 | for k, p, source in _donormalize( |
|
334 | for k, p, source in _donormalize( | |
335 | files, default, root, cwd, auditor, warn |
|
335 | files, default, root, cwd, auditor, warn | |
336 | ): |
|
336 | ): | |
337 | kindpats.append((k, p, pat)) |
|
337 | kindpats.append((k, p, pat)) | |
338 | continue |
|
338 | continue | |
339 | elif kind == b'include': |
|
339 | elif kind == b'include': | |
340 | try: |
|
340 | try: | |
341 | fullpath = os.path.join(root, util.localpath(pat)) |
|
341 | fullpath = os.path.join(root, util.localpath(pat)) | |
342 | includepats = readpatternfile(fullpath, warn) |
|
342 | includepats = readpatternfile(fullpath, warn) | |
343 | for k, p, source in _donormalize( |
|
343 | for k, p, source in _donormalize( | |
344 | includepats, default, root, cwd, auditor, warn |
|
344 | includepats, default, root, cwd, auditor, warn | |
345 | ): |
|
345 | ): | |
346 | kindpats.append((k, p, source or pat)) |
|
346 | kindpats.append((k, p, source or pat)) | |
347 | except error.Abort as inst: |
|
347 | except error.Abort as inst: | |
348 | raise error.Abort( |
|
348 | raise error.Abort( | |
349 | b'%s: %s' |
|
349 | b'%s: %s' | |
350 | % (pat, inst[0]) # pytype: disable=unsupported-operands |
|
350 | % (pat, inst[0]) # pytype: disable=unsupported-operands | |
351 | ) |
|
351 | ) | |
352 | except IOError as inst: |
|
352 | except IOError as inst: | |
353 | if warn: |
|
353 | if warn: | |
354 | warn( |
|
354 | warn( | |
355 | _(b"skipping unreadable pattern file '%s': %s\n") |
|
355 | _(b"skipping unreadable pattern file '%s': %s\n") | |
356 | % (pat, stringutil.forcebytestr(inst.strerror)) |
|
356 | % (pat, stringutil.forcebytestr(inst.strerror)) | |
357 | ) |
|
357 | ) | |
358 | continue |
|
358 | continue | |
359 | # else: re or relre - which cannot be normalized |
|
359 | # else: re or relre - which cannot be normalized | |
360 | kindpats.append((kind, pat, b'')) |
|
360 | kindpats.append((kind, pat, b'')) | |
361 | return kindpats |
|
361 | return kindpats | |
362 |
|
362 | |||
363 |
|
363 | |||
364 | class basematcher(object): |
|
364 | class basematcher(object): | |
365 | def __init__(self, badfn=None): |
|
365 | def __init__(self, badfn=None): | |
366 | if badfn is not None: |
|
366 | if badfn is not None: | |
367 | self.bad = badfn |
|
367 | self.bad = badfn | |
368 |
|
368 | |||
369 | def __call__(self, fn): |
|
369 | def __call__(self, fn): | |
370 | return self.matchfn(fn) |
|
370 | return self.matchfn(fn) | |
371 |
|
371 | |||
372 | # Callbacks related to how the matcher is used by dirstate.walk. |
|
372 | # Callbacks related to how the matcher is used by dirstate.walk. | |
373 | # Subscribers to these events must monkeypatch the matcher object. |
|
373 | # Subscribers to these events must monkeypatch the matcher object. | |
374 | def bad(self, f, msg): |
|
374 | def bad(self, f, msg): | |
375 | '''Callback from dirstate.walk for each explicit file that can't be |
|
375 | '''Callback from dirstate.walk for each explicit file that can't be | |
376 | found/accessed, with an error message.''' |
|
376 | found/accessed, with an error message.''' | |
377 |
|
377 | |||
378 | # If an explicitdir is set, it will be called when an explicitly listed |
|
|||
379 | # directory is visited. |
|
|||
380 | explicitdir = None |
|
|||
381 |
|
||||
382 | # If an traversedir is set, it will be called when a directory discovered |
|
378 | # If an traversedir is set, it will be called when a directory discovered | |
383 | # by recursive traversal is visited. |
|
379 | # by recursive traversal is visited. | |
384 | traversedir = None |
|
380 | traversedir = None | |
385 |
|
381 | |||
386 | @propertycache |
|
382 | @propertycache | |
387 | def _files(self): |
|
383 | def _files(self): | |
388 | return [] |
|
384 | return [] | |
389 |
|
385 | |||
390 | def files(self): |
|
386 | def files(self): | |
391 | '''Explicitly listed files or patterns or roots: |
|
387 | '''Explicitly listed files or patterns or roots: | |
392 | if no patterns or .always(): empty list, |
|
388 | if no patterns or .always(): empty list, | |
393 | if exact: list exact files, |
|
389 | if exact: list exact files, | |
394 | if not .anypats(): list all files and dirs, |
|
390 | if not .anypats(): list all files and dirs, | |
395 | else: optimal roots''' |
|
391 | else: optimal roots''' | |
396 | return self._files |
|
392 | return self._files | |
397 |
|
393 | |||
398 | @propertycache |
|
394 | @propertycache | |
399 | def _fileset(self): |
|
395 | def _fileset(self): | |
400 | return set(self._files) |
|
396 | return set(self._files) | |
401 |
|
397 | |||
402 | def exact(self, f): |
|
398 | def exact(self, f): | |
403 | '''Returns True if f is in .files().''' |
|
399 | '''Returns True if f is in .files().''' | |
404 | return f in self._fileset |
|
400 | return f in self._fileset | |
405 |
|
401 | |||
406 | def matchfn(self, f): |
|
402 | def matchfn(self, f): | |
407 | return False |
|
403 | return False | |
408 |
|
404 | |||
409 | def visitdir(self, dir): |
|
405 | def visitdir(self, dir): | |
410 | '''Decides whether a directory should be visited based on whether it |
|
406 | '''Decides whether a directory should be visited based on whether it | |
411 | has potential matches in it or one of its subdirectories. This is |
|
407 | has potential matches in it or one of its subdirectories. This is | |
412 | based on the match's primary, included, and excluded patterns. |
|
408 | based on the match's primary, included, and excluded patterns. | |
413 |
|
409 | |||
414 | Returns the string 'all' if the given directory and all subdirectories |
|
410 | Returns the string 'all' if the given directory and all subdirectories | |
415 | should be visited. Otherwise returns True or False indicating whether |
|
411 | should be visited. Otherwise returns True or False indicating whether | |
416 | the given directory should be visited. |
|
412 | the given directory should be visited. | |
417 | ''' |
|
413 | ''' | |
418 | return True |
|
414 | return True | |
419 |
|
415 | |||
420 | def visitchildrenset(self, dir): |
|
416 | def visitchildrenset(self, dir): | |
421 | '''Decides whether a directory should be visited based on whether it |
|
417 | '''Decides whether a directory should be visited based on whether it | |
422 | has potential matches in it or one of its subdirectories, and |
|
418 | has potential matches in it or one of its subdirectories, and | |
423 | potentially lists which subdirectories of that directory should be |
|
419 | potentially lists which subdirectories of that directory should be | |
424 | visited. This is based on the match's primary, included, and excluded |
|
420 | visited. This is based on the match's primary, included, and excluded | |
425 | patterns. |
|
421 | patterns. | |
426 |
|
422 | |||
427 | This function is very similar to 'visitdir', and the following mapping |
|
423 | This function is very similar to 'visitdir', and the following mapping | |
428 | can be applied: |
|
424 | can be applied: | |
429 |
|
425 | |||
430 | visitdir | visitchildrenlist |
|
426 | visitdir | visitchildrenlist | |
431 | ----------+------------------- |
|
427 | ----------+------------------- | |
432 | False | set() |
|
428 | False | set() | |
433 | 'all' | 'all' |
|
429 | 'all' | 'all' | |
434 | True | 'this' OR non-empty set of subdirs -or files- to visit |
|
430 | True | 'this' OR non-empty set of subdirs -or files- to visit | |
435 |
|
431 | |||
436 | Example: |
|
432 | Example: | |
437 | Assume matchers ['path:foo/bar', 'rootfilesin:qux'], we would return |
|
433 | Assume matchers ['path:foo/bar', 'rootfilesin:qux'], we would return | |
438 | the following values (assuming the implementation of visitchildrenset |
|
434 | the following values (assuming the implementation of visitchildrenset | |
439 | is capable of recognizing this; some implementations are not). |
|
435 | is capable of recognizing this; some implementations are not). | |
440 |
|
436 | |||
441 | '' -> {'foo', 'qux'} |
|
437 | '' -> {'foo', 'qux'} | |
442 | 'baz' -> set() |
|
438 | 'baz' -> set() | |
443 | 'foo' -> {'bar'} |
|
439 | 'foo' -> {'bar'} | |
444 | # Ideally this would be 'all', but since the prefix nature of matchers |
|
440 | # Ideally this would be 'all', but since the prefix nature of matchers | |
445 | # is applied to the entire matcher, we have to downgrade this to |
|
441 | # is applied to the entire matcher, we have to downgrade this to | |
446 | # 'this' due to the non-prefix 'rootfilesin'-kind matcher being mixed |
|
442 | # 'this' due to the non-prefix 'rootfilesin'-kind matcher being mixed | |
447 | # in. |
|
443 | # in. | |
448 | 'foo/bar' -> 'this' |
|
444 | 'foo/bar' -> 'this' | |
449 | 'qux' -> 'this' |
|
445 | 'qux' -> 'this' | |
450 |
|
446 | |||
451 | Important: |
|
447 | Important: | |
452 | Most matchers do not know if they're representing files or |
|
448 | Most matchers do not know if they're representing files or | |
453 | directories. They see ['path:dir/f'] and don't know whether 'f' is a |
|
449 | directories. They see ['path:dir/f'] and don't know whether 'f' is a | |
454 | file or a directory, so visitchildrenset('dir') for most matchers will |
|
450 | file or a directory, so visitchildrenset('dir') for most matchers will | |
455 | return {'f'}, but if the matcher knows it's a file (like exactmatcher |
|
451 | return {'f'}, but if the matcher knows it's a file (like exactmatcher | |
456 | does), it may return 'this'. Do not rely on the return being a set |
|
452 | does), it may return 'this'. Do not rely on the return being a set | |
457 | indicating that there are no files in this dir to investigate (or |
|
453 | indicating that there are no files in this dir to investigate (or | |
458 | equivalently that if there are files to investigate in 'dir' that it |
|
454 | equivalently that if there are files to investigate in 'dir' that it | |
459 | will always return 'this'). |
|
455 | will always return 'this'). | |
460 | ''' |
|
456 | ''' | |
461 | return b'this' |
|
457 | return b'this' | |
462 |
|
458 | |||
463 | def always(self): |
|
459 | def always(self): | |
464 | '''Matcher will match everything and .files() will be empty -- |
|
460 | '''Matcher will match everything and .files() will be empty -- | |
465 | optimization might be possible.''' |
|
461 | optimization might be possible.''' | |
466 | return False |
|
462 | return False | |
467 |
|
463 | |||
468 | def isexact(self): |
|
464 | def isexact(self): | |
469 | '''Matcher will match exactly the list of files in .files() -- |
|
465 | '''Matcher will match exactly the list of files in .files() -- | |
470 | optimization might be possible.''' |
|
466 | optimization might be possible.''' | |
471 | return False |
|
467 | return False | |
472 |
|
468 | |||
473 | def prefix(self): |
|
469 | def prefix(self): | |
474 | '''Matcher will match the paths in .files() recursively -- |
|
470 | '''Matcher will match the paths in .files() recursively -- | |
475 | optimization might be possible.''' |
|
471 | optimization might be possible.''' | |
476 | return False |
|
472 | return False | |
477 |
|
473 | |||
478 | def anypats(self): |
|
474 | def anypats(self): | |
479 | '''None of .always(), .isexact(), and .prefix() is true -- |
|
475 | '''None of .always(), .isexact(), and .prefix() is true -- | |
480 | optimizations will be difficult.''' |
