Show More
@@ -1,1617 +1,1637 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('dirstate') |
|
27 | rustmod = policy.importrust('dirstate') | |
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(cwd, kindpats, ctx=None, listsubrepos=False, badfn=None): |
|
60 | def _expandsets(cwd, 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(cwd, pat, badfn=badfn)) |
|
71 | matchers.append(ctx.matchfileset(cwd, 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(cwd, pat, badfn=badfn) |
|
75 | sm = ctx.sub(subpath).matchfileset(cwd, 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, |
|
120 | matchercls, | |
121 | root, |
|
121 | root, | |
122 | cwd, |
|
122 | cwd, | |
123 | kindpats, |
|
123 | kindpats, | |
124 | ctx=None, |
|
124 | ctx=None, | |
125 | listsubrepos=False, |
|
125 | listsubrepos=False, | |
126 | badfn=None, |
|
126 | badfn=None, | |
127 | ): |
|
127 | ): | |
128 | matchers = [] |
|
128 | matchers = [] | |
129 | fms, kindpats = _expandsets( |
|
129 | fms, kindpats = _expandsets( | |
130 | cwd, |
|
130 | cwd, | |
131 | kindpats, |
|
131 | kindpats, | |
132 | ctx=ctx, |
|
132 | ctx=ctx, | |
133 | listsubrepos=listsubrepos, |
|
133 | listsubrepos=listsubrepos, | |
134 | badfn=badfn, |
|
134 | badfn=badfn, | |
135 | ) |
|
135 | ) | |
136 | if kindpats: |
|
136 | if kindpats: | |
137 | m = matchercls(root, kindpats, badfn=badfn) |
|
137 | m = matchercls(root, kindpats, badfn=badfn) | |
138 | matchers.append(m) |
|
138 | matchers.append(m) | |
139 | if fms: |
|
139 | if fms: | |
140 | matchers.extend(fms) |
|
140 | matchers.extend(fms) | |
141 | if not matchers: |
|
141 | if not matchers: | |
142 | return nevermatcher(badfn=badfn) |
|
142 | return nevermatcher(badfn=badfn) | |
143 | if len(matchers) == 1: |
|
143 | if len(matchers) == 1: | |
144 | return matchers[0] |
|
144 | return matchers[0] | |
145 | return unionmatcher(matchers) |
|
145 | return unionmatcher(matchers) | |
146 |
|
146 | |||
147 |
|
147 | |||
148 | def match( |
|
148 | def match( | |
149 | root, |
|
149 | root, | |
150 | cwd, |
|
150 | cwd, | |
151 | patterns=None, |
|
151 | patterns=None, | |
152 | include=None, |
|
152 | include=None, | |
153 | exclude=None, |
|
153 | exclude=None, | |
154 | default=b'glob', |
|
154 | default=b'glob', | |
155 | auditor=None, |
|
155 | auditor=None, | |
156 | ctx=None, |
|
156 | ctx=None, | |
157 | listsubrepos=False, |
|
157 | listsubrepos=False, | |
158 | warn=None, |
|
158 | warn=None, | |
159 | badfn=None, |
|
159 | badfn=None, | |
160 | icasefs=False, |
|
160 | icasefs=False, | |
161 | ): |
|
161 | ): | |
162 | r"""build an object to match a set of file patterns |
|
162 | r"""build an object to match a set of file patterns | |
163 |
|
163 | |||
164 | arguments: |
|
164 | arguments: | |
165 | root - the canonical root of the tree you're matching against |
|
165 | root - the canonical root of the tree you're matching against | |
166 | cwd - the current working directory, if relevant |
|
166 | cwd - the current working directory, if relevant | |
167 | patterns - patterns to find |
|
167 | patterns - patterns to find | |
168 | include - patterns to include (unless they are excluded) |
|
168 | include - patterns to include (unless they are excluded) | |
169 | exclude - patterns to exclude (even if they are included) |
|
169 | exclude - patterns to exclude (even if they are included) | |
170 | default - if a pattern in patterns has no explicit type, assume this one |
|
170 | default - if a pattern in patterns has no explicit type, assume this one | |
171 | auditor - optional path auditor |
|
171 | auditor - optional path auditor | |
172 | ctx - optional changecontext |
|
172 | ctx - optional changecontext | |
173 | listsubrepos - if True, recurse into subrepositories |
|
173 | listsubrepos - if True, recurse into subrepositories | |
174 | warn - optional function used for printing warnings |
|
174 | warn - optional function used for printing warnings | |
175 | badfn - optional bad() callback for this matcher instead of the default |
|
175 | badfn - optional bad() callback for this matcher instead of the default | |
176 | icasefs - make a matcher for wdir on case insensitive filesystems, which |
|
176 | icasefs - make a matcher for wdir on case insensitive filesystems, which | |
177 | normalizes the given patterns to the case in the filesystem |
|
177 | normalizes the given patterns to the case in the filesystem | |
178 |
|
178 | |||
179 | a pattern is one of: |
|
179 | a pattern is one of: | |
180 | 'glob:<glob>' - a glob relative to cwd |
|
180 | 'glob:<glob>' - a glob relative to cwd | |
181 | 're:<regexp>' - a regular expression |
|
181 | 're:<regexp>' - a regular expression | |
182 | 'path:<path>' - a path relative to repository root, which is matched |
|
182 | 'path:<path>' - a path relative to repository root, which is matched | |
183 | recursively |
|
183 | recursively | |
184 | 'rootfilesin:<path>' - a path relative to repository root, which is |
|
184 | 'rootfilesin:<path>' - a path relative to repository root, which is | |
185 | matched non-recursively (will not match subdirectories) |
|
185 | matched non-recursively (will not match subdirectories) | |
186 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
186 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
187 | 'relpath:<path>' - a path relative to cwd |
|
187 | 'relpath:<path>' - a path relative to cwd | |
188 | 'relre:<regexp>' - a regexp that needn't match the start of a name |
|
188 | 'relre:<regexp>' - a regexp that needn't match the start of a name | |
189 | 'set:<fileset>' - a fileset expression |
|
189 | 'set:<fileset>' - a fileset expression | |