|
476 | optimizations will be difficult.''' | |
481 | return not self.always() and not self.isexact() and not self.prefix() |
|
477 | return not self.always() and not self.isexact() and not self.prefix() | |
482 |
|
478 | |||
483 |
|
479 | |||
484 | class alwaysmatcher(basematcher): |
|
480 | class alwaysmatcher(basematcher): | |
485 | '''Matches everything.''' |
|
481 | '''Matches everything.''' | |
486 |
|
482 | |||
487 | def __init__(self, badfn=None): |
|
483 | def __init__(self, badfn=None): | |
488 | super(alwaysmatcher, self).__init__(badfn) |
|
484 | super(alwaysmatcher, self).__init__(badfn) | |
489 |
|
485 | |||
490 | def always(self): |
|
486 | def always(self): | |
491 | return True |
|
487 | return True | |
492 |
|
488 | |||
493 | def matchfn(self, f): |
|
489 | def matchfn(self, f): | |
494 | return True |
|
490 | return True | |
495 |
|
491 | |||
496 | def visitdir(self, dir): |
|
492 | def visitdir(self, dir): | |
497 | return b'all' |
|
493 | return b'all' | |
498 |
|
494 | |||
499 | def visitchildrenset(self, dir): |
|
495 | def visitchildrenset(self, dir): | |
500 | return b'all' |
|
496 | return b'all' | |
501 |
|
497 | |||
502 | def __repr__(self): |
|
498 | def __repr__(self): | |
503 | return r'<alwaysmatcher>' |
|
499 | return r'<alwaysmatcher>' | |
504 |
|
500 | |||
505 |
|
501 | |||
506 | class nevermatcher(basematcher): |
|
502 | class nevermatcher(basematcher): | |
507 | '''Matches nothing.''' |
|
503 | '''Matches nothing.''' | |
508 |
|
504 | |||
509 | def __init__(self, badfn=None): |
|
505 | def __init__(self, badfn=None): | |
510 | super(nevermatcher, self).__init__(badfn) |
|
506 | super(nevermatcher, self).__init__(badfn) | |
511 |
|
507 | |||
512 | # It's a little weird to say that the nevermatcher is an exact matcher |
|
508 | # It's a little weird to say that the nevermatcher is an exact matcher | |
513 | # or a prefix matcher, but it seems to make sense to let callers take |
|
509 | # or a prefix matcher, but it seems to make sense to let callers take | |
514 | # fast paths based on either. There will be no exact matches, nor any |
|
510 | # fast paths based on either. There will be no exact matches, nor any | |
515 | # prefixes (files() returns []), so fast paths iterating over them should |
|
511 | # prefixes (files() returns []), so fast paths iterating over them should | |
516 | # be efficient (and correct). |
|
512 | # be efficient (and correct). | |
517 | def isexact(self): |
|
513 | def isexact(self): | |
518 | return True |
|
514 | return True | |
519 |
|
515 | |||
520 | def prefix(self): |
|
516 | def prefix(self): | |
521 | return True |
|
517 | return True | |
522 |
|
518 | |||
523 | def visitdir(self, dir): |
|
519 | def visitdir(self, dir): | |
524 | return False |
|
520 | return False | |
525 |
|
521 | |||
526 | def visitchildrenset(self, dir): |
|
522 | def visitchildrenset(self, dir): | |
527 | return set() |
|
523 | return set() | |
528 |
|
524 | |||
529 | def __repr__(self): |
|
525 | def __repr__(self): | |
530 | return r'<nevermatcher>' |
|
526 | return r'<nevermatcher>' | |
531 |
|
527 | |||
532 |
|
528 | |||
533 | class predicatematcher(basematcher): |
|
529 | class predicatematcher(basematcher): | |
534 | """A matcher adapter for a simple boolean function""" |
|
530 | """A matcher adapter for a simple boolean function""" | |
535 |
|
531 | |||
536 | def __init__(self, predfn, predrepr=None, badfn=None): |
|
532 | def __init__(self, predfn, predrepr=None, badfn=None): | |
537 | super(predicatematcher, self).__init__(badfn) |
|
533 | super(predicatematcher, self).__init__(badfn) | |
538 | self.matchfn = predfn |
|
534 | self.matchfn = predfn | |
539 | self._predrepr = predrepr |
|
535 | self._predrepr = predrepr | |
540 |
|
536 | |||
541 | @encoding.strmethod |
|
537 | @encoding.strmethod | |
542 | def __repr__(self): |
|
538 | def __repr__(self): | |
543 | s = stringutil.buildrepr(self._predrepr) or pycompat.byterepr( |
|
539 | s = stringutil.buildrepr(self._predrepr) or pycompat.byterepr( | |
544 | self.matchfn |
|
540 | self.matchfn | |
545 | ) |
|
541 | ) | |
546 | return b'<predicatenmatcher pred=%s>' % s |
|
542 | return b'<predicatenmatcher pred=%s>' % s | |
547 |
|
543 | |||
548 |
|
544 | |||
549 | class patternmatcher(basematcher): |
|
545 | class patternmatcher(basematcher): | |
550 | """Matches a set of (kind, pat, source) against a 'root' directory. |
|
546 | """Matches a set of (kind, pat, source) against a 'root' directory. | |
551 |
|
547 | |||
552 | >>> kindpats = [ |
|
548 | >>> kindpats = [ | |
553 | ... (b're', br'.*\.c$', b''), |
|
549 | ... (b're', br'.*\.c$', b''), | |
554 | ... (b'path', b'foo/a', b''), |
|
550 | ... (b'path', b'foo/a', b''), | |
555 | ... (b'relpath', b'b', b''), |
|
551 | ... (b'relpath', b'b', b''), | |
556 | ... (b'glob', b'*.h', b''), |
|
552 | ... (b'glob', b'*.h', b''), | |
557 | ... ] |
|
553 | ... ] | |
558 | >>> m = patternmatcher(b'foo', kindpats) |
|
554 | >>> m = patternmatcher(b'foo', kindpats) | |
559 | >>> m(b'main.c') # matches re:.*\.c$ |
|
555 | >>> m(b'main.c') # matches re:.*\.c$ | |
560 | True |
|
556 | True | |
561 | >>> m(b'b.txt') |
|
557 | >>> m(b'b.txt') | |
562 | False |
|
558 | False | |
563 | >>> m(b'foo/a') # matches path:foo/a |
|
559 | >>> m(b'foo/a') # matches path:foo/a | |
564 | True |
|
560 | True | |
565 | >>> m(b'a') # does not match path:b, since 'root' is 'foo' |
|
561 | >>> m(b'a') # does not match path:b, since 'root' is 'foo' | |
566 | False |
|
562 | False | |
567 | >>> m(b'b') # matches relpath:b, since 'root' is 'foo' |
|
563 | >>> m(b'b') # matches relpath:b, since 'root' is 'foo' | |
568 | True |
|
564 | True | |
569 | >>> m(b'lib.h') # matches glob:*.h |
|
565 | >>> m(b'lib.h') # matches glob:*.h | |
570 | True |
|
566 | True | |
571 |
|
567 | |||
572 | >>> m.files() |
|
568 | >>> m.files() | |
573 | ['', 'foo/a', 'b', ''] |
|
569 | ['', 'foo/a', 'b', ''] | |
574 | >>> m.exact(b'foo/a') |
|
570 | >>> m.exact(b'foo/a') | |
575 | True |
|
571 | True | |
576 | >>> m.exact(b'b') |
|
572 | >>> m.exact(b'b') | |
577 | True |
|
573 | True | |
578 | >>> m.exact(b'lib.h') # exact matches are for (rel)path kinds |
|
574 | >>> m.exact(b'lib.h') # exact matches are for (rel)path kinds | |
579 | False |
|
575 | False | |
580 | """ |
|
576 | """ | |
581 |
|
577 | |||
582 | def __init__(self, root, kindpats, badfn=None): |
|
578 | def __init__(self, root, kindpats, badfn=None): | |
583 | super(patternmatcher, self).__init__(badfn) |
|
579 | super(patternmatcher, self).__init__(badfn) | |
584 |
|
580 | |||
585 | self._files = _explicitfiles(kindpats) |
|
581 | self._files = _explicitfiles(kindpats) | |
586 | self._prefix = _prefix(kindpats) |
|
582 | self._prefix = _prefix(kindpats) | |
587 | self._pats, self.matchfn = _buildmatch(kindpats, b'$', root) |
|
583 | self._pats, self.matchfn = _buildmatch(kindpats, b'$', root) | |
588 |
|
584 | |||
589 | @propertycache |
|
585 | @propertycache | |
590 | def _dirs(self): |
|
586 | def _dirs(self): | |
591 | return set(pathutil.dirs(self._fileset)) |
|
587 | return set(pathutil.dirs(self._fileset)) | |
592 |
|
588 | |||
593 | def visitdir(self, dir): |
|
589 | def visitdir(self, dir): | |
594 | if self._prefix and dir in self._fileset: |
|
590 | if self._prefix and dir in self._fileset: | |
595 | return b'all' |
|
591 | return b'all' | |
596 | return ( |
|
592 | return ( | |
597 | dir in self._fileset |
|
593 | dir in self._fileset | |
598 | or dir in self._dirs |
|
594 | or dir in self._dirs | |
599 | or any( |
|
595 | or any( | |
600 | parentdir in self._fileset |
|
596 | parentdir in self._fileset | |
601 | for parentdir in pathutil.finddirs(dir) |
|
597 | for parentdir in pathutil.finddirs(dir) | |
602 | ) |
|
598 | ) | |
603 | ) |
|
599 | ) | |
604 |
|
600 | |||
605 | def visitchildrenset(self, dir): |
|
601 | def visitchildrenset(self, dir): | |
606 | ret = self.visitdir(dir) |
|
602 | ret = self.visitdir(dir) | |
607 | if ret is True: |
|
603 | if ret is True: | |
608 | return b'this' |
|
604 | return b'this' | |
609 | elif not ret: |
|
605 | elif not ret: | |
610 | return set() |
|
606 | return set() | |
611 | assert ret == b'all' |
|
607 | assert ret == b'all' | |
612 | return b'all' |
|
608 | return b'all' | |
613 |
|
609 | |||
614 | def prefix(self): |
|
610 | def prefix(self): | |
615 | return self._prefix |
|
611 | return self._prefix | |
616 |
|
612 | |||
617 | @encoding.strmethod |
|
613 | @encoding.strmethod | |
618 | def __repr__(self): |
|
614 | def __repr__(self): | |
619 | return b'<patternmatcher patterns=%r>' % pycompat.bytestr(self._pats) |
|
615 | return b'<patternmatcher patterns=%r>' % pycompat.bytestr(self._pats) | |
620 |
|
616 | |||
621 |
|
617 | |||
622 | # This is basically a reimplementation of pathutil.dirs that stores the |
|
618 | # This is basically a reimplementation of pathutil.dirs that stores the | |
623 | # children instead of just a count of them, plus a small optional optimization |
|
619 | # children instead of just a count of them, plus a small optional optimization | |
624 | # to avoid some directories we don't need. |
|
620 | # to avoid some directories we don't need. | |
625 | class _dirchildren(object): |
|
621 | class _dirchildren(object): | |
626 | def __init__(self, paths, onlyinclude=None): |
|
622 | def __init__(self, paths, onlyinclude=None): | |
627 | self._dirs = {} |
|
623 | self._dirs = {} | |
628 | self._onlyinclude = onlyinclude or [] |
|
624 | self._onlyinclude = onlyinclude or [] | |
629 | addpath = self.addpath |
|
625 | addpath = self.addpath | |
630 | for f in paths: |
|
626 | for f in paths: | |
631 | addpath(f) |
|
627 | addpath(f) | |
632 |
|
628 | |||
633 | def addpath(self, path): |
|
629 | def addpath(self, path): | |
634 | if path == b'': |
|
630 | if path == b'': | |
635 | return |
|
631 | return | |
636 | dirs = self._dirs |
|
632 | dirs = self._dirs | |
637 | findsplitdirs = _dirchildren._findsplitdirs |
|
633 | findsplitdirs = _dirchildren._findsplitdirs | |
638 | for d, b in findsplitdirs(path): |
|
634 | for d, b in findsplitdirs(path): | |
639 | if d not in self._onlyinclude: |
|
635 | if d not in self._onlyinclude: | |
640 | continue |
|
636 | continue | |
641 | dirs.setdefault(d, set()).add(b) |
|
637 | dirs.setdefault(d, set()).add(b) | |
642 |
|
638 | |||
643 | @staticmethod |
|
639 | @staticmethod | |
644 | def _findsplitdirs(path): |
|
640 | def _findsplitdirs(path): | |
645 | # yields (dirname, basename) tuples, walking back to the root. This is |
|
641 | # yields (dirname, basename) tuples, walking back to the root. This is | |
646 | # very similar to pathutil.finddirs, except: |
|
642 | # very similar to pathutil.finddirs, except: | |
647 | # - produces a (dirname, basename) tuple, not just 'dirname' |
|
643 | # - produces a (dirname, basename) tuple, not just 'dirname' | |
648 | # Unlike manifest._splittopdir, this does not suffix `dirname` with a |
|
644 | # Unlike manifest._splittopdir, this does not suffix `dirname` with a | |
649 | # slash. |
|
645 | # slash. | |
650 | oldpos = len(path) |
|
646 | oldpos = len(path) | |
651 | pos = path.rfind(b'/') |
|
647 | pos = path.rfind(b'/') | |
652 | while pos != -1: |
|
648 | while pos != -1: | |
653 | yield path[:pos], path[pos + 1 : oldpos] |
|
649 | yield path[:pos], path[pos + 1 : oldpos] | |
654 | oldpos = pos |
|
650 | oldpos = pos | |
655 | pos = path.rfind(b'/', 0, pos) |
|
651 | pos = path.rfind(b'/', 0, pos) | |
656 | yield b'', path[:oldpos] |
|
652 | yield b'', path[:oldpos] | |
657 |
|
653 | |||
658 | def get(self, path): |
|
654 | def get(self, path): | |
659 | return self._dirs.get(path, set()) |
|
655 | return self._dirs.get(path, set()) | |
660 |
|
656 | |||
661 |
|
657 | |||
662 | class includematcher(basematcher): |
|
658 | class includematcher(basematcher): | |
663 | def __init__(self, root, kindpats, badfn=None): |
|
659 | def __init__(self, root, kindpats, badfn=None): | |
664 | super(includematcher, self).__init__(badfn) |
|
660 | super(includematcher, self).__init__(badfn) | |
665 |
|
661 | |||
666 | self._pats, self.matchfn = _buildmatch(kindpats, b'(?:/|$)', root) |
|
662 | self._pats, self.matchfn = _buildmatch(kindpats, b'(?:/|$)', root) | |
667 | self._prefix = _prefix(kindpats) |
|
663 | self._prefix = _prefix(kindpats) | |
668 | roots, dirs, parents = _rootsdirsandparents(kindpats) |
|
664 | roots, dirs, parents = _rootsdirsandparents(kindpats) | |
669 | # roots are directories which are recursively included. |
|
665 | # roots are directories which are recursively included. | |
670 | self._roots = set(roots) |
|
666 | self._roots = set(roots) | |
671 | # dirs are directories which are non-recursively included. |
|
667 | # dirs are directories which are non-recursively included. | |
672 | self._dirs = set(dirs) |
|
668 | self._dirs = set(dirs) | |
673 | # parents are directories which are non-recursively included because |
|
669 | # parents are directories which are non-recursively included because | |
674 | # they are needed to get to items in _dirs or _roots. |
|
670 | # they are needed to get to items in _dirs or _roots. | |
675 | self._parents = parents |
|
671 | self._parents = parents | |
676 |
|
672 | |||
677 | def visitdir(self, dir): |
|
673 | def visitdir(self, dir): | |
678 | if self._prefix and dir in self._roots: |
|
674 | if self._prefix and dir in self._roots: | |
679 | return b'all' |
|
675 | return b'all' | |
680 | return ( |
|
676 | return ( | |
681 | dir in self._roots |
|
677 | dir in self._roots | |
682 | or dir in self._dirs |
|
678 | or dir in self._dirs | |
683 | or dir in self._parents |
|
679 | or dir in self._parents | |
684 | or any( |
|
680 | or any( | |
685 | parentdir in self._roots for parentdir in pathutil.finddirs(dir) |
|
681 | parentdir in self._roots for parentdir in pathutil.finddirs(dir) | |
686 | ) |
|
682 | ) | |
687 | ) |
|
683 | ) | |
688 |
|
684 | |||
689 | @propertycache |
|
685 | @propertycache | |
690 | def _allparentschildren(self): |
|
686 | def _allparentschildren(self): | |
691 | # It may seem odd that we add dirs, roots, and parents, and then |
|
687 | # It may seem odd that we add dirs, roots, and parents, and then | |
692 | # restrict to only parents. This is to catch the case of: |
|
688 | # restrict to only parents. This is to catch the case of: | |
693 | # dirs = ['foo/bar'] |
|
689 | # dirs = ['foo/bar'] | |
694 | # parents = ['foo'] |
|
690 | # parents = ['foo'] | |
695 | # if we asked for the children of 'foo', but had only added |
|
691 | # if we asked for the children of 'foo', but had only added | |
696 | # self._parents, we wouldn't be able to respond ['bar']. |
|
692 | # self._parents, we wouldn't be able to respond ['bar']. | |
697 | return _dirchildren( |
|
693 | return _dirchildren( | |
698 | itertools.chain(self._dirs, self._roots, self._parents), |
|
694 | itertools.chain(self._dirs, self._roots, self._parents), | |
699 | onlyinclude=self._parents, |
|
695 | onlyinclude=self._parents, | |
700 | ) |
|
696 | ) | |
701 |
|
697 | |||
702 | def visitchildrenset(self, dir): |
|
698 | def visitchildrenset(self, dir): | |
703 | if self._prefix and dir in self._roots: |
|
699 | if self._prefix and dir in self._roots: | |
704 | return b'all' |
|
700 | return b'all' | |
705 | # Note: this does *not* include the 'dir in self._parents' case from |
|
701 | # Note: this does *not* include the 'dir in self._parents' case from | |
706 | # visitdir, that's handled below. |
|
702 | # visitdir, that's handled below. | |
707 | if ( |
|
703 | if ( | |
708 | b'' in self._roots |
|
704 | b'' in self._roots | |
709 | or dir in self._roots |
|
705 | or dir in self._roots | |
710 | or dir in self._dirs |
|
706 | or dir in self._dirs | |
711 | or any( |
|
707 | or any( | |
712 | parentdir in self._roots for parentdir in pathutil.finddirs(dir) |
|
708 | parentdir in self._roots for parentdir in pathutil.finddirs(dir) | |
713 | ) |
|
709 | ) | |
714 | ): |
|
710 | ): | |
715 | return b'this' |
|
711 | return b'this' | |
716 |
|
712 | |||
717 | if dir in self._parents: |
|
713 | if dir in self._parents: | |
718 | return self._allparentschildren.get(dir) or set() |
|
714 | return self._allparentschildren.get(dir) or set() | |
719 | return set() |
|
715 | return set() | |
720 |
|
716 | |||
721 | @encoding.strmethod |
|
717 | @encoding.strmethod | |
722 | def __repr__(self): |
|
718 | def __repr__(self): | |
723 | return b'<includematcher includes=%r>' % pycompat.bytestr(self._pats) |
|
719 | return b'<includematcher includes=%r>' % pycompat.bytestr(self._pats) | |
724 |
|
720 | |||
725 |
|
721 | |||
726 | class exactmatcher(basematcher): |
|
722 | class exactmatcher(basematcher): | |
727 | r'''Matches the input files exactly. They are interpreted as paths, not |
|
723 | r'''Matches the input files exactly. They are interpreted as paths, not | |
728 | patterns (so no kind-prefixes). |
|
724 | patterns (so no kind-prefixes). | |
729 |
|
725 | |||
730 | >>> m = exactmatcher([b'a.txt', br're:.*\.c$']) |
|
726 | >>> m = exactmatcher([b'a.txt', br're:.*\.c$']) | |
731 | >>> m(b'a.txt') |
|
727 | >>> m(b'a.txt') | |
732 | True |
|
728 | True | |
733 | >>> m(b'b.txt') |
|
729 | >>> m(b'b.txt') | |
734 | False |
|
730 | False | |
735 |
|
731 | |||
736 | Input files that would be matched are exactly those returned by .files() |
|
732 | Input files that would be matched are exactly those returned by .files() | |
737 | >>> m.files() |
|
733 | >>> m.files() | |
738 | ['a.txt', 're:.*\\.c$'] |
|
734 | ['a.txt', 're:.*\\.c$'] | |
739 |
|
735 | |||
740 | So pattern 're:.*\.c$' is not considered as a regex, but as a file name |
|
736 | So pattern 're:.*\.c$' is not considered as a regex, but as a file name | |
741 | >>> m(b'main.c') |
|
737 | >>> m(b'main.c') | |
742 | False |
|
738 | False | |
743 | >>> m(br're:.*\.c$') |
|
739 | >>> m(br're:.*\.c$') | |
744 | True |
|
740 | True | |
745 | ''' |
|
741 | ''' | |
746 |
|
742 | |||
747 | def __init__(self, files, badfn=None): |
|
743 | def __init__(self, files, badfn=None): | |
748 | super(exactmatcher, self).__init__(badfn) |
|
744 | super(exactmatcher, self).__init__(badfn) | |
749 |
|
745 | |||
750 | if isinstance(files, list): |
|
746 | if isinstance(files, list): | |
751 | self._files = files |
|
747 | self._files = files | |
752 | else: |
|
748 | else: | |
753 | self._files = list(files) |
|
749 | self._files = list(files) | |
754 |
|
750 | |||
755 | matchfn = basematcher.exact |
|
751 | matchfn = basematcher.exact | |
756 |
|
752 | |||
757 | @propertycache |
|
753 | @propertycache | |
758 | def _dirs(self): |
|
754 | def _dirs(self): | |
759 | return set(pathutil.dirs(self._fileset)) |
|
755 | return set(pathutil.dirs(self._fileset)) | |
760 |
|
756 | |||
761 | def visitdir(self, dir): |
|
757 | def visitdir(self, dir): | |
762 | return dir in self._dirs |
|
758 | return dir in self._dirs | |
763 |
|
759 | |||
764 | def visitchildrenset(self, dir): |
|
760 | def visitchildrenset(self, dir): | |
765 | if not self._fileset or dir not in self._dirs: |
|
761 | if not self._fileset or dir not in self._dirs: | |
766 | return set() |
|
762 | return set() | |
767 |
|
763 | |||
768 | candidates = self._fileset | self._dirs - {b''} |
|
764 | candidates = self._fileset | self._dirs - {b''} | |
769 | if dir != b'': |
|
765 | if dir != b'': | |
770 | d = dir + b'/' |
|
766 | d = dir + b'/' | |
771 | candidates = set(c[len(d) :] for c in candidates if c.startswith(d)) |
|
767 | candidates = set(c[len(d) :] for c in candidates if c.startswith(d)) | |
772 | # self._dirs includes all of the directories, recursively, so if |
|
768 | # self._dirs includes all of the directories, recursively, so if | |
773 | # we're attempting to match foo/bar/baz.txt, it'll have '', 'foo', |
|
769 | # we're attempting to match foo/bar/baz.txt, it'll have '', 'foo', | |
774 | # 'foo/bar' in it. Thus we can safely ignore a candidate that has a |
|
770 | # 'foo/bar' in it. Thus we can safely ignore a candidate that has a | |
775 | # '/' in it, indicating a it's for a subdir-of-a-subdir; the |
|
771 | # '/' in it, indicating a it's for a subdir-of-a-subdir; the | |
776 | # immediate subdir will be in there without a slash. |
|
772 | # immediate subdir will be in there without a slash. | |
777 | ret = {c for c in candidates if b'/' not in c} |
|
773 | ret = {c for c in candidates if b'/' not in c} | |
778 | # We really do not expect ret to be empty, since that would imply that |
|
774 | # We really do not expect ret to be empty, since that would imply that | |
779 | # there's something in _dirs that didn't have a file in _fileset. |
|
775 | # there's something in _dirs that didn't have a file in _fileset. | |
780 | assert ret |
|
776 | assert ret | |
781 | return ret |
|
777 | return ret | |
782 |
|
778 | |||
783 | def isexact(self): |
|
779 | def isexact(self): | |
784 | return True |
|
780 | return True | |
785 |
|
781 | |||
786 | @encoding.strmethod |
|
782 | @encoding.strmethod | |
787 | def __repr__(self): |
|
783 | def __repr__(self): | |
788 | return b'<exactmatcher files=%r>' % self._files |
|
784 | return b'<exactmatcher files=%r>' % self._files | |
789 |
|
785 | |||
790 |
|
786 | |||
791 | class differencematcher(basematcher): |
|
787 | class differencematcher(basematcher): | |
792 | '''Composes two matchers by matching if the first matches and the second |
|
788 | '''Composes two matchers by matching if the first matches and the second | |
793 | does not. |
|
789 | does not. | |
794 |
|
790 | |||
795 |
The second matcher's non-matching-attributes (bad, |
|
791 | The second matcher's non-matching-attributes (bad, traversedir) are ignored. | |
796 | traversedir) are ignored. |
|
|||
797 | ''' |
|
792 | ''' | |
798 |
|
793 | |||
799 | def __init__(self, m1, m2): |
|
794 | def __init__(self, m1, m2): | |
800 | super(differencematcher, self).__init__() |
|
795 | super(differencematcher, self).__init__() | |
801 | self._m1 = m1 |
|
796 | self._m1 = m1 | |
802 | self._m2 = m2 |
|
797 | self._m2 = m2 | |
803 | self.bad = m1.bad |
|
798 | self.bad = m1.bad | |
804 | self.explicitdir = m1.explicitdir |
|
|||
805 | self.traversedir = m1.traversedir |
|
799 | self.traversedir = m1.traversedir | |
806 |
|
800 | |||
807 | def matchfn(self, f): |
|
801 | def matchfn(self, f): | |
808 | return self._m1(f) and not self._m2(f) |
|
802 | return self._m1(f) and not self._m2(f) | |
809 |
|
803 | |||
810 | @propertycache |
|
804 | @propertycache | |
811 | def _files(self): |
|
805 | def _files(self): | |
812 | if self.isexact(): |
|
806 | if self.isexact(): | |
813 | return [f for f in self._m1.files() if self(f)] |
|
807 | return [f for f in self._m1.files() if self(f)] | |