190 | 'include:<path>' - a file of patterns to read and include |
|
190 | 'include:<path>' - a file of patterns to read and include | |
191 | 'subinclude:<path>' - a file of patterns to match against files under |
|
191 | 'subinclude:<path>' - a file of patterns to match against files under | |
192 | the same directory |
|
192 | the same directory | |
193 | '<something>' - a pattern of the specified default type |
|
193 | '<something>' - a pattern of the specified default type | |
194 |
|
194 | |||
195 | >>> def _match(root, *args, **kwargs): |
|
195 | >>> def _match(root, *args, **kwargs): | |
196 | ... return match(util.localpath(root), *args, **kwargs) |
|
196 | ... return match(util.localpath(root), *args, **kwargs) | |
197 |
|
197 | |||
198 | Usually a patternmatcher is returned: |
|
198 | Usually a patternmatcher is returned: | |
199 | >>> _match(b'/foo', b'.', [b're:.*\.c$', b'path:foo/a', b'*.py']) |
|
199 | >>> _match(b'/foo', b'.', [b're:.*\.c$', b'path:foo/a', b'*.py']) | |
200 | <patternmatcher patterns='.*\\.c$|foo/a(?:/|$)|[^/]*\\.py$'> |
|
200 | <patternmatcher patterns='.*\\.c$|foo/a(?:/|$)|[^/]*\\.py$'> | |
201 |
|
201 | |||
202 | Combining 'patterns' with 'include' (resp. 'exclude') gives an |
|
202 | Combining 'patterns' with 'include' (resp. 'exclude') gives an | |
203 | intersectionmatcher (resp. a differencematcher): |
|
203 | intersectionmatcher (resp. a differencematcher): | |
204 | >>> type(_match(b'/foo', b'.', [b're:.*\.c$'], include=[b'path:lib'])) |
|
204 | >>> type(_match(b'/foo', b'.', [b're:.*\.c$'], include=[b'path:lib'])) | |
205 | <class 'mercurial.match.intersectionmatcher'> |
|
205 | <class 'mercurial.match.intersectionmatcher'> | |
206 | >>> type(_match(b'/foo', b'.', [b're:.*\.c$'], exclude=[b'path:build'])) |
|
206 | >>> type(_match(b'/foo', b'.', [b're:.*\.c$'], exclude=[b'path:build'])) | |
207 | <class 'mercurial.match.differencematcher'> |
|
207 | <class 'mercurial.match.differencematcher'> | |
208 |
|
208 | |||
209 | Notice that, if 'patterns' is empty, an alwaysmatcher is returned: |
|
209 | Notice that, if 'patterns' is empty, an alwaysmatcher is returned: | |
210 | >>> _match(b'/foo', b'.', []) |
|
210 | >>> _match(b'/foo', b'.', []) | |
211 | <alwaysmatcher> |
|
211 | <alwaysmatcher> | |
212 |
|
212 | |||
213 | The 'default' argument determines which kind of pattern is assumed if a |
|
213 | The 'default' argument determines which kind of pattern is assumed if a | |
214 | pattern has no prefix: |
|
214 | pattern has no prefix: | |
215 | >>> _match(b'/foo', b'.', [b'.*\.c$'], default=b're') |
|
215 | >>> _match(b'/foo', b'.', [b'.*\.c$'], default=b're') | |
216 | <patternmatcher patterns='.*\\.c$'> |
|
216 | <patternmatcher patterns='.*\\.c$'> | |
217 | >>> _match(b'/foo', b'.', [b'main.py'], default=b'relpath') |
|
217 | >>> _match(b'/foo', b'.', [b'main.py'], default=b'relpath') | |
218 | <patternmatcher patterns='main\\.py(?:/|$)'> |
|
218 | <patternmatcher patterns='main\\.py(?:/|$)'> | |
219 | >>> _match(b'/foo', b'.', [b'main.py'], default=b're') |
|
219 | >>> _match(b'/foo', b'.', [b'main.py'], default=b're') | |
220 | <patternmatcher patterns='main.py'> |
|
220 | <patternmatcher patterns='main.py'> | |
221 |
|
221 | |||
222 | The primary use of matchers is to check whether a value (usually a file |
|
222 | The primary use of matchers is to check whether a value (usually a file | |
223 | name) matches againset one of the patterns given at initialization. There |
|
223 | name) matches againset one of the patterns given at initialization. There | |
224 | are two ways of doing this check. |
|
224 | are two ways of doing this check. | |
225 |
|
225 | |||
226 | >>> m = _match(b'/foo', b'', [b're:.*\.c$', b'relpath:a']) |
|
226 | >>> m = _match(b'/foo', b'', [b're:.*\.c$', b'relpath:a']) | |
227 |
|
227 | |||
228 | 1. Calling the matcher with a file name returns True if any pattern |
|
228 | 1. Calling the matcher with a file name returns True if any pattern | |
229 | matches that file name: |
|
229 | matches that file name: | |
230 | >>> m(b'a') |
|
230 | >>> m(b'a') | |
231 | True |
|
231 | True | |
232 | >>> m(b'main.c') |
|
232 | >>> m(b'main.c') | |
233 | True |
|
233 | True | |
234 | >>> m(b'test.py') |
|
234 | >>> m(b'test.py') | |
235 | False |
|
235 | False | |
236 |
|
236 | |||
237 | 2. Using the exact() method only returns True if the file name matches one |
|
237 | 2. Using the exact() method only returns True if the file name matches one | |
238 | of the exact patterns (i.e. not re: or glob: patterns): |
|
238 | of the exact patterns (i.e. not re: or glob: patterns): | |
239 | >>> m.exact(b'a') |
|
239 | >>> m.exact(b'a') | |
240 | True |
|
240 | True | |
241 | >>> m.exact(b'main.c') |
|
241 | >>> m.exact(b'main.c') | |
242 | False |
|
242 | False | |
243 | """ |
|
243 | """ | |
244 | assert os.path.isabs(root) |
|
244 | assert os.path.isabs(root) | |
245 | cwd = os.path.join(root, util.localpath(cwd)) |
|
245 | cwd = os.path.join(root, util.localpath(cwd)) | |
246 | normalize = _donormalize |
|
246 | normalize = _donormalize | |
247 | if icasefs: |
|
247 | if icasefs: | |
248 | dirstate = ctx.repo().dirstate |
|
248 | dirstate = ctx.repo().dirstate | |
249 | dsnormalize = dirstate.normalize |
|
249 | dsnormalize = dirstate.normalize | |
250 |
|
250 | |||
251 | def normalize(patterns, default, root, cwd, auditor, warn): |
|
251 | def normalize(patterns, default, root, cwd, auditor, warn): | |
252 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) |
|
252 | kp = _donormalize(patterns, default, root, cwd, auditor, warn) | |
253 | kindpats = [] |
|
253 | kindpats = [] | |
254 | for kind, pats, source in kp: |