814 | # If m1 is not an exact matcher, we can't easily figure out the set of |
|
808 | # If m1 is not an exact matcher, we can't easily figure out the set of | |
815 | # files, because its files() are not always files. For example, if |
|
809 | # files, because its files() are not always files. For example, if | |
816 | # m1 is "path:dir" and m2 is "rootfileins:.", we don't |
|
810 | # m1 is "path:dir" and m2 is "rootfileins:.", we don't | |
817 | # want to remove "dir" from the set even though it would match m2, |
|
811 | # want to remove "dir" from the set even though it would match m2, | |
818 | # because the "dir" in m1 may not be a file. |
|
812 | # because the "dir" in m1 may not be a file. | |
819 | return self._m1.files() |
|
813 | return self._m1.files() | |
820 |
|
814 | |||
821 | def visitdir(self, dir): |
|
815 | def visitdir(self, dir): | |
822 | if self._m2.visitdir(dir) == b'all': |
|
816 | if self._m2.visitdir(dir) == b'all': | |
823 | return False |
|
817 | return False | |
824 | elif not self._m2.visitdir(dir): |
|
818 | elif not self._m2.visitdir(dir): | |
825 | # m2 does not match dir, we can return 'all' here if possible |
|
819 | # m2 does not match dir, we can return 'all' here if possible | |
826 | return self._m1.visitdir(dir) |
|
820 | return self._m1.visitdir(dir) | |
827 | return bool(self._m1.visitdir(dir)) |
|
821 | return bool(self._m1.visitdir(dir)) | |
828 |
|
822 | |||
829 | def visitchildrenset(self, dir): |
|
823 | def visitchildrenset(self, dir): | |
830 | m2_set = self._m2.visitchildrenset(dir) |
|
824 | m2_set = self._m2.visitchildrenset(dir) | |
831 | if m2_set == b'all': |
|
825 | if m2_set == b'all': | |
832 | return set() |
|
826 | return set() | |
833 | m1_set = self._m1.visitchildrenset(dir) |
|
827 | m1_set = self._m1.visitchildrenset(dir) | |
834 | # Possible values for m1: 'all', 'this', set(...), set() |
|
828 | # Possible values for m1: 'all', 'this', set(...), set() | |
835 | # Possible values for m2: 'this', set(...), set() |
|
829 | # Possible values for m2: 'this', set(...), set() | |
836 | # If m2 has nothing under here that we care about, return m1, even if |
|
830 | # If m2 has nothing under here that we care about, return m1, even if | |
837 | # it's 'all'. This is a change in behavior from visitdir, which would |
|
831 | # it's 'all'. This is a change in behavior from visitdir, which would | |
838 | # return True, not 'all', for some reason. |
|
832 | # return True, not 'all', for some reason. | |
839 | if not m2_set: |
|
833 | if not m2_set: | |
840 | return m1_set |
|
834 | return m1_set | |
841 | if m1_set in [b'all', b'this']: |
|
835 | if m1_set in [b'all', b'this']: | |
842 | # Never return 'all' here if m2_set is any kind of non-empty (either |
|
836 | # Never return 'all' here if m2_set is any kind of non-empty (either | |
843 | # 'this' or set(foo)), since m2 might return set() for a |
|
837 | # 'this' or set(foo)), since m2 might return set() for a | |
844 | # subdirectory. |
|
838 | # subdirectory. | |
845 | return b'this' |
|
839 | return b'this' | |
846 | # Possible values for m1: set(...), set() |
|
840 | # Possible values for m1: set(...), set() | |
847 | # Possible values for m2: 'this', set(...) |
|
841 | # Possible values for m2: 'this', set(...) | |
848 | # We ignore m2's set results. They're possibly incorrect: |
|
842 | # We ignore m2's set results. They're possibly incorrect: | |
849 | # m1 = path:dir/subdir, m2=rootfilesin:dir, visitchildrenset(''): |
|
843 | # m1 = path:dir/subdir, m2=rootfilesin:dir, visitchildrenset(''): | |
850 | # m1 returns {'dir'}, m2 returns {'dir'}, if we subtracted we'd |
|
844 | # m1 returns {'dir'}, m2 returns {'dir'}, if we subtracted we'd | |
851 | # return set(), which is *not* correct, we still need to visit 'dir'! |
|
845 | # return set(), which is *not* correct, we still need to visit 'dir'! | |
852 | return m1_set |
|
846 | return m1_set | |
853 |
|
847 | |||
854 | def isexact(self): |
|
848 | def isexact(self): | |
855 | return self._m1.isexact() |
|
849 | return self._m1.isexact() | |
856 |
|
850 | |||
857 | @encoding.strmethod |
|
851 | @encoding.strmethod | |
858 | def __repr__(self): |
|
852 | def __repr__(self): | |
859 | return b'<differencematcher m1=%r, m2=%r>' % (self._m1, self._m2) |
|
853 | return b'<differencematcher m1=%r, m2=%r>' % (self._m1, self._m2) | |
860 |
|
854 | |||
861 |
|
855 | |||
862 | def intersectmatchers(m1, m2): |
|
856 | def intersectmatchers(m1, m2): | |
863 | '''Composes two matchers by matching if both of them match. |
|
857 | '''Composes two matchers by matching if both of them match. | |
864 |
|
858 | |||
865 |
The second matcher's non-matching-attributes (bad, |
|
859 | The second matcher's non-matching-attributes (bad, traversedir) are ignored. | |
866 | traversedir) are ignored. |
|
|||
867 | ''' |
|
860 | ''' | |
868 | if m1 is None or m2 is None: |
|
861 | if m1 is None or m2 is None: | |
869 | return m1 or m2 |
|
862 | return m1 or m2 | |
870 | if m1.always(): |
|
863 | if m1.always(): | |
871 | m = copy.copy(m2) |
|
864 | m = copy.copy(m2) | |
872 | # TODO: Consider encapsulating these things in a class so there's only |
|
865 | # TODO: Consider encapsulating these things in a class so there's only | |
873 | # one thing to copy from m1. |
|
866 | # one thing to copy from m1. | |
874 | m.bad = m1.bad |
|
867 | m.bad = m1.bad | |
875 | m.explicitdir = m1.explicitdir |
|
|||
876 | m.traversedir = m1.traversedir |
|
868 | m.traversedir = m1.traversedir | |
877 | return m |
|
869 | return m | |
878 | if m2.always(): |
|
870 | if m2.always(): | |
879 | m = copy.copy(m1) |
|
871 | m = copy.copy(m1) | |
880 | return m |
|
872 | return m | |
881 | return intersectionmatcher(m1, m2) |
|
873 | return intersectionmatcher(m1, m2) | |
882 |
|
874 | |||
883 |
|
875 | |||
884 | class intersectionmatcher(basematcher): |
|
876 | class intersectionmatcher(basematcher): | |
885 | def __init__(self, m1, m2): |
|
877 | def __init__(self, m1, m2): | |
886 | super(intersectionmatcher, self).__init__() |
|
878 | super(intersectionmatcher, self).__init__() | |
887 | self._m1 = m1 |
|
879 | self._m1 = m1 | |
888 | self._m2 = m2 |
|
880 | self._m2 = m2 | |
889 | self.bad = m1.bad |
|
881 | self.bad = m1.bad | |
890 | self.explicitdir = m1.explicitdir |
|
|||
891 | self.traversedir = m1.traversedir |
|
882 | self.traversedir = m1.traversedir | |
892 |
|
883 | |||
893 | @propertycache |
|
884 | @propertycache | |
894 | def _files(self): |
|
885 | def _files(self): | |
895 | if self.isexact(): |
|
886 | if self.isexact(): | |
896 | m1, m2 = self._m1, self._m2 |
|
887 | m1, m2 = self._m1, self._m2 | |
897 | if not m1.isexact(): |
|
888 | if not m1.isexact(): | |
898 | m1, m2 = m2, m1 |
|
889 | m1, m2 = m2, m1 | |
899 | return [f for f in m1.files() if m2(f)] |
|
890 | return [f for f in m1.files() if m2(f)] | |
900 | # It neither m1 nor m2 is an exact matcher, we can't easily intersect |
|
891 | # It neither m1 nor m2 is an exact matcher, we can't easily intersect | |
901 | # the set of files, because their files() are not always files. For |
|
892 | # the set of files, because their files() are not always files. For | |
902 | # example, if intersecting a matcher "-I glob:foo.txt" with matcher of |
|
893 | # example, if intersecting a matcher "-I glob:foo.txt" with matcher of | |
903 | # "path:dir2", we don't want to remove "dir2" from the set. |
|
894 | # "path:dir2", we don't want to remove "dir2" from the set. | |
904 | return self._m1.files() + self._m2.files() |
|
895 | return self._m1.files() + self._m2.files() | |
905 |
|
896 | |||
906 | def matchfn(self, f): |
|
897 | def matchfn(self, f): | |
907 | return self._m1(f) and self._m2(f) |
|
898 | return self._m1(f) and self._m2(f) | |
908 |
|
899 | |||
909 | def visitdir(self, dir): |
|
900 | def visitdir(self, dir): | |
910 | visit1 = self._m1.visitdir(dir) |
|
901 | visit1 = self._m1.visitdir(dir) | |
911 | if visit1 == b'all': |
|
902 | if visit1 == b'all': | |
912 | return self._m2.visitdir(dir) |
|
903 | return self._m2.visitdir(dir) | |
913 | # bool() because visit1=True + visit2='all' should not be 'all' |
|
904 | # bool() because visit1=True + visit2='all' should not be 'all' | |
914 | return bool(visit1 and self._m2.visitdir(dir)) |
|
905 | return bool(visit1 and self._m2.visitdir(dir)) | |
915 |
|
906 | |||
916 | def visitchildrenset(self, dir): |
|
907 | def visitchildrenset(self, dir): | |
917 | m1_set = self._m1.visitchildrenset(dir) |
|
908 | m1_set = self._m1.visitchildrenset(dir) | |
918 | if not m1_set: |
|
909 | if not m1_set: | |
919 | return set() |
|
910 | return set() | |
920 | m2_set = self._m2.visitchildrenset(dir) |
|
911 | m2_set = self._m2.visitchildrenset(dir) | |
921 | if not m2_set: |
|
912 | if not m2_set: | |
922 | return set() |
|
913 | return set() | |
923 |
|
914 | |||
924 | if m1_set == b'all': |
|
915 | if m1_set == b'all': | |
925 | return m2_set |
|
916 | return m2_set | |
926 | elif m2_set == b'all': |
|
917 | elif m2_set == b'all': | |
927 | return m1_set |
|
918 | return m1_set | |
928 |
|
919 | |||
929 | if m1_set == b'this' or m2_set == b'this': |
|
920 | if m1_set == b'this' or m2_set == b'this': | |
930 | return b'this' |
|
921 | return b'this' | |
931 |
|
922 | |||
932 | assert isinstance(m1_set, set) and isinstance(m2_set, set) |
|
923 | assert isinstance(m1_set, set) and isinstance(m2_set, set) | |
933 | return m1_set.intersection(m2_set) |
|
924 | return m1_set.intersection(m2_set) | |
934 |
|
925 | |||
935 | def always(self): |
|
926 | def always(self): | |
936 | return self._m1.always() and self._m2.always() |
|
927 | return self._m1.always() and self._m2.always() | |
937 |
|
928 | |||
938 | def isexact(self): |
|
929 | def isexact(self): | |
939 | return self._m1.isexact() or self._m2.isexact() |
|
930 | return self._m1.isexact() or self._m2.isexact() | |
940 |
|
931 | |||
941 | @encoding.strmethod |
|
932 | @encoding.strmethod | |
942 | def __repr__(self): |
|
933 | def __repr__(self): | |
943 | return b'<intersectionmatcher m1=%r, m2=%r>' % (self._m1, self._m2) |
|
934 | return b'<intersectionmatcher m1=%r, m2=%r>' % (self._m1, self._m2) | |
944 |
|
935 | |||
945 |
|
936 | |||
946 | class subdirmatcher(basematcher): |
|
937 | class subdirmatcher(basematcher): | |
947 | """Adapt a matcher to work on a subdirectory only. |
|
938 | """Adapt a matcher to work on a subdirectory only. | |
948 |
|
939 | |||
949 | The paths are remapped to remove/insert the path as needed: |
|
940 | The paths are remapped to remove/insert the path as needed: | |
950 |
|
941 | |||
951 | >>> from . import pycompat |
|
942 | >>> from . import pycompat | |
952 | >>> m1 = match(b'root', b'', [b'a.txt', b'sub/b.txt']) |
|
943 | >>> m1 = match(b'root', b'', [b'a.txt', b'sub/b.txt']) | |
953 | >>> m2 = subdirmatcher(b'sub', m1) |
|
944 | >>> m2 = subdirmatcher(b'sub', m1) | |
954 | >>> m2(b'a.txt') |
|
945 | >>> m2(b'a.txt') | |
955 | False |
|
946 | False | |
956 | >>> m2(b'b.txt') |
|
947 | >>> m2(b'b.txt') | |
957 | True |
|
948 | True | |
958 | >>> m2.matchfn(b'a.txt') |
|
949 | >>> m2.matchfn(b'a.txt') | |
959 | False |
|
950 | False | |
960 | >>> m2.matchfn(b'b.txt') |
|
951 | >>> m2.matchfn(b'b.txt') | |
961 | True |
|
952 | True | |
962 | >>> m2.files() |
|
953 | >>> m2.files() | |
963 | ['b.txt'] |
|
954 | ['b.txt'] | |
964 | >>> m2.exact(b'b.txt') |
|
955 | >>> m2.exact(b'b.txt') | |
965 | True |
|
956 | True | |
966 | >>> def bad(f, msg): |
|
957 | >>> def bad(f, msg): | |
967 | ... print(pycompat.sysstr(b"%s: %s" % (f, msg))) |
|
958 | ... print(pycompat.sysstr(b"%s: %s" % (f, msg))) | |
968 | >>> m1.bad = bad |
|
959 | >>> m1.bad = bad | |
969 | >>> m2.bad(b'x.txt', b'No such file') |
|
960 | >>> m2.bad(b'x.txt', b'No such file') | |
970 | sub/x.txt: No such file |
|
961 | sub/x.txt: No such file | |
971 | """ |
|
962 | """ | |
972 |
|
963 | |||
973 | def __init__(self, path, matcher): |
|
964 | def __init__(self, path, matcher): | |
974 | super(subdirmatcher, self).__init__() |
|
965 | super(subdirmatcher, self).__init__() | |
975 | self._path = path |
|
966 | self._path = path | |
976 | self._matcher = matcher |
|
967 | self._matcher = matcher | |
977 | self._always = matcher.always() |
|
968 | self._always = matcher.always() | |
978 |
|
969 | |||
979 | self._files = [ |
|
970 | self._files = [ | |
980 | f[len(path) + 1 :] |
|
971 | f[len(path) + 1 :] | |
981 | for f in matcher._files |
|
972 | for f in matcher._files | |
982 | if f.startswith(path + b"/") |
|
973 | if f.startswith(path + b"/") | |
983 | ] |
|
974 | ] | |
984 |
|
975 | |||
985 | # If the parent repo had a path to this subrepo and the matcher is |
|
976 | # If the parent repo had a path to this subrepo and the matcher is | |
986 | # a prefix matcher, this submatcher always matches. |
|
977 | # a prefix matcher, this submatcher always matches. | |
987 | if matcher.prefix(): |
|
978 | if matcher.prefix(): | |
988 | self._always = any(f == path for f in matcher._files) |
|
979 | self._always = any(f == path for f in matcher._files) | |
989 |
|
980 | |||
990 | def bad(self, f, msg): |
|
981 | def bad(self, f, msg): | |
991 | self._matcher.bad(self._path + b"/" + f, msg) |
|
982 | self._matcher.bad(self._path + b"/" + f, msg) | |
992 |
|
983 | |||
993 | def matchfn(self, f): |
|
984 | def matchfn(self, f): | |
994 | # Some information is lost in the superclass's constructor, so we |
|
985 | # Some information is lost in the superclass's constructor, so we | |
995 | # can not accurately create the matching function for the subdirectory |
|
986 | # can not accurately create the matching function for the subdirectory | |
996 | # from the inputs. Instead, we override matchfn() and visitdir() to |
|
987 | # from the inputs. Instead, we override matchfn() and visitdir() to | |
997 | # call the original matcher with the subdirectory path prepended. |
|
988 | # call the original matcher with the subdirectory path prepended. | |
998 | return self._matcher.matchfn(self._path + b"/" + f) |
|
989 | return self._matcher.matchfn(self._path + b"/" + f) | |
999 |
|
990 | |||
1000 | def visitdir(self, dir): |
|
991 | def visitdir(self, dir): | |
1001 | if dir == b'': |
|
992 | if dir == b'': | |
1002 | dir = self._path |
|
993 | dir = self._path | |
1003 | else: |
|
994 | else: | |
1004 | dir = self._path + b"/" + dir |
|
995 | dir = self._path + b"/" + dir | |
1005 | return self._matcher.visitdir(dir) |
|
996 | return self._matcher.visitdir(dir) | |
1006 |
|
997 | |||
1007 | def visitchildrenset(self, dir): |
|
998 | def visitchildrenset(self, dir): | |
1008 | if dir == b'': |
|
999 | if dir == b'': | |
1009 | dir = self._path |
|
1000 | dir = self._path | |
1010 | else: |
|
1001 | else: | |
1011 | dir = self._path + b"/" + dir |
|
1002 | dir = self._path + b"/" + dir | |
1012 | return self._matcher.visitchildrenset(dir) |
|
1003 | return self._matcher.visitchildrenset(dir) | |
1013 |
|
1004 | |||
1014 | def always(self): |
|
1005 | def always(self): | |
1015 | return self._always |
|
1006 | return self._always | |
1016 |
|
1007 | |||
1017 | def prefix(self): |
|
1008 | def prefix(self): | |
1018 | return self._matcher.prefix() and not self._always |
|
1009 | return self._matcher.prefix() and not self._always | |
1019 |
|
1010 | |||
1020 | @encoding.strmethod |
|
1011 | @encoding.strmethod | |
1021 | def __repr__(self): |
|
1012 | def __repr__(self): | |
1022 | return b'<subdirmatcher path=%r, matcher=%r>' % ( |
|
1013 | return b'<subdirmatcher path=%r, matcher=%r>' % ( | |
1023 | self._path, |
|
1014 | self._path, | |
1024 | self._matcher, |
|
1015 | self._matcher, | |
1025 | ) |
|
1016 | ) | |
1026 |
|
1017 | |||
1027 |
|
1018 | |||
1028 | class prefixdirmatcher(basematcher): |
|
1019 | class prefixdirmatcher(basematcher): | |
1029 | """Adapt a matcher to work on a parent directory. |
|
1020 | """Adapt a matcher to work on a parent directory. | |
1030 |
|
1021 | |||
1031 |
The matcher's non-matching-attributes (bad, |
|
1022 | The matcher's non-matching-attributes (bad, traversedir) are ignored. | |
1032 | ignored. |
|
|||
1033 |
|
1023 | |||
1034 | The prefix path should usually be the relative path from the root of |
|
1024 | The prefix path should usually be the relative path from the root of | |
1035 | this matcher to the root of the wrapped matcher. |
|
1025 | this matcher to the root of the wrapped matcher. | |
1036 |
|
1026 | |||
1037 | >>> m1 = match(util.localpath(b'root/d/e'), b'f', [b'../a.txt', b'b.txt']) |
|
1027 | >>> m1 = match(util.localpath(b'root/d/e'), b'f', [b'../a.txt', b'b.txt']) | |
1038 | >>> m2 = prefixdirmatcher(b'd/e', m1) |
|
1028 | >>> m2 = prefixdirmatcher(b'd/e', m1) | |
1039 | >>> m2(b'a.txt') |
|
1029 | >>> m2(b'a.txt') | |
1040 | False |
|
1030 | False | |
1041 | >>> m2(b'd/e/a.txt') |
|
1031 | >>> m2(b'd/e/a.txt') | |
1042 | True |
|
1032 | True | |
1043 | >>> m2(b'd/e/b.txt') |
|
1033 | >>> m2(b'd/e/b.txt') | |
1044 | False |
|
1034 | False | |
1045 | >>> m2.files() |
|
1035 | >>> m2.files() | |
1046 | ['d/e/a.txt', 'd/e/f/b.txt'] |
|
1036 | ['d/e/a.txt', 'd/e/f/b.txt'] | |
1047 | >>> m2.exact(b'd/e/a.txt') |
|
1037 | >>> m2.exact(b'd/e/a.txt') | |
1048 | True |
|
1038 | True | |
1049 | >>> m2.visitdir(b'd') |
|
1039 | >>> m2.visitdir(b'd') | |
1050 | True |
|
1040 | True | |
1051 | >>> m2.visitdir(b'd/e') |
|
1041 | >>> m2.visitdir(b'd/e') | |
1052 | True |
|
1042 | True | |
1053 | >>> m2.visitdir(b'd/e/f') |
|
1043 | >>> m2.visitdir(b'd/e/f') | |
1054 | True |
|
1044 | True | |
1055 | >>> m2.visitdir(b'd/e/g') |
|
1045 | >>> m2.visitdir(b'd/e/g') | |
1056 | False |
|
1046 | False | |
1057 | >>> m2.visitdir(b'd/ef') |
|
1047 | >>> m2.visitdir(b'd/ef') | |
1058 | False |
|
1048 | False | |
1059 | """ |
|
1049 | """ | |
1060 |
|
1050 | |||
1061 | def __init__(self, path, matcher, badfn=None): |
|
1051 | def __init__(self, path, matcher, badfn=None): | |
1062 | super(prefixdirmatcher, self).__init__(badfn) |
|
1052 | super(prefixdirmatcher, self).__init__(badfn) | |
1063 | if not path: |
|
1053 | if not path: | |
1064 | raise error.ProgrammingError(b'prefix path must not be empty') |
|
1054 | raise error.ProgrammingError(b'prefix path must not be empty') | |
1065 | self._path = path |
|
1055 | self._path = path | |
1066 | self._pathprefix = path + b'/' |
|
1056 | self._pathprefix = path + b'/' | |
1067 | self._matcher = matcher |
|
1057 | self._matcher = matcher | |
1068 |
|
1058 | |||
1069 | @propertycache |
|
1059 | @propertycache | |
1070 | def _files(self): |
|
1060 | def _files(self): | |
1071 | return [self._pathprefix + f for f in self._matcher._files] |
|
1061 | return [self._pathprefix + f for f in self._matcher._files] | |
1072 |
|
1062 | |||
1073 | def matchfn(self, f): |
|
1063 | def matchfn(self, f): | |
1074 | if not f.startswith(self._pathprefix): |
|
1064 | if not f.startswith(self._pathprefix): | |
1075 | return False |
|
1065 | return False | |
1076 | return self._matcher.matchfn(f[len(self._pathprefix) :]) |
|
1066 | return self._matcher.matchfn(f[len(self._pathprefix) :]) | |
1077 |
|
1067 | |||
1078 | @propertycache |
|
1068 | @propertycache | |
1079 | def _pathdirs(self): |
|
1069 | def _pathdirs(self): | |
1080 | return set(pathutil.finddirs(self._path)) |
|
1070 | return set(pathutil.finddirs(self._path)) | |
1081 |
|
1071 | |||
1082 | def visitdir(self, dir): |
|
1072 | def visitdir(self, dir): | |
1083 | if dir == self._path: |
|
1073 | if dir == self._path: | |
1084 | return self._matcher.visitdir(b'') |
|
1074 | return self._matcher.visitdir(b'') | |
1085 | if dir.startswith(self._pathprefix): |
|
1075 | if dir.startswith(self._pathprefix): | |
1086 | return self._matcher.visitdir(dir[len(self._pathprefix) :]) |
|
1076 | return self._matcher.visitdir(dir[len(self._pathprefix) :]) | |
1087 | return dir in self._pathdirs |
|
1077 | return dir in self._pathdirs | |
1088 |
|
1078 | |||
1089 | def visitchildrenset(self, dir): |
|
1079 | def visitchildrenset(self, dir): | |
1090 | if dir == self._path: |
|
1080 | if dir == self._path: | |
1091 | return self._matcher.visitchildrenset(b'') |
|
1081 | return self._matcher.visitchildrenset(b'') | |
1092 | if dir.startswith(self._pathprefix): |
|
1082 | if dir.startswith(self._pathprefix): | |
1093 | return self._matcher.visitchildrenset(dir[len(self._pathprefix) :]) |
|
1083 | return self._matcher.visitchildrenset(dir[len(self._pathprefix) :]) | |
1094 | if dir in self._pathdirs: |
|
1084 | if dir in self._pathdirs: | |
1095 | return b'this' |
|
1085 | return b'this' | |
1096 | return set() |
|
1086 | return set() | |
1097 |
|
1087 | |||
1098 | def isexact(self): |
|
1088 | def isexact(self): | |
1099 | return self._matcher.isexact() |
|
1089 | return self._matcher.isexact() | |
1100 |
|
1090 | |||
1101 | def prefix(self): |
|
1091 | def prefix(self): | |
1102 | return self._matcher.prefix() |
|
1092 | return self._matcher.prefix() | |
1103 |
|
1093 | |||
1104 | @encoding.strmethod |
|
1094 | @encoding.strmethod | |
1105 | def __repr__(self): |
|
1095 | def __repr__(self): | |
1106 | return b'<prefixdirmatcher path=%r, matcher=%r>' % ( |
|
1096 | return b'<prefixdirmatcher path=%r, matcher=%r>' % ( | |
1107 | pycompat.bytestr(self._path), |
|
1097 | pycompat.bytestr(self._path), | |
1108 | self._matcher, |
|
1098 | self._matcher, | |
1109 | ) |
|
1099 | ) | |
1110 |
|
1100 | |||
1111 |
|
1101 | |||
1112 | class unionmatcher(basematcher): |
|
1102 | class unionmatcher(basematcher): | |
1113 | """A matcher that is the union of several matchers. |
|
1103 | """A matcher that is the union of several matchers. | |
1114 |
|
1104 | |||
1115 |
The non-matching-attributes (bad, |
|
1105 | The non-matching-attributes (bad, traversedir) are taken from the first | |
1116 |
|
|
1106 | matcher. | |
1117 | """ |
|
1107 | """ | |
1118 |
|
1108 | |||
1119 | def __init__(self, matchers): |
|
1109 | def __init__(self, matchers): | |
1120 | m1 = matchers[0] |
|
1110 | m1 = matchers[0] | |
1121 | super(unionmatcher, self).__init__() |
|
1111 | super(unionmatcher, self).__init__() | |
1122 | self.explicitdir = m1.explicitdir |
|
|||
1123 | self.traversedir = m1.traversedir |
|
1112 | self.traversedir = m1.traversedir | |
1124 | self._matchers = matchers |
|
1113 | self._matchers = matchers | |
1125 |
|
1114 | |||
1126 | def matchfn(self, f): |
|
1115 | def matchfn(self, f): | |
1127 | for match in self._matchers: |
|
1116 | for match in self._matchers: | |
1128 | if match(f): |
|
1117 | if match(f): | |
1129 | return True |
|
1118 | return True | |
1130 | return False |
|
1119 | return False | |
1131 |
|
1120 | |||
1132 | def visitdir(self, dir): |
|
1121 | def visitdir(self, dir): | |
1133 | r = False |
|
1122 | r = False | |