|
254 | for kind, pats, source in kp: | |
255 | if kind not in (b're', b'relre'): # regex can't be normalized |
|
255 | if kind not in (b're', b'relre'): # regex can't be normalized | |
256 | p = pats |
|
256 | p = pats | |
257 | pats = dsnormalize(pats) |
|
257 | pats = dsnormalize(pats) | |
258 |
|
258 | |||
259 | # Preserve the original to handle a case only rename. |
|
259 | # Preserve the original to handle a case only rename. | |
260 | if p != pats and p in dirstate: |
|
260 | if p != pats and p in dirstate: | |
261 | kindpats.append((kind, p, source)) |
|
261 | kindpats.append((kind, p, source)) | |
262 |
|
262 | |||
263 | kindpats.append((kind, pats, source)) |
|
263 | kindpats.append((kind, pats, source)) | |
264 | return kindpats |
|
264 | return kindpats | |
265 |
|
265 | |||
266 | if patterns: |
|
266 | if patterns: | |
267 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) |
|
267 | kindpats = normalize(patterns, default, root, cwd, auditor, warn) | |
268 | if _kindpatsalwaysmatch(kindpats): |
|
268 | if _kindpatsalwaysmatch(kindpats): | |
269 | m = alwaysmatcher(badfn) |
|
269 | m = alwaysmatcher(badfn) | |
270 | else: |
|
270 | else: | |
271 | m = _buildkindpatsmatcher( |
|
271 | m = _buildkindpatsmatcher( | |
272 | patternmatcher, |
|
272 | patternmatcher, | |
273 | root, |
|
273 | root, | |
274 | cwd, |
|
274 | cwd, | |
275 | kindpats, |
|
275 | kindpats, | |
276 | ctx=ctx, |
|
276 | ctx=ctx, | |
277 | listsubrepos=listsubrepos, |
|
277 | listsubrepos=listsubrepos, | |
278 | badfn=badfn, |
|
278 | badfn=badfn, | |
279 | ) |
|
279 | ) | |
280 | else: |
|
280 | else: | |
281 | # It's a little strange that no patterns means to match everything. |
|
281 | # It's a little strange that no patterns means to match everything. | |
282 | # Consider changing this to match nothing (probably using nevermatcher). |
|
282 | # Consider changing this to match nothing (probably using nevermatcher). | |
283 | m = alwaysmatcher(badfn) |
|
283 | m = alwaysmatcher(badfn) | |
284 |
|
284 | |||
285 | if include: |
|
285 | if include: | |
286 | kindpats = normalize(include, b'glob', root, cwd, auditor, warn) |
|
286 | kindpats = normalize(include, b'glob', root, cwd, auditor, warn) | |
287 | im = _buildkindpatsmatcher( |
|
287 | im = _buildkindpatsmatcher( | |
288 | includematcher, |
|
288 | includematcher, | |
289 | root, |
|
289 | root, | |
290 | cwd, |
|
290 | cwd, | |
291 | kindpats, |
|
291 | kindpats, | |
292 | ctx=ctx, |
|
292 | ctx=ctx, | |
293 | listsubrepos=listsubrepos, |
|
293 | listsubrepos=listsubrepos, | |
294 | badfn=None, |
|
294 | badfn=None, | |
295 | ) |
|
295 | ) | |
296 | m = intersectmatchers(m, im) |
|
296 | m = intersectmatchers(m, im) | |
297 | if exclude: |
|
297 | if exclude: | |
298 | kindpats = normalize(exclude, b'glob', root, cwd, auditor, warn) |
|
298 | kindpats = normalize(exclude, b'glob', root, cwd, auditor, warn) | |
299 | em = _buildkindpatsmatcher( |
|
299 | em = _buildkindpatsmatcher( | |
300 | includematcher, |
|
300 | includematcher, | |
301 | root, |
|
301 | root, | |
302 | cwd, |
|
302 | cwd, | |
303 | kindpats, |
|
303 | kindpats, | |
304 | ctx=ctx, |
|
304 | ctx=ctx, | |
305 | listsubrepos=listsubrepos, |
|
305 | listsubrepos=listsubrepos, | |
306 | badfn=None, |
|
306 | badfn=None, | |
307 | ) |
|
307 | ) | |
308 | m = differencematcher(m, em) |
|
308 | m = differencematcher(m, em) | |
309 | return m |
|
309 | return m | |
310 |
|
310 | |||
311 |
|
311 | |||
312 | def exact(files, badfn=None): |
|
312 | def exact(files, badfn=None): | |
313 | return exactmatcher(files, badfn=badfn) |
|
313 | return exactmatcher(files, badfn=badfn) | |
314 |
|
314 | |||
315 |
|
315 | |||
316 | def always(badfn=None): |
|
316 | def always(badfn=None): | |
317 | return alwaysmatcher(badfn) |
|
317 | return alwaysmatcher(badfn) | |
318 |
|
318 | |||
319 |
|
319 | |||
320 | def never(badfn=None): |
|
320 | def never(badfn=None): | |
321 | return nevermatcher(badfn) |
|
321 | return nevermatcher(badfn) | |
322 |
|
322 | |||
323 |
|
323 | |||
324 | def badmatch(match, badfn): |
|
324 | def badmatch(match, badfn): | |
325 | """Make a copy of the given matcher, replacing its bad method with the given |
|
325 | """Make a copy of the given matcher, replacing its bad method with the given | |
326 | one. |
|
326 | one. | |
327 | """ |
|
327 | """ | |
328 | m = copy.copy(match) |
|
328 | m = copy.copy(match) | |
329 | m.bad = badfn |
|
329 | m.bad = badfn | |
330 | return m |
|
330 | return m | |
331 |
|
331 | |||
332 |
|
332 | |||
333 | def _donormalize(patterns, default, root, cwd, auditor=None, warn=None): |
|
333 | def _donormalize(patterns, default, root, cwd, auditor=None, warn=None): | |
334 | """Convert 'kind:pat' from the patterns list to tuples with kind and |
|
334 | """Convert 'kind:pat' from the patterns list to tuples with kind and | |
335 | normalized and rooted patterns and with listfiles expanded.""" |
|
335 | normalized and rooted patterns and with listfiles expanded.""" | |
336 | kindpats = [] |
|
336 | kindpats = [] | |
337 | for kind, pat in [_patsplit(p, default) for p in patterns]: |
|
337 | for kind, pat in [_patsplit(p, default) for p in patterns]: | |
338 | if kind in cwdrelativepatternkinds: |
|
338 | if kind in cwdrelativepatternkinds: | |
339 | pat = pathutil.canonpath(root, cwd, pat, auditor=auditor) |
|
339 | pat = pathutil.canonpath(root, cwd, pat, auditor=auditor) | |
340 | elif kind in (b'relglob', b'path', b'rootfilesin', b'rootglob'): |
|
340 | elif kind in (b'relglob', b'path', b'rootfilesin', b'rootglob'): | |
341 | pat = util.normpath(pat) |
|
341 | pat = util.normpath(pat) | |