1134 | for m in self._matchers: |
|
1123 | for m in self._matchers: | |
1135 | v = m.visitdir(dir) |
|
1124 | v = m.visitdir(dir) | |
1136 | if v == b'all': |
|
1125 | if v == b'all': | |
1137 | return v |
|
1126 | return v | |
1138 | r |= v |
|
1127 | r |= v | |
1139 | return r |
|
1128 | return r | |
1140 |
|
1129 | |||
1141 | def visitchildrenset(self, dir): |
|
1130 | def visitchildrenset(self, dir): | |
1142 | r = set() |
|
1131 | r = set() | |
1143 | this = False |
|
1132 | this = False | |
1144 | for m in self._matchers: |
|
1133 | for m in self._matchers: | |
1145 | v = m.visitchildrenset(dir) |
|
1134 | v = m.visitchildrenset(dir) | |
1146 | if not v: |
|
1135 | if not v: | |
1147 | continue |
|
1136 | continue | |
1148 | if v == b'all': |
|
1137 | if v == b'all': | |
1149 | return v |
|
1138 | return v | |
1150 | if this or v == b'this': |
|
1139 | if this or v == b'this': | |
1151 | this = True |
|
1140 | this = True | |
1152 | # don't break, we might have an 'all' in here. |
|
1141 | # don't break, we might have an 'all' in here. | |
1153 | continue |
|
1142 | continue | |
1154 | assert isinstance(v, set) |
|
1143 | assert isinstance(v, set) | |
1155 | r = r.union(v) |
|
1144 | r = r.union(v) | |
1156 | if this: |
|
1145 | if this: | |
1157 | return b'this' |
|
1146 | return b'this' | |
1158 | return r |
|
1147 | return r | |
1159 |
|
1148 | |||
1160 | @encoding.strmethod |
|
1149 | @encoding.strmethod | |
1161 | def __repr__(self): |
|
1150 | def __repr__(self): | |
1162 | return b'<unionmatcher matchers=%r>' % self._matchers |
|
1151 | return b'<unionmatcher matchers=%r>' % self._matchers | |
1163 |
|
1152 | |||
1164 |
|
1153 | |||
1165 | def patkind(pattern, default=None): |
|
1154 | def patkind(pattern, default=None): | |
1166 | '''If pattern is 'kind:pat' with a known kind, return kind. |
|
1155 | '''If pattern is 'kind:pat' with a known kind, return kind. | |
1167 |
|
1156 | |||
1168 | >>> patkind(br're:.*\.c$') |
|
1157 | >>> patkind(br're:.*\.c$') | |
1169 | 're' |
|
1158 | 're' | |
1170 | >>> patkind(b'glob:*.c') |
|
1159 | >>> patkind(b'glob:*.c') | |
1171 | 'glob' |
|
1160 | 'glob' | |
1172 | >>> patkind(b'relpath:test.py') |
|
1161 | >>> patkind(b'relpath:test.py') | |
1173 | 'relpath' |
|
1162 | 'relpath' | |
1174 | >>> patkind(b'main.py') |
|
1163 | >>> patkind(b'main.py') | |
1175 | >>> patkind(b'main.py', default=b're') |
|
1164 | >>> patkind(b'main.py', default=b're') | |
1176 | 're' |
|
1165 | 're' | |
1177 | ''' |
|
1166 | ''' | |
1178 | return _patsplit(pattern, default)[0] |
|
1167 | return _patsplit(pattern, default)[0] | |
1179 |
|
1168 | |||
1180 |
|
1169 | |||
1181 | def _patsplit(pattern, default): |
|
1170 | def _patsplit(pattern, default): | |
1182 | """Split a string into the optional pattern kind prefix and the actual |
|
1171 | """Split a string into the optional pattern kind prefix and the actual | |
1183 | pattern.""" |
|
1172 | pattern.""" | |
1184 | if b':' in pattern: |
|
1173 | if b':' in pattern: | |
1185 | kind, pat = pattern.split(b':', 1) |
|
1174 | kind, pat = pattern.split(b':', 1) | |
1186 | if kind in allpatternkinds: |
|
1175 | if kind in allpatternkinds: | |
1187 | return kind, pat |
|
1176 | return kind, pat | |
1188 | return default, pattern |
|
1177 | return default, pattern | |
1189 |
|
1178 | |||
1190 |
|
1179 | |||
1191 | def _globre(pat): |
|
1180 | def _globre(pat): | |
1192 | r'''Convert an extended glob string to a regexp string. |
|
1181 | r'''Convert an extended glob string to a regexp string. | |
1193 |
|
1182 | |||
1194 | >>> from . import pycompat |
|
1183 | >>> from . import pycompat | |
1195 | >>> def bprint(s): |
|
1184 | >>> def bprint(s): | |
1196 | ... print(pycompat.sysstr(s)) |
|
1185 | ... print(pycompat.sysstr(s)) | |
1197 | >>> bprint(_globre(br'?')) |
|
1186 | >>> bprint(_globre(br'?')) | |
1198 | . |
|
1187 | . | |
1199 | >>> bprint(_globre(br'*')) |
|
1188 | >>> bprint(_globre(br'*')) | |
1200 | [^/]* |
|
1189 | [^/]* | |
1201 | >>> bprint(_globre(br'**')) |
|
1190 | >>> bprint(_globre(br'**')) | |
1202 | .* |
|
1191 | .* | |
1203 | >>> bprint(_globre(br'**/a')) |
|
1192 | >>> bprint(_globre(br'**/a')) | |
1204 | (?:.*/)?a |
|
1193 | (?:.*/)?a | |
1205 | >>> bprint(_globre(br'a/**/b')) |
|
1194 | >>> bprint(_globre(br'a/**/b')) | |
1206 | a/(?:.*/)?b |
|
1195 | a/(?:.*/)?b | |
1207 | >>> bprint(_globre(br'[a*?!^][^b][!c]')) |
|
1196 | >>> bprint(_globre(br'[a*?!^][^b][!c]')) | |
1208 | [a*?!^][\^b][^c] |
|
1197 | [a*?!^][\^b][^c] | |
1209 | >>> bprint(_globre(br'{a,b}')) |
|
1198 | >>> bprint(_globre(br'{a,b}')) | |
1210 | (?:a|b) |
|
1199 | (?:a|b) | |
1211 | >>> bprint(_globre(br'.\*\?')) |
|
1200 | >>> bprint(_globre(br'.\*\?')) | |
1212 | \.\*\? |
|
1201 | \.\*\? | |
1213 | ''' |
|
1202 | ''' | |
1214 | i, n = 0, len(pat) |
|
1203 | i, n = 0, len(pat) | |
1215 | res = b'' |
|
1204 | res = b'' | |
1216 | group = 0 |
|
1205 | group = 0 | |
1217 | escape = util.stringutil.regexbytesescapemap.get |
|
1206 | escape = util.stringutil.regexbytesescapemap.get | |
1218 |
|
1207 | |||
1219 | def peek(): |
|
1208 | def peek(): | |
1220 | return i < n and pat[i : i + 1] |
|
1209 | return i < n and pat[i : i + 1] | |
1221 |
|
1210 | |||
1222 | while i < n: |
|
1211 | while i < n: | |
1223 | c = pat[i : i + 1] |
|
1212 | c = pat[i : i + 1] | |
1224 | i += 1 |
|
1213 | i += 1 | |
1225 | if c not in b'*?[{},\\': |
|
1214 | if c not in b'*?[{},\\': | |
1226 | res += escape(c, c) |
|
1215 | res += escape(c, c) | |
1227 | elif c == b'*': |
|
1216 | elif c == b'*': | |
1228 | if peek() == b'*': |
|
1217 | if peek() == b'*': | |
1229 | i += 1 |
|
1218 | i += 1 | |
1230 | if peek() == b'/': |
|
1219 | if peek() == b'/': | |
1231 | i += 1 |
|
1220 | i += 1 | |
1232 | res += b'(?:.*/)?' |
|
1221 | res += b'(?:.*/)?' | |
1233 | else: |
|
1222 | else: | |
1234 | res += b'.*' |
|
1223 | res += b'.*' | |
1235 | else: |
|
1224 | else: | |
1236 | res += b'[^/]*' |
|
1225 | res += b'[^/]*' | |
1237 | elif c == b'?': |
|
1226 | elif c == b'?': | |
1238 | res += b'.' |
|
1227 | res += b'.' | |
1239 | elif c == b'[': |
|
1228 | elif c == b'[': | |
1240 | j = i |
|
1229 | j = i | |
1241 | if j < n and pat[j : j + 1] in b'!]': |
|
1230 | if j < n and pat[j : j + 1] in b'!]': | |
1242 | j += 1 |
|
1231 | j += 1 | |
1243 | while j < n and pat[j : j + 1] != b']': |
|
1232 | while j < n and pat[j : j + 1] != b']': | |
1244 | j += 1 |
|
1233 | j += 1 | |
1245 | if j >= n: |
|
1234 | if j >= n: | |
1246 | res += b'\\[' |
|
1235 | res += b'\\[' | |
1247 | else: |
|
1236 | else: | |
1248 | stuff = pat[i:j].replace(b'\\', b'\\\\') |
|
1237 | stuff = pat[i:j].replace(b'\\', b'\\\\') | |
1249 | i = j + 1 |
|
1238 | i = j + 1 | |
1250 | if stuff[0:1] == b'!': |
|
1239 | if stuff[0:1] == b'!': | |
1251 | stuff = b'^' + stuff[1:] |
|
1240 | stuff = b'^' + stuff[1:] | |
1252 | elif stuff[0:1] == b'^': |
|
1241 | elif stuff[0:1] == b'^': | |
1253 | stuff = b'\\' + stuff |
|
1242 | stuff = b'\\' + stuff | |
1254 | res = b'%s[%s]' % (res, stuff) |
|
1243 | res = b'%s[%s]' % (res, stuff) | |
1255 | elif c == b'{': |
|
1244 | elif c == b'{': | |
1256 | group += 1 |
|
1245 | group += 1 | |
1257 | res += b'(?:' |
|
1246 | res += b'(?:' | |
1258 | elif c == b'}' and group: |
|
1247 | elif c == b'}' and group: | |
1259 | res += b')' |
|
1248 | res += b')' | |
1260 | group -= 1 |
|
1249 | group -= 1 | |
1261 | elif c == b',' and group: |
|
1250 | elif c == b',' and group: | |
1262 | res += b'|' |
|
1251 | res += b'|' | |
1263 | elif c == b'\\': |
|
1252 | elif c == b'\\': | |
1264 | p = peek() |
|
1253 | p = peek() | |
1265 | if p: |
|
1254 | if p: | |
1266 | i += 1 |
|
1255 | i += 1 | |
1267 | res += escape(p, p) |
|
1256 | res += escape(p, p) | |
1268 | else: |
|
1257 | else: | |
1269 | res += escape(c, c) |
|
1258 | res += escape(c, c) | |
1270 | else: |
|
1259 | else: | |
1271 | res += escape(c, c) |
|
1260 | res += escape(c, c) | |
1272 | return res |
|
1261 | return res | |
1273 |
|
1262 | |||
1274 |
|
1263 | |||
1275 | def _regex(kind, pat, globsuffix): |
|
1264 | def _regex(kind, pat, globsuffix): | |
1276 | '''Convert a (normalized) pattern of any kind into a |
|
1265 | '''Convert a (normalized) pattern of any kind into a | |
1277 | regular expression. |
|
1266 | regular expression. | |
1278 | globsuffix is appended to the regexp of globs.''' |
|
1267 | globsuffix is appended to the regexp of globs.''' | |
1279 |
|
1268 | |||
1280 | if rustmod is not None: |
|
1269 | if rustmod is not None: | |
1281 | try: |
|
1270 | try: | |
1282 | return rustmod.build_single_regex(kind, pat, globsuffix) |
|
1271 | return rustmod.build_single_regex(kind, pat, globsuffix) | |
1283 | except rustmod.PatternError: |
|
1272 | except rustmod.PatternError: | |
1284 | raise error.ProgrammingError( |
|
1273 | raise error.ProgrammingError( | |
1285 | b'not a regex pattern: %s:%s' % (kind, pat) |
|
1274 | b'not a regex pattern: %s:%s' % (kind, pat) | |
1286 | ) |
|
1275 | ) | |
1287 |
|
1276 | |||
1288 | if not pat and kind in (b'glob', b'relpath'): |
|
1277 | if not pat and kind in (b'glob', b'relpath'): | |
1289 | return b'' |
|
1278 | return b'' | |
1290 | if kind == b're': |
|
1279 | if kind == b're': | |
1291 | return pat |
|
1280 | return pat | |
1292 | if kind in (b'path', b'relpath'): |
|
1281 | if kind in (b'path', b'relpath'): | |
1293 | if pat == b'.': |
|
1282 | if pat == b'.': | |
1294 | return b'' |
|
1283 | return b'' | |
1295 | return util.stringutil.reescape(pat) + b'(?:/|$)' |
|
1284 | return util.stringutil.reescape(pat) + b'(?:/|$)' | |
1296 | if kind == b'rootfilesin': |
|
1285 | if kind == b'rootfilesin': | |
1297 | if pat == b'.': |
|
1286 | if pat == b'.': | |
1298 | escaped = b'' |
|
1287 | escaped = b'' | |
1299 | else: |
|
1288 | else: | |
1300 | # Pattern is a directory name. |
|
1289 | # Pattern is a directory name. | |
1301 | escaped = util.stringutil.reescape(pat) + b'/' |
|
1290 | escaped = util.stringutil.reescape(pat) + b'/' | |
1302 | # Anything after the pattern must be a non-directory. |
|
1291 | # Anything after the pattern must be a non-directory. | |
1303 | return escaped + b'[^/]+$' |
|
1292 | return escaped + b'[^/]+$' | |
1304 | if kind == b'relglob': |
|
1293 | if kind == b'relglob': | |
1305 | globre = _globre(pat) |
|
1294 | globre = _globre(pat) | |
1306 | if globre.startswith(b'[^/]*'): |
|
1295 | if globre.startswith(b'[^/]*'): | |
1307 | # When pat has the form *XYZ (common), make the returned regex more |
|
1296 | # When pat has the form *XYZ (common), make the returned regex more | |
1308 | # legible by returning the regex for **XYZ instead of **/*XYZ. |
|
1297 | # legible by returning the regex for **XYZ instead of **/*XYZ. | |
1309 | return b'.*' + globre[len(b'[^/]*') :] + globsuffix |
|
1298 | return b'.*' + globre[len(b'[^/]*') :] + globsuffix | |
1310 | return b'(?:|.*/)' + globre + globsuffix |
|
1299 | return b'(?:|.*/)' + globre + globsuffix | |
1311 | if kind == b'relre': |
|
1300 | if kind == b'relre': | |
1312 | if pat.startswith(b'^'): |
|
1301 | if pat.startswith(b'^'): | |
1313 | return pat |
|
1302 | return pat | |
1314 | return b'.*' + pat |
|
1303 | return b'.*' + pat | |
1315 | if kind in (b'glob', b'rootglob'): |
|
1304 | if kind in (b'glob', b'rootglob'): | |
1316 | return _globre(pat) + globsuffix |
|
1305 | return _globre(pat) + globsuffix | |