342 | elif kind in (b'listfile', b'listfile0'): |
|
342 | elif kind in (b'listfile', b'listfile0'): | |
343 | try: |
|
343 | try: | |
344 | files = util.readfile(pat) |
|
344 | files = util.readfile(pat) | |
345 | if kind == b'listfile0': |
|
345 | if kind == b'listfile0': | |
346 | files = files.split(b'\0') |
|
346 | files = files.split(b'\0') | |
347 | else: |
|
347 | else: | |
348 | files = files.splitlines() |
|
348 | files = files.splitlines() | |
349 | files = [f for f in files if f] |
|
349 | files = [f for f in files if f] | |
350 | except EnvironmentError: |
|
350 | except EnvironmentError: | |
351 | raise error.Abort(_(b"unable to read file list (%s)") % pat) |
|
351 | raise error.Abort(_(b"unable to read file list (%s)") % pat) | |
352 | for k, p, source in _donormalize( |
|
352 | for k, p, source in _donormalize( | |
353 | files, default, root, cwd, auditor, warn |
|
353 | files, default, root, cwd, auditor, warn | |
354 | ): |
|
354 | ): | |
355 | kindpats.append((k, p, pat)) |
|
355 | kindpats.append((k, p, pat)) | |
356 | continue |
|
356 | continue | |
357 | elif kind == b'include': |
|
357 | elif kind == b'include': | |
358 | try: |
|
358 | try: | |
359 | fullpath = os.path.join(root, util.localpath(pat)) |
|
359 | fullpath = os.path.join(root, util.localpath(pat)) | |
360 | includepats = readpatternfile(fullpath, warn) |
|
360 | includepats = readpatternfile(fullpath, warn) | |
361 | for k, p, source in _donormalize( |
|
361 | for k, p, source in _donormalize( | |
362 | includepats, default, root, cwd, auditor, warn |
|
362 | includepats, default, root, cwd, auditor, warn | |
363 | ): |
|
363 | ): | |
364 | kindpats.append((k, p, source or pat)) |
|
364 | kindpats.append((k, p, source or pat)) | |
365 | except error.Abort as inst: |
|
365 | except error.Abort as inst: | |
366 | raise error.Abort( |
|
366 | raise error.Abort( | |
367 | b'%s: %s' |
|
367 | b'%s: %s' | |
368 | % ( |
|
368 | % ( | |
369 | pat, |
|
369 | pat, | |
370 | inst.message, |
|
370 | inst.message, | |
371 | ) # pytype: disable=unsupported-operands |
|
371 | ) # pytype: disable=unsupported-operands | |
372 | ) |
|
372 | ) | |
373 | except IOError as inst: |
|
373 | except IOError as inst: | |
374 | if warn: |
|
374 | if warn: | |
375 | warn( |
|
375 | warn( | |
376 | _(b"skipping unreadable pattern file '%s': %s\n") |
|
376 | _(b"skipping unreadable pattern file '%s': %s\n") | |
377 | % (pat, stringutil.forcebytestr(inst.strerror)) |
|
377 | % (pat, stringutil.forcebytestr(inst.strerror)) | |
378 | ) |
|
378 | ) | |
379 | continue |
|
379 | continue | |
380 | # else: re or relre - which cannot be normalized |
|
380 | # else: re or relre - which cannot be normalized | |
381 | kindpats.append((kind, pat, b'')) |
|
381 | kindpats.append((kind, pat, b'')) | |
382 | return kindpats |
|
382 | return kindpats | |
383 |
|
383 | |||
384 |
|
384 | |||
385 | class basematcher(object): |
|
385 | class basematcher(object): | |
386 | def __init__(self, badfn=None): |
|
386 | def __init__(self, badfn=None): | |
387 | if badfn is not None: |
|
387 | if badfn is not None: | |
388 | self.bad = badfn |
|
388 | self.bad = badfn | |
389 |
|
389 | |||
390 | def __call__(self, fn): |
|
390 | def __call__(self, fn): | |
391 | return self.matchfn(fn) |
|
391 | return self.matchfn(fn) | |
392 |
|
392 | |||
393 | # Callbacks related to how the matcher is used by dirstate.walk. |
|
393 | # Callbacks related to how the matcher is used by dirstate.walk. | |
394 | # Subscribers to these events must monkeypatch the matcher object. |
|
394 | # Subscribers to these events must monkeypatch the matcher object. | |
395 | def bad(self, f, msg): |
|
395 | def bad(self, f, msg): | |
396 | """Callback from dirstate.walk for each explicit file that can't be |
|
396 | """Callback from dirstate.walk for each explicit file that can't be | |
397 | found/accessed, with an error message.""" |
|
397 | found/accessed, with an error message.""" | |
398 |
|
398 | |||
399 | # If an traversedir is set, it will be called when a directory discovered |
|
399 | # If an traversedir is set, it will be called when a directory discovered | |
400 | # by recursive traversal is visited. |
|
400 | # by recursive traversal is visited. | |
401 | traversedir = None |
|
401 | traversedir = None | |
402 |
|
402 | |||
403 | @propertycache |
|
403 | @propertycache | |
404 | def _files(self): |
|
404 | def _files(self): | |
405 | return [] |
|
405 | return [] | |
406 |
|
406 | |||
407 | def files(self): |
|
407 | def files(self): | |
408 | """Explicitly listed files or patterns or roots: |
|
408 | """Explicitly listed files or patterns or roots: | |
409 | if no patterns or .always(): empty list, |
|
409 | if no patterns or .always(): empty list, | |
410 | if exact: list exact files, |
|
410 | if exact: list exact files, | |
411 | if not .anypats(): list all files and dirs, |
|
411 | if not .anypats(): list all files and dirs, | |
412 | else: optimal roots""" |
|
412 | else: optimal roots""" | |
413 | return self._files |
|
413 | return self._files | |
414 |
|
414 | |||
415 | @propertycache |
|
415 | @propertycache | |
416 | def _fileset(self): |
|
416 | def _fileset(self): | |
417 | return set(self._files) |
|
417 | return set(self._files) | |
418 |
|
418 | |||
419 | def exact(self, f): |
|
419 | def exact(self, f): | |
420 | '''Returns True if f is in .files().''' |
|
420 | '''Returns True if f is in .files().''' | |
421 | return f in self._fileset |
|
421 | return f in self._fileset | |
422 |
|
422 | |||
423 | def matchfn(self, f): |
|
423 | def matchfn(self, f): | |
424 | return False |
|
424 | return False | |
425 |
|
425 | |||
426 | def visitdir(self, dir): |
|
426 | def visitdir(self, dir): | |
427 | """Decides whether a directory should be visited based on whether it |
|
427 | """Decides whether a directory should be visited based on whether it | |
428 | has potential matches in it or one of its subdirectories. This is |
|
428 | has potential matches in it or one of its subdirectories. This is | |
429 | based on the match's primary, included, and excluded patterns. |
|
429 | based on the match's primary, included, and excluded patterns. | |
430 |
|
430 | |||
431 | Returns the string 'all' if the given directory and all subdirectories |
|
431 | Returns the string 'all' if the given directory and all subdirectories | |
432 | should be visited. Otherwise returns True or False indicating whether |
|
432 | should be visited. Otherwise returns True or False indicating whether | |
433 | the given directory should be visited. |
|
433 | the given directory should be visited. | |
434 | """ |
|
434 | """ | |
435 | return True |
|
435 | return True | |
436 |
|
436 | |||
437 | def visitchildrenset(self, dir): |
|
437 | def visitchildrenset(self, dir): | |
438 | """Decides whether a directory should be visited based on whether it |
|
438 | """Decides whether a directory should be visited based on whether it | |
439 | has potential matches in it or one of its subdirectories, and |
|
439 | has potential matches in it or one of its subdirectories, and | |
440 | potentially lists which subdirectories of that directory should be |
|
440 | potentially lists which subdirectories of that directory should be | |
441 | visited. This is based on the match's primary, included, and excluded |
|
441 | visited. This is based on the match's primary, included, and excluded | |
442 | patterns. |
|
442 | patterns. | |
443 |
|
443 | |||
444 | This function is very similar to 'visitdir', and the following mapping |
|
444 | This function is very similar to 'visitdir', and the following mapping | |
445 | can be applied: |
|
445 | can be applied: | |
446 |
|
446 | |||
447 | visitdir | visitchildrenlist |
|
447 | visitdir | visitchildrenlist | |
448 | ----------+------------------- |
|
448 | ----------+------------------- | |
449 | False | set() |
|
449 | False | set() | |
450 | 'all' | 'all' |
|
450 | 'all' | 'all' | |
451 | True | 'this' OR non-empty set of subdirs -or files- to visit |
|
451 | True | 'this' OR non-empty set of subdirs -or files- to visit | |
452 |
|
452 | |||
453 | Example: |
|
453 | Example: | |
454 | Assume matchers ['path:foo/bar', 'rootfilesin:qux'], we would return |
|
454 | Assume matchers ['path:foo/bar', 'rootfilesin:qux'], we would return | |
455 | the following values (assuming the implementation of visitchildrenset |
|
455 | the following values (assuming the implementation of visitchildrenset | |
456 | is capable of recognizing this; some implementations are not). |
|
456 | is capable of recognizing this; some implementations are not). | |
457 |
|
457 | |||
458 | '' -> {'foo', 'qux'} |
|
458 | '' -> {'foo', 'qux'} | |
459 | 'baz' -> set() |
|
459 | 'baz' -> set() | |
460 | 'foo' -> {'bar'} |
|
460 | 'foo' -> {'bar'} | |
461 | # Ideally this would be 'all', but since the prefix nature of matchers |
|
461 | # Ideally this would be 'all', but since the prefix nature of matchers | |
462 | # is applied to the entire matcher, we have to downgrade this to |
|
462 | # is applied to the entire matcher, we have to downgrade this to | |
463 | # 'this' due to the non-prefix 'rootfilesin'-kind matcher being mixed |
|
463 | # 'this' due to the non-prefix 'rootfilesin'-kind matcher being mixed | |
464 | # in. |
|
464 | # in. | |
465 | 'foo/bar' -> 'this' |
|
465 | 'foo/bar' -> 'this' | |
466 | 'qux' -> 'this' |
|
466 | 'qux' -> 'this' | |
467 |
|
467 | |||
468 | Important: |
|
468 | Important: | |
469 | Most matchers do not know if they're representing files or |
|
469 | Most matchers do not know if they're representing files or | |
470 | directories. They see ['path:dir/f'] and don't know whether 'f' is a |
|
470 | directories. They see ['path:dir/f'] and don't know whether 'f' is a | |
471 | file or a directory, so visitchildrenset('dir') for most matchers will |
|
471 | file or a directory, so visitchildrenset('dir') for most matchers will | |
472 | return {'f'}, but if the matcher knows it's a file (like exactmatcher |
|
472 | return {'f'}, but if the matcher knows it's a file (like exactmatcher | |
473 | does), it may return 'this'. Do not rely on the return being a set |
|
473 | does), it may return 'this'. Do not rely on the return being a set | |
474 | indicating that there are no files in this dir to investigate (or |
|
474 | indicating that there are no files in this dir to investigate (or | |
475 | equivalently that if there are files to investigate in 'dir' that it |
|
475 | equivalently that if there are files to investigate in 'dir' that it | |
476 | will always return 'this'). |
|
476 | will always return 'this'). | |
477 | """ |
|
477 | """ | |
478 | return b'this' |
|
478 | return b'this' | |
479 |
|
479 | |||
480 | def always(self): |
|
480 | def always(self): | |
481 | """Matcher will match everything and .files() will be empty -- |
|
481 | """Matcher will match everything and .files() will be empty -- | |
482 | optimization might be possible.""" |
|
482 | optimization might be possible.""" | |
483 | return False |
|
483 | return False | |
484 |
|
484 | |||
485 | def isexact(self): |
|
485 | def isexact(self): | |
486 | """Matcher will match exactly the list of files in .files() -- |
|
486 | """Matcher will match exactly the list of files in .files() -- | |