1317 | raise error.ProgrammingError(b'not a regex pattern: %s:%s' % (kind, pat)) |
|
1306 | raise error.ProgrammingError(b'not a regex pattern: %s:%s' % (kind, pat)) | |
1318 |
|
1307 | |||
1319 |
|
1308 | |||
1320 | def _buildmatch(kindpats, globsuffix, root): |
|
1309 | def _buildmatch(kindpats, globsuffix, root): | |
1321 | '''Return regexp string and a matcher function for kindpats. |
|
1310 | '''Return regexp string and a matcher function for kindpats. | |
1322 | globsuffix is appended to the regexp of globs.''' |
|
1311 | globsuffix is appended to the regexp of globs.''' | |
1323 | matchfuncs = [] |
|
1312 | matchfuncs = [] | |
1324 |
|
1313 | |||
1325 | subincludes, kindpats = _expandsubinclude(kindpats, root) |
|
1314 | subincludes, kindpats = _expandsubinclude(kindpats, root) | |
1326 | if subincludes: |
|
1315 | if subincludes: | |
1327 | submatchers = {} |
|
1316 | submatchers = {} | |
1328 |
|
1317 | |||
1329 | def matchsubinclude(f): |
|
1318 | def matchsubinclude(f): | |
1330 | for prefix, matcherargs in subincludes: |
|
1319 | for prefix, matcherargs in subincludes: | |
1331 | if f.startswith(prefix): |
|
1320 | if f.startswith(prefix): | |
1332 | mf = submatchers.get(prefix) |
|
1321 | mf = submatchers.get(prefix) | |
1333 | if mf is None: |
|
1322 | if mf is None: | |
1334 | mf = match(*matcherargs) |
|
1323 | mf = match(*matcherargs) | |
1335 | submatchers[prefix] = mf |
|
1324 | submatchers[prefix] = mf | |
1336 |
|
1325 | |||
1337 | if mf(f[len(prefix) :]): |
|
1326 | if mf(f[len(prefix) :]): | |
1338 | return True |
|
1327 | return True | |
1339 | return False |
|
1328 | return False | |
1340 |
|
1329 | |||
1341 | matchfuncs.append(matchsubinclude) |
|
1330 | matchfuncs.append(matchsubinclude) | |
1342 |
|
1331 | |||
1343 | regex = b'' |
|
1332 | regex = b'' | |
1344 | if kindpats: |
|
1333 | if kindpats: | |
1345 | if all(k == b'rootfilesin' for k, p, s in kindpats): |
|
1334 | if all(k == b'rootfilesin' for k, p, s in kindpats): | |
1346 | dirs = {p for k, p, s in kindpats} |
|
1335 | dirs = {p for k, p, s in kindpats} | |
1347 |
|
1336 | |||
1348 | def mf(f): |
|
1337 | def mf(f): | |
1349 | i = f.rfind(b'/') |
|
1338 | i = f.rfind(b'/') | |
1350 | if i >= 0: |
|
1339 | if i >= 0: | |
1351 | dir = f[:i] |
|
1340 | dir = f[:i] | |
1352 | else: |
|
1341 | else: | |
1353 | dir = b'.' |
|
1342 | dir = b'.' | |
1354 | return dir in dirs |
|
1343 | return dir in dirs | |
1355 |
|
1344 | |||
1356 | regex = b'rootfilesin: %s' % stringutil.pprint(list(sorted(dirs))) |
|
1345 | regex = b'rootfilesin: %s' % stringutil.pprint(list(sorted(dirs))) | |
1357 | matchfuncs.append(mf) |
|
1346 | matchfuncs.append(mf) | |
1358 | else: |
|
1347 | else: | |
1359 | regex, mf = _buildregexmatch(kindpats, globsuffix) |
|
1348 | regex, mf = _buildregexmatch(kindpats, globsuffix) | |
1360 | matchfuncs.append(mf) |
|
1349 | matchfuncs.append(mf) | |
1361 |
|
1350 | |||
1362 | if len(matchfuncs) == 1: |
|
1351 | if len(matchfuncs) == 1: | |
1363 | return regex, matchfuncs[0] |
|
1352 | return regex, matchfuncs[0] | |
1364 | else: |
|
1353 | else: | |
1365 | return regex, lambda f: any(mf(f) for mf in matchfuncs) |
|
1354 | return regex, lambda f: any(mf(f) for mf in matchfuncs) | |
1366 |
|
1355 | |||
1367 |
|
1356 | |||
1368 | MAX_RE_SIZE = 20000 |
|
1357 | MAX_RE_SIZE = 20000 | |
1369 |
|
1358 | |||
1370 |
|
1359 | |||
1371 | def _joinregexes(regexps): |
|
1360 | def _joinregexes(regexps): | |
1372 | """gather multiple regular expressions into a single one""" |
|
1361 | """gather multiple regular expressions into a single one""" | |
1373 | return b'|'.join(regexps) |
|
1362 | return b'|'.join(regexps) | |
1374 |
|
1363 | |||
1375 |
|
1364 | |||
1376 | def _buildregexmatch(kindpats, globsuffix): |
|
1365 | def _buildregexmatch(kindpats, globsuffix): | |
1377 | """Build a match function from a list of kinds and kindpats, |
|
1366 | """Build a match function from a list of kinds and kindpats, | |
1378 | return regexp string and a matcher function. |
|
1367 | return regexp string and a matcher function. | |
1379 |
|
1368 | |||
1380 | Test too large input |
|
1369 | Test too large input | |
1381 | >>> _buildregexmatch([ |
|
1370 | >>> _buildregexmatch([ | |
1382 | ... (b'relglob', b'?' * MAX_RE_SIZE, b'') |
|
1371 | ... (b'relglob', b'?' * MAX_RE_SIZE, b'') | |
1383 | ... ], b'$') |
|
1372 | ... ], b'$') | |
1384 | Traceback (most recent call last): |
|
1373 | Traceback (most recent call last): | |
1385 | ... |
|
1374 | ... | |
1386 | Abort: matcher pattern is too long (20009 bytes) |
|
1375 | Abort: matcher pattern is too long (20009 bytes) | |
1387 | """ |
|
1376 | """ | |
1388 | try: |
|
1377 | try: | |
1389 | allgroups = [] |
|
1378 | allgroups = [] | |
1390 | regexps = [_regex(k, p, globsuffix) for (k, p, s) in kindpats] |
|
1379 | regexps = [_regex(k, p, globsuffix) for (k, p, s) in kindpats] | |
1391 | fullregexp = _joinregexes(regexps) |
|
1380 | fullregexp = _joinregexes(regexps) | |
1392 |
|
1381 | |||
1393 | startidx = 0 |
|
1382 | startidx = 0 | |
1394 | groupsize = 0 |
|
1383 | groupsize = 0 | |
1395 | for idx, r in enumerate(regexps): |
|
1384 | for idx, r in enumerate(regexps): | |
1396 | piecesize = len(r) |
|
1385 | piecesize = len(r) | |
1397 | if piecesize > MAX_RE_SIZE: |
|
1386 | if piecesize > MAX_RE_SIZE: | |
1398 | msg = _(b"matcher pattern is too long (%d bytes)") % piecesize |
|
1387 | msg = _(b"matcher pattern is too long (%d bytes)") % piecesize | |
1399 | raise error.Abort(msg) |
|
1388 | raise error.Abort(msg) | |
1400 | elif (groupsize + piecesize) > MAX_RE_SIZE: |
|
1389 | elif (groupsize + piecesize) > MAX_RE_SIZE: | |
1401 | group = regexps[startidx:idx] |
|
1390 | group = regexps[startidx:idx] | |
1402 | allgroups.append(_joinregexes(group)) |
|
1391 | allgroups.append(_joinregexes(group)) | |
1403 | startidx = idx |
|
1392 | startidx = idx | |
1404 | groupsize = 0 |
|
1393 | groupsize = 0 | |
1405 | groupsize += piecesize + 1 |
|
1394 | groupsize += piecesize + 1 | |
1406 |
|
1395 | |||
1407 | if startidx == 0: |
|
1396 | if startidx == 0: | |
1408 | matcher = _rematcher(fullregexp) |
|
1397 | matcher = _rematcher(fullregexp) | |
1409 | func = lambda s: bool(matcher(s)) |
|
1398 | func = lambda s: bool(matcher(s)) | |
1410 | else: |
|
1399 | else: | |
1411 | group = regexps[startidx:] |
|
1400 | group = regexps[startidx:] | |
1412 | allgroups.append(_joinregexes(group)) |
|
1401 | allgroups.append(_joinregexes(group)) | |
1413 | allmatchers = [_rematcher(g) for g in allgroups] |
|
1402 | allmatchers = [_rematcher(g) for g in allgroups] | |
1414 | func = lambda s: any(m(s) for m in allmatchers) |
|
1403 | func = lambda s: any(m(s) for m in allmatchers) | |
1415 | return fullregexp, func |
|
1404 | return fullregexp, func | |
1416 | except re.error: |
|
1405 | except re.error: | |
1417 | for k, p, s in kindpats: |
|
1406 | for k, p, s in kindpats: | |
1418 | try: |
|
1407 | try: | |
1419 | _rematcher(_regex(k, p, globsuffix)) |
|
1408 | _rematcher(_regex(k, p, globsuffix)) | |
1420 | except re.error: |
|
1409 | except re.error: | |
1421 | if s: |
|
1410 | if s: | |
1422 | raise error.Abort( |
|
1411 | raise error.Abort( | |
1423 | _(b"%s: invalid pattern (%s): %s") % (s, k, p) |
|
1412 | _(b"%s: invalid pattern (%s): %s") % (s, k, p) | |
1424 | ) |
|
1413 | ) | |
1425 | else: |
|
1414 | else: | |
1426 | raise error.Abort(_(b"invalid pattern (%s): %s") % (k, p)) |
|
1415 | raise error.Abort(_(b"invalid pattern (%s): %s") % (k, p)) | |
1427 | raise error.Abort(_(b"invalid pattern")) |
|
1416 | raise error.Abort(_(b"invalid pattern")) | |
1428 |
|
1417 | |||
1429 |
|
1418 | |||
1430 | def _patternrootsanddirs(kindpats): |
|
1419 | def _patternrootsanddirs(kindpats): | |
1431 | '''Returns roots and directories corresponding to each pattern. |
|
1420 | '''Returns roots and directories corresponding to each pattern. | |
1432 |
|
1421 | |||
1433 | This calculates the roots and directories exactly matching the patterns and |
|
1422 | This calculates the roots and directories exactly matching the patterns and | |
1434 | returns a tuple of (roots, dirs) for each. It does not return other |
|
1423 | returns a tuple of (roots, dirs) for each. It does not return other | |
1435 | directories which may also need to be considered, like the parent |
|
1424 | directories which may also need to be considered, like the parent | |
1436 | directories. |
|
1425 | directories. | |
1437 | ''' |
|
1426 | ''' | |
1438 | r = [] |
|
1427 | r = [] | |
1439 | d = [] |
|
1428 | d = [] | |
1440 | for kind, pat, source in kindpats: |
|
1429 | for kind, pat, source in kindpats: | |
1441 | if kind in (b'glob', b'rootglob'): # find the non-glob prefix |
|
1430 | if kind in (b'glob', b'rootglob'): # find the non-glob prefix | |
1442 | root = [] |
|
1431 | root = [] | |
1443 | for p in pat.split(b'/'): |
|
1432 | for p in pat.split(b'/'): | |
1444 | if b'[' in p or b'{' in p or b'*' in p or b'?' in p: |
|
1433 | if b'[' in p or b'{' in p or b'*' in p or b'?' in p: | |
1445 | break |
|
1434 | break | |
1446 | root.append(p) |
|
1435 | root.append(p) | |
1447 | r.append(b'/'.join(root)) |
|
1436 | r.append(b'/'.join(root)) | |
1448 | elif kind in (b'relpath', b'path'): |
|
1437 | elif kind in (b'relpath', b'path'): | |
1449 | if pat == b'.': |
|
1438 | if pat == b'.': | |
1450 | pat = b'' |
|
1439 | pat = b'' | |
1451 | r.append(pat) |
|
1440 | r.append(pat) | |
1452 | elif kind in (b'rootfilesin',): |
|
1441 | elif kind in (b'rootfilesin',): | |
1453 | if pat == b'.': |
|
1442 | if pat == b'.': | |
1454 | pat = b'' |
|
1443 | pat = b'' | |
1455 | d.append(pat) |
|
1444 | d.append(pat) | |
1456 | else: # relglob, re, relre |
|
1445 | else: # relglob, re, relre | |
1457 | r.append(b'') |
|
1446 | r.append(b'') | |
1458 | return r, d |
|
1447 | return r, d | |
1459 |
|
1448 | |||
1460 |
|
1449 | |||
1461 | def _roots(kindpats): |
|
1450 | def _roots(kindpats): | |
1462 | '''Returns root directories to match recursively from the given patterns.''' |
|
1451 | '''Returns root directories to match recursively from the given patterns.''' | |
1463 | roots, dirs = _patternrootsanddirs(kindpats) |
|
1452 | roots, dirs = _patternrootsanddirs(kindpats) | |
1464 | return roots |
|
1453 | return roots | |
1465 |
|
1454 | |||
1466 |
|
1455 | |||
1467 | def _rootsdirsandparents(kindpats): |
|
1456 | def _rootsdirsandparents(kindpats): | |
1468 | '''Returns roots and exact directories from patterns. |
|
1457 | '''Returns roots and exact directories from patterns. | |
1469 |
|
1458 | |||
1470 | `roots` are directories to match recursively, `dirs` should |
|
1459 | `roots` are directories to match recursively, `dirs` should | |
1471 | be matched non-recursively, and `parents` are the implicitly required |
|
1460 | be matched non-recursively, and `parents` are the implicitly required | |
1472 | directories to walk to items in either roots or dirs. |
|
1461 | directories to walk to items in either roots or dirs. | |
1473 |
|
1462 | |||
1474 | Returns a tuple of (roots, dirs, parents). |
|
1463 | Returns a tuple of (roots, dirs, parents). | |
1475 |
|
1464 | |||
1476 | >>> r = _rootsdirsandparents( |
|
1465 | >>> r = _rootsdirsandparents( | |
1477 | ... [(b'glob', b'g/h/*', b''), (b'glob', b'g/h', b''), |
|
1466 | ... [(b'glob', b'g/h/*', b''), (b'glob', b'g/h', b''), | |
1478 | ... (b'glob', b'g*', b'')]) |