487 | optimization might be possible.""" |
|
487 | optimization might be possible.""" | |
488 | return False |
|
488 | return False | |
489 |
|
489 | |||
490 | def prefix(self): |
|
490 | def prefix(self): | |
491 | """Matcher will match the paths in .files() recursively -- |
|
491 | """Matcher will match the paths in .files() recursively -- | |
492 | optimization might be possible.""" |
|
492 | optimization might be possible.""" | |
493 | return False |
|
493 | return False | |
494 |
|
494 | |||
495 | def anypats(self): |
|
495 | def anypats(self): | |
496 | """None of .always(), .isexact(), and .prefix() is true -- |
|
496 | """None of .always(), .isexact(), and .prefix() is true -- | |
497 | optimizations will be difficult.""" |
|
497 | optimizations will be difficult.""" | |
498 | return not self.always() and not self.isexact() and not self.prefix() |
|
498 | return not self.always() and not self.isexact() and not self.prefix() | |
499 |
|
499 | |||
500 |
|
500 | |||
501 | class alwaysmatcher(basematcher): |
|
501 | class alwaysmatcher(basematcher): | |
502 | '''Matches everything.''' |
|
502 | '''Matches everything.''' | |
503 |
|
503 | |||
504 | def __init__(self, badfn=None): |
|
504 | def __init__(self, badfn=None): | |
505 | super(alwaysmatcher, self).__init__(badfn) |
|
505 | super(alwaysmatcher, self).__init__(badfn) | |
506 |
|
506 | |||
507 | def always(self): |
|
507 | def always(self): | |
508 | return True |
|
508 | return True | |
509 |
|
509 | |||
510 | def matchfn(self, f): |
|
510 | def matchfn(self, f): | |
511 | return True |
|
511 | return True | |
512 |
|
512 | |||
513 | def visitdir(self, dir): |
|
513 | def visitdir(self, dir): | |
514 | return b'all' |
|
514 | return b'all' | |
515 |
|
515 | |||
516 | def visitchildrenset(self, dir): |
|
516 | def visitchildrenset(self, dir): | |
517 | return b'all' |
|
517 | return b'all' | |
518 |
|
518 | |||
519 | def __repr__(self): |
|
519 | def __repr__(self): | |
520 | return r'<alwaysmatcher>' |
|
520 | return r'<alwaysmatcher>' | |
521 |
|
521 | |||
522 |
|
522 | |||
523 | class nevermatcher(basematcher): |
|
523 | class nevermatcher(basematcher): | |
524 | '''Matches nothing.''' |
|
524 | '''Matches nothing.''' | |
525 |
|
525 | |||
526 | def __init__(self, badfn=None): |
|
526 | def __init__(self, badfn=None): | |
527 | super(nevermatcher, self).__init__(badfn) |
|
527 | super(nevermatcher, self).__init__(badfn) | |
528 |
|
528 | |||
529 | # It's a little weird to say that the nevermatcher is an exact matcher |
|
529 | # It's a little weird to say that the nevermatcher is an exact matcher | |
530 | # or a prefix matcher, but it seems to make sense to let callers take |
|
530 | # or a prefix matcher, but it seems to make sense to let callers take | |
531 | # fast paths based on either. There will be no exact matches, nor any |
|
531 | # fast paths based on either. There will be no exact matches, nor any | |
532 | # prefixes (files() returns []), so fast paths iterating over them should |
|
532 | # prefixes (files() returns []), so fast paths iterating over them should | |
533 | # be efficient (and correct). |
|
533 | # be efficient (and correct). | |
534 | def isexact(self): |
|
534 | def isexact(self): | |
535 | return True |
|
535 | return True | |
536 |
|
536 | |||
537 | def prefix(self): |
|
537 | def prefix(self): | |
538 | return True |
|
538 | return True | |
539 |
|
539 | |||
540 | def visitdir(self, dir): |
|
540 | def visitdir(self, dir): | |
541 | return False |
|
541 | return False | |
542 |
|
542 | |||
543 | def visitchildrenset(self, dir): |
|
543 | def visitchildrenset(self, dir): | |
544 | return set() |
|
544 | return set() | |
545 |
|
545 | |||
546 | def __repr__(self): |
|
546 | def __repr__(self): | |
547 | return r'<nevermatcher>' |
|
547 | return r'<nevermatcher>' | |
548 |
|
548 | |||
549 |
|
549 | |||
550 | class predicatematcher(basematcher): |
|
550 | class predicatematcher(basematcher): | |
551 | """A matcher adapter for a simple boolean function""" |
|
551 | """A matcher adapter for a simple boolean function""" | |
552 |
|
552 | |||
553 | def __init__(self, predfn, predrepr=None, badfn=None): |
|
553 | def __init__(self, predfn, predrepr=None, badfn=None): | |
554 | super(predicatematcher, self).__init__(badfn) |
|
554 | super(predicatematcher, self).__init__(badfn) | |
555 | self.matchfn = predfn |
|
555 | self.matchfn = predfn | |
556 | self._predrepr = predrepr |
|
556 | self._predrepr = predrepr | |
557 |
|
557 | |||
558 | @encoding.strmethod |
|
558 | @encoding.strmethod | |
559 | def __repr__(self): |
|
559 | def __repr__(self): | |
560 | s = stringutil.buildrepr(self._predrepr) or pycompat.byterepr( |
|
560 | s = stringutil.buildrepr(self._predrepr) or pycompat.byterepr( | |
561 | self.matchfn |
|
561 | self.matchfn | |
562 | ) |
|
562 | ) | |
563 | return b'<predicatenmatcher pred=%s>' % s |
|
563 | return b'<predicatenmatcher pred=%s>' % s | |
564 |
|
564 | |||
565 |
|
565 | |||
|
566 | def path_or_parents_in_set(path, prefix_set): | |||
|
567 | """Returns True if `path` (or any parent of `path`) is in `prefix_set`.""" | |||
|
568 | l = len(prefix_set) | |||
|
569 | if l == 0: | |||
|
570 | return False | |||
|
571 | if path in prefix_set: | |||
|
572 | return True | |||
|
573 | # If there's more than 5 paths in prefix_set, it's *probably* quicker to | |||
|
574 | # "walk up" the directory hierarchy instead, with the assumption that most | |||
|
575 | # directory hierarchies are relatively shallow and hash lookup is cheap. | |||
|
576 | if l > 5: | |||
|
577 | return any( | |||
|
578 | parentdir in prefix_set for parentdir in pathutil.finddirs(path) | |||
|
579 | ) | |||
|
580 | ||||
|
581 | # FIXME: Ideally we'd never get to this point if this is the case - we'd | |||
|
582 | # recognize ourselves as an 'always' matcher and skip this. | |||
|
583 | if b'' in prefix_set: | |||
|
584 | return True | |||
|
585 | ||||
|
586 | if pycompat.ispy3: | |||
|
587 | sl = ord(b'/') | |||
|
588 | else: | |||
|
589 | sl = '/' | |||
|
590 | ||||
|
591 | # We already checked that path isn't in prefix_set exactly, so | |||
|
592 | # `path[len(pf)] should never raise IndexError. | |||
|
593 | return any(path.startswith(pf) and path[len(pf)] == sl for pf in prefix_set) | |||
|
594 | ||||
|
595 | ||||
566 | class patternmatcher(basematcher): |
|
596 | class patternmatcher(basematcher): | |
567 | r"""Matches a set of (kind, pat, source) against a 'root' directory. |
|
597 | r"""Matches a set of (kind, pat, source) against a 'root' directory. | |
568 |
|
598 | |||
569 | >>> kindpats = [ |
|
599 | >>> kindpats = [ | |
570 | ... (b're', br'.*\.c$', b''), |
|
600 | ... (b're', br'.*\.c$', b''), | |
571 | ... (b'path', b'foo/a', b''), |
|
601 | ... (b'path', b'foo/a', b''), | |
572 | ... (b'relpath', b'b', b''), |
|
602 | ... (b'relpath', b'b', b''), | |
573 | ... (b'glob', b'*.h', b''), |
|
603 | ... (b'glob', b'*.h', b''), | |
574 | ... ] |
|
604 | ... ] | |
575 | >>> m = patternmatcher(b'foo', kindpats) |
|
605 | >>> m = patternmatcher(b'foo', kindpats) | |
576 | >>> m(b'main.c') # matches re:.*\.c$ |
|
606 | >>> m(b'main.c') # matches re:.*\.c$ | |
577 | True |
|
607 | True | |
578 | >>> m(b'b.txt') |
|
608 | >>> m(b'b.txt') | |
579 | False |
|
609 | False | |
580 | >>> m(b'foo/a') # matches path:foo/a |
|
610 | >>> m(b'foo/a') # matches path:foo/a | |
581 | True |
|
611 | True | |
582 | >>> m(b'a') # does not match path:b, since 'root' is 'foo' |
|
612 | >>> m(b'a') # does not match path:b, since 'root' is 'foo' | |
583 | False |
|
613 | False | |
584 | >>> m(b'b') # matches relpath:b, since 'root' is 'foo' |
|
614 | >>> m(b'b') # matches relpath:b, since 'root' is 'foo' | |
585 | True |
|
615 | True | |
586 | >>> m(b'lib.h') # matches glob:*.h |
|
616 | >>> m(b'lib.h') # matches glob:*.h | |
587 | True |
|
617 | True | |
588 |
|
618 | |||
589 | >>> m.files() |
|
619 | >>> m.files() | |
590 | ['', 'foo/a', 'b', ''] |
|
620 | ['', 'foo/a', 'b', ''] | |
591 | >>> m.exact(b'foo/a') |
|
621 | >>> m.exact(b'foo/a') | |
592 | True |
|
622 | True | |
593 | >>> m.exact(b'b') |
|
623 | >>> m.exact(b'b') | |
594 | True |
|
624 | True | |
595 | >>> m.exact(b'lib.h') # exact matches are for (rel)path kinds |
|
625 | >>> m.exact(b'lib.h') # exact matches are for (rel)path kinds | |
596 | False |
|
626 | False | |
597 | """ |
|
627 | """ | |
598 |
|
628 | |||
599 | def __init__(self, root, kindpats, badfn=None): |
|
629 | def __init__(self, root, kindpats, badfn=None): | |
600 | super(patternmatcher, self).__init__(badfn) |
|
630 | super(patternmatcher, self).__init__(badfn) | |
601 |
|
631 | |||
602 | self._files = _explicitfiles(kindpats) |
|
632 | self._files = _explicitfiles(kindpats) | |
603 | self._prefix = _prefix(kindpats) |
|
633 | self._prefix = _prefix(kindpats) | |
604 | self._pats, self.matchfn = _buildmatch(kindpats, b'$', root) |
|
634 | self._pats, self.matchfn = _buildmatch(kindpats, b'$', root) | |
605 |
|
635 | |||
606 | @propertycache |
|
636 | @propertycache | |
607 | def _dirs(self): |
|
637 | def _dirs(self): | |
608 | return set(pathutil.dirs(self._fileset)) |
|
638 | return set(pathutil.dirs(self._fileset)) | |
609 |
|
639 | |||
610 | def visitdir(self, dir): |
|
640 | def visitdir(self, dir): | |
611 | if self._prefix and dir in self._fileset: |
|
641 | if self._prefix and dir in self._fileset: | |
612 | return b'all' |
|
642 | return b'all' | |
613 | return ( |
|
643 | return ( | |
614 |
dir in self._ |
|
644 | dir in self._dirs | |
615 |
or dir |
|
645 | or path_or_parents_in_set(dir, self._fileset) | |
616 | or any( |
|
|||
617 | parentdir in self._fileset |
|
|||
618 | for parentdir in pathutil.finddirs(dir) |
|
|||
619 | ) |
|
|||
620 | ) |
|
646 | ) | |
621 |
|
647 | |||
622 | def visitchildrenset(self, dir): |
|
648 | def visitchildrenset(self, dir): | |
623 | ret = self.visitdir(dir) |
|
649 | ret = self.visitdir(dir) | |
624 | if ret is True: |
|
650 | if ret is True: | |
625 | return b'this' |
|
651 | return b'this' | |
626 | elif not ret: |
|
652 | elif not ret: | |
627 | return set() |
|
653 | return set() | |
628 | assert ret == b'all' |
|
654 | assert ret == b'all' | |
629 | return b'all' |
|
655 | return b'all' | |
630 |
|
656 | |||
631 | def prefix(self): |
|
657 | def prefix(self): | |
632 | return self._prefix |
|
658 | return self._prefix | |
633 |
|
659 | |||
634 | @encoding.strmethod |
|
660 | @encoding.strmethod | |
635 | def __repr__(self): |
|
661 | def __repr__(self): | |
636 | return b'<patternmatcher patterns=%r>' % pycompat.bytestr(self._pats) |
|
662 | return b'<patternmatcher patterns=%r>' % pycompat.bytestr(self._pats) | |
637 |
|
663 | |||
638 |
|
664 | |||
639 | # This is basically a reimplementation of pathutil.dirs that stores the |
|
665 | # This is basically a reimplementation of pathutil.dirs that stores the | |
640 | # children instead of just a count of them, plus a small optional optimization |
|
666 | # children instead of just a count of them, plus a small optional optimization | |
641 | # to avoid some directories we don't need. |
|
667 | # to avoid some directories we don't need. | |
642 | class _dirchildren(object): |
|
668 | class _dirchildren(object): | |
643 | def __init__(self, paths, onlyinclude=None): |
|
669 | def __init__(self, paths, onlyinclude=None): | |
644 | self._dirs = {} |
|
670 | self._dirs = {} | |
645 | self._onlyinclude = onlyinclude or [] |
|
671 | self._onlyinclude = onlyinclude or [] | |
646 | addpath = self.addpath |
|
672 | addpath = self.addpath | |
647 | for f in paths: |
|
673 | for f in paths: | |
648 | addpath(f) |
|
674 | addpath(f) | |
649 |
|
675 | |||
650 | def addpath(self, path): |
|
676 | def addpath(self, path): | |
651 | if path == b'': |
|
677 | if path == b'': | |
652 | return |
|
678 | return | |
653 | dirs = self._dirs |
|
679 | dirs = self._dirs | |
654 | findsplitdirs = _dirchildren._findsplitdirs |
|
680 | findsplitdirs = _dirchildren._findsplitdirs | |
655 | for d, b in findsplitdirs(path): |
|
681 | for d, b in findsplitdirs(path): | |
656 | if d not in self._onlyinclude: |
|
682 | if d not in self._onlyinclude: | |
657 | continue |
|
683 | continue | |
658 | dirs.setdefault(d, set()).add(b) |
|
684 | dirs.setdefault(d, set()).add(b) | |
659 |
|
685 | |||
660 | @staticmethod |
|
686 | @staticmethod | |
661 | def _findsplitdirs(path): |
|
687 | def _findsplitdirs(path): | |
662 | # yields (dirname, basename) tuples, walking back to the root. This is |
|
688 | # yields (dirname, basename) tuples, walking back to the root. This is | |
663 | # very similar to pathutil.finddirs, except: |
|
689 | # very similar to pathutil.finddirs, except: | |
664 | # - produces a (dirname, basename) tuple, not just 'dirname' |
|
690 | # - produces a (dirname, basename) tuple, not just 'dirname' | |
665 | # Unlike manifest._splittopdir, this does not suffix `dirname` with a |
|
691 | # Unlike manifest._splittopdir, this does not suffix `dirname` with a | |
666 | # slash. |
|
692 | # slash. | |
667 | oldpos = len(path) |
|
693 | oldpos = len(path) | |
668 | pos = path.rfind(b'/') |
|
694 | pos = path.rfind(b'/') | |
669 | while pos != -1: |
|
695 | while pos != -1: | |
670 | yield path[:pos], path[pos + 1 : oldpos] |
|
696 | yield path[:pos], path[pos + 1 : oldpos] | |
671 | oldpos = pos |
|
697 | oldpos = pos | |
672 | pos = path.rfind(b'/', 0, pos) |
|
698 | pos = path.rfind(b'/', 0, pos) | |
673 | yield b'', path[:oldpos] |
|
699 | yield b'', path[:oldpos] | |
674 |
|
700 | |||
675 | def get(self, path): |
|
701 | def get(self, path): | |
676 | return self._dirs.get(path, set()) |
|
702 | return self._dirs.get(path, set()) | |
677 |
|
703 | |||
678 |
|
704 | |||
679 | class includematcher(basematcher): |
|
705 | class includematcher(basematcher): | |
680 | def __init__(self, root, kindpats, badfn=None): |
|
706 | def __init__(self, root, kindpats, badfn=None): | |
681 | super(includematcher, self).__init__(badfn) |
|
707 | super(includematcher, self).__init__(badfn) | |
682 | if rustmod is not None: |
|
708 | if rustmod is not None: | |
683 | # We need to pass the patterns to Rust because they can contain |
|
709 | # We need to pass the patterns to Rust because they can contain | |
684 | # patterns from the user interface |
|
710 | # patterns from the user interface | |
685 | self._kindpats = kindpats |
|
711 | self._kindpats = kindpats | |
686 | self._pats, self.matchfn = _buildmatch(kindpats, b'(?:/|$)', root) |
|
712 | self._pats, self.matchfn = _buildmatch(kindpats, b'(?:/|$)', root) | |
687 | self._prefix = _prefix(kindpats) |
|
713 | self._prefix = _prefix(kindpats) | |
688 | roots, dirs, parents = _rootsdirsandparents(kindpats) |
|
714 | roots, dirs, parents = _rootsdirsandparents(kindpats) | |
689 | # roots are directories which are recursively included. |
|
715 | # roots are directories which are recursively included. | |
690 | self._roots = set(roots) |
|
716 | self._roots = set(roots) | |
691 | # dirs are directories which are non-recursively included. |
|
717 | # dirs are directories which are non-recursively included. | |
692 | self._dirs = set(dirs) |
|
718 | self._dirs = set(dirs) | |
693 | # parents are directories which are non-recursively included because |
|
719 | # parents are directories which are non-recursively included because | |
694 | # they are needed to get to items in _dirs or _roots. |
|
720 | # they are needed to get to items in _dirs or _roots. | |
695 | self._parents = parents |
|
721 | self._parents = parents | |
696 |
|
722 | |||
697 | def visitdir(self, dir): |
|
723 | def visitdir(self, dir): | |
698 | if self._prefix and dir in self._roots: |
|
724 | if self._prefix and dir in self._roots: | |
699 | return b'all' |
|
725 | return b'all' | |
700 | return ( |
|
726 | return ( | |
701 |
dir in self._r |
|
727 | dir in self._dirs | |
702 | or dir in self._dirs |
|
|||
703 | or dir in self._parents |
|
728 | or dir in self._parents | |
704 | or any( |
|
729 | or path_or_parents_in_set(dir, self._roots) | |
705 | parentdir in self._roots for parentdir in pathutil.finddirs(dir) |
|
|||
706 | ) |
|
|||
707 | ) |
|
730 | ) | |
708 |
|
731 | |||
709 | @propertycache |
|
732 | @propertycache | |
710 | def _allparentschildren(self): |
|
733 | def _allparentschildren(self): | |
711 | # It may seem odd that we add dirs, roots, and parents, and then |
|
734 | # It may seem odd that we add dirs, roots, and parents, and then | |
712 | # restrict to only parents. This is to catch the case of: |
|
735 | # restrict to only parents. This is to catch the case of: | |
713 | # dirs = ['foo/bar'] |
|
736 | # dirs = ['foo/bar'] | |
714 | # parents = ['foo'] |
|
737 | # parents = ['foo'] | |
715 | # if we asked for the children of 'foo', but had only added |
|
738 | # if we asked for the children of 'foo', but had only added | |
716 | # self._parents, we wouldn't be able to respond ['bar']. |
|
739 | # self._parents, we wouldn't be able to respond ['bar']. | |
717 | return