|
1467 | ... (b'glob', b'g*', b'')]) | |
1479 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1468 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1480 | (['g/h', 'g/h', ''], []) ['', 'g'] |
|
1469 | (['g/h', 'g/h', ''], []) ['', 'g'] | |
1481 | >>> r = _rootsdirsandparents( |
|
1470 | >>> r = _rootsdirsandparents( | |
1482 | ... [(b'rootfilesin', b'g/h', b''), (b'rootfilesin', b'', b'')]) |
|
1471 | ... [(b'rootfilesin', b'g/h', b''), (b'rootfilesin', b'', b'')]) | |
1483 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1472 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1484 | ([], ['g/h', '']) ['', 'g'] |
|
1473 | ([], ['g/h', '']) ['', 'g'] | |
1485 | >>> r = _rootsdirsandparents( |
|
1474 | >>> r = _rootsdirsandparents( | |
1486 | ... [(b'relpath', b'r', b''), (b'path', b'p/p', b''), |
|
1475 | ... [(b'relpath', b'r', b''), (b'path', b'p/p', b''), | |
1487 | ... (b'path', b'', b'')]) |
|
1476 | ... (b'path', b'', b'')]) | |
1488 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1477 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1489 | (['r', 'p/p', ''], []) ['', 'p'] |
|
1478 | (['r', 'p/p', ''], []) ['', 'p'] | |
1490 | >>> r = _rootsdirsandparents( |
|
1479 | >>> r = _rootsdirsandparents( | |
1491 | ... [(b'relglob', b'rg*', b''), (b're', b're/', b''), |
|
1480 | ... [(b'relglob', b'rg*', b''), (b're', b're/', b''), | |
1492 | ... (b'relre', b'rr', b'')]) |
|
1481 | ... (b'relre', b'rr', b'')]) | |
1493 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1482 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1494 | (['', '', ''], []) [''] |
|
1483 | (['', '', ''], []) [''] | |
1495 | ''' |
|
1484 | ''' | |
1496 | r, d = _patternrootsanddirs(kindpats) |
|
1485 | r, d = _patternrootsanddirs(kindpats) | |
1497 |
|
1486 | |||
1498 | p = set() |
|
1487 | p = set() | |
1499 | # Add the parents as non-recursive/exact directories, since they must be |
|
1488 | # Add the parents as non-recursive/exact directories, since they must be | |
1500 | # scanned to get to either the roots or the other exact directories. |
|
1489 | # scanned to get to either the roots or the other exact directories. | |
1501 | p.update(pathutil.dirs(d)) |
|
1490 | p.update(pathutil.dirs(d)) | |
1502 | p.update(pathutil.dirs(r)) |
|
1491 | p.update(pathutil.dirs(r)) | |
1503 |
|
1492 | |||
1504 | # FIXME: all uses of this function convert these to sets, do so before |
|
1493 | # FIXME: all uses of this function convert these to sets, do so before | |
1505 | # returning. |
|
1494 | # returning. | |
1506 | # FIXME: all uses of this function do not need anything in 'roots' and |
|
1495 | # FIXME: all uses of this function do not need anything in 'roots' and | |
1507 | # 'dirs' to also be in 'parents', consider removing them before returning. |
|
1496 | # 'dirs' to also be in 'parents', consider removing them before returning. | |
1508 | return r, d, p |
|
1497 | return r, d, p | |
1509 |
|
1498 | |||
1510 |
|
1499 | |||
1511 | def _explicitfiles(kindpats): |
|
1500 | def _explicitfiles(kindpats): | |
1512 | '''Returns the potential explicit filenames from the patterns. |
|
1501 | '''Returns the potential explicit filenames from the patterns. | |
1513 |
|
1502 | |||
1514 | >>> _explicitfiles([(b'path', b'foo/bar', b'')]) |
|
1503 | >>> _explicitfiles([(b'path', b'foo/bar', b'')]) | |
1515 | ['foo/bar'] |
|
1504 | ['foo/bar'] | |
1516 | >>> _explicitfiles([(b'rootfilesin', b'foo/bar', b'')]) |
|
1505 | >>> _explicitfiles([(b'rootfilesin', b'foo/bar', b'')]) | |
1517 | [] |
|
1506 | [] | |
1518 | ''' |
|
1507 | ''' | |
1519 | # Keep only the pattern kinds where one can specify filenames (vs only |
|
1508 | # Keep only the pattern kinds where one can specify filenames (vs only | |
1520 | # directory names). |
|
1509 | # directory names). | |
1521 | filable = [kp for kp in kindpats if kp[0] not in (b'rootfilesin',)] |
|
1510 | filable = [kp for kp in kindpats if kp[0] not in (b'rootfilesin',)] | |
1522 | return _roots(filable) |
|
1511 | return _roots(filable) | |
1523 |
|
1512 | |||
1524 |
|
1513 | |||
1525 | def _prefix(kindpats): |
|
1514 | def _prefix(kindpats): | |
1526 | '''Whether all the patterns match a prefix (i.e. recursively)''' |
|
1515 | '''Whether all the patterns match a prefix (i.e. recursively)''' | |
1527 | for kind, pat, source in kindpats: |
|
1516 | for kind, pat, source in kindpats: | |
1528 | if kind not in (b'path', b'relpath'): |
|
1517 | if kind not in (b'path', b'relpath'): | |
1529 | return False |
|
1518 | return False | |
1530 | return True |
|
1519 | return True | |
1531 |
|
1520 | |||
1532 |
|
1521 | |||
1533 | _commentre = None |
|
1522 | _commentre = None | |
1534 |
|
1523 | |||
1535 |
|
1524 | |||
1536 | def readpatternfile(filepath, warn, sourceinfo=False): |
|
1525 | def readpatternfile(filepath, warn, sourceinfo=False): | |
1537 | '''parse a pattern file, returning a list of |
|
1526 | '''parse a pattern file, returning a list of | |
1538 | patterns. These patterns should be given to compile() |
|
1527 | patterns. These patterns should be given to compile() | |
1539 | to be validated and converted into a match function. |
|
1528 | to be validated and converted into a match function. | |
1540 |
|
1529 | |||
1541 | trailing white space is dropped. |
|
1530 | trailing white space is dropped. | |
1542 | the escape character is backslash. |
|
1531 | the escape character is backslash. | |
1543 | comments start with #. |
|
1532 | comments start with #. | |
1544 | empty lines are skipped. |
|
1533 | empty lines are skipped. | |
1545 |
|
1534 | |||
1546 | lines can be of the following formats: |
|
1535 | lines can be of the following formats: | |
1547 |
|
1536 | |||
1548 | syntax: regexp # defaults following lines to non-rooted regexps |
|
1537 | syntax: regexp # defaults following lines to non-rooted regexps | |
1549 | syntax: glob # defaults following lines to non-rooted globs |
|
1538 | syntax: glob # defaults following lines to non-rooted globs | |
1550 | re:pattern # non-rooted regular expression |
|
1539 | re:pattern # non-rooted regular expression | |
1551 | glob:pattern # non-rooted glob |
|
1540 | glob:pattern # non-rooted glob | |
1552 | rootglob:pat # rooted glob (same root as ^ in regexps) |
|
1541 | rootglob:pat # rooted glob (same root as ^ in regexps) | |
1553 | pattern # pattern of the current default type |
|
1542 | pattern # pattern of the current default type | |
1554 |
|
1543 | |||
1555 | if sourceinfo is set, returns a list of tuples: |
|
1544 | if sourceinfo is set, returns a list of tuples: | |
1556 | (pattern, lineno, originalline). |
|
1545 | (pattern, lineno, originalline). | |
1557 | This is useful to debug ignore patterns. |
|
1546 | This is useful to debug ignore patterns. | |
1558 | ''' |
|
1547 | ''' | |
1559 |
|
1548 | |||
1560 | if rustmod is not None: |
|
1549 | if rustmod is not None: | |
1561 | result, warnings = rustmod.read_pattern_file( |
|
1550 | result, warnings = rustmod.read_pattern_file( | |
1562 | filepath, bool(warn), sourceinfo, |
|
1551 | filepath, bool(warn), sourceinfo, | |
1563 | ) |
|
1552 | ) | |
1564 |
|
1553 | |||
1565 | for warning_params in warnings: |
|
1554 | for warning_params in warnings: | |
1566 | # Can't be easily emitted from Rust, because it would require |
|
1555 | # Can't be easily emitted from Rust, because it would require | |
1567 | # a mechanism for both gettext and calling the `warn` function. |
|
1556 | # a mechanism for both gettext and calling the `warn` function. | |
1568 | warn(_(b"%s: ignoring invalid syntax '%s'\n") % warning_params) |
|
1557 | warn(_(b"%s: ignoring invalid syntax '%s'\n") % warning_params) | |
1569 |
|
1558 | |||
1570 | return result |
|
1559 | return result | |
1571 |
|
1560 | |||
1572 | syntaxes = { |
|
1561 | syntaxes = { | |
1573 | b're': b'relre:', |
|
1562 | b're': b'relre:', | |
1574 | b'regexp': b'relre:', |
|
1563 | b'regexp': b'relre:', | |
1575 | b'glob': b'relglob:', |
|
1564 | b'glob': b'relglob:', | |
1576 | b'rootglob': b'rootglob:', |
|
1565 | b'rootglob': b'rootglob:', | |
1577 | b'include': b'include', |
|
1566 | b'include': b'include', | |
1578 | b'subinclude': b'subinclude', |
|
1567 | b'subinclude': b'subinclude', | |
1579 | } |
|
1568 | } | |
1580 | syntax = b'relre:' |
|
1569 | syntax = b'relre:' | |
1581 | patterns = [] |
|
1570 | patterns = [] | |
1582 |
|
1571 | |||
1583 | fp = open(filepath, b'rb') |
|
1572 | fp = open(filepath, b'rb') | |
1584 | for lineno, line in enumerate(util.iterfile(fp), start=1): |
|
1573 | for lineno, line in enumerate(util.iterfile(fp), start=1): | |
1585 | if b"#" in line: |
|
1574 | if b"#" in line: | |
1586 | global _commentre |
|
1575 | global _commentre | |
1587 | if not _commentre: |
|
1576 | if not _commentre: | |
1588 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') |
|
1577 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') | |
1589 | # remove comments prefixed by an even number of escapes |
|
1578 | # remove comments prefixed by an even number of escapes | |
1590 | m = _commentre.search(line) |
|
1579 | m = _commentre.search(line) | |
1591 | if m: |
|
1580 | if m: | |
1592 | line = line[: m.end(1)] |
|
1581 | line = line[: m.end(1)] | |
1593 | # fixup properly escaped comments that survived the above |
|
1582 | # fixup properly escaped comments that survived the above | |
1594 | line = line.replace(b"\\#", b"#") |
|
1583 | line = line.replace(b"\\#", b"#") | |
1595 | line = line.rstrip() |
|
1584 | line = line.rstrip() | |
1596 | if not line: |
|
1585 | if not line: | |
1597 | continue |
|
1586 | continue | |
1598 |
|
1587 | |||
1599 | if line.startswith(b'syntax:'): |
|
1588 | if line.startswith(b'syntax:'): | |
1600 | s = line[7:].strip() |
|
1589 | s = line[7:].strip() | |
1601 | try: |
|
1590 | try: | |
1602 | syntax = syntaxes[s] |
|
1591 | syntax = syntaxes[s] | |
1603 | except KeyError: |
|
1592 | except KeyError: | |
1604 | if warn: |
|
1593 | if warn: | |
1605 | warn( |
|
1594 | warn( | |
1606 | _(b"%s: ignoring invalid syntax '%s'\n") % (filepath, s) |
|
1595 | _(b"%s: ignoring invalid syntax '%s'\n") % (filepath, s) | |
1607 | ) |
|
1596 | ) | |
1608 | continue |
|
1597 | continue | |
1609 |
|
1598 | |||
1610 | linesyntax = syntax |
|
1599 | linesyntax = syntax | |
1611 | for s, rels in pycompat.iteritems(syntaxes): |
|
1600 | for s, rels in pycompat.iteritems(syntaxes): | |
1612 | if line.startswith(rels): |
|
1601 | if line.startswith(rels): | |
1613 | linesyntax = rels |
|
1602 | linesyntax = rels | |
1614 | line = line[len(rels) :] |
|
1603 | line = line[len(rels) :] | |
1615 | break |
|
1604 | break | |
1616 | elif line.startswith(s + b':'): |
|
1605 | elif line.startswith(s + b':'): | |
1617 | linesyntax = rels |
|
1606 | linesyntax = rels | |
1618 | line = line[len(s) + 1 :] |
|
1607 | line = line[len(s) + 1 :] | |
1619 | break |
|
1608 | break | |
1620 | if sourceinfo: |
|
1609 | if sourceinfo: | |
1621 | patterns.append((linesyntax + line, lineno, line)) |
|
1610 | patterns.append((linesyntax + line, lineno, line)) | |
1622 | else: |
|
1611 | else: | |
1623 | patterns.append(linesyntax + line) |
|
1612 | patterns.append(linesyntax + line) | |
1624 | fp.close() |
|
1613 | fp.close() | |
1625 | return patterns |
|
1614 | return patterns |
General Comments 0
You need to be logged in to leave comments.
Login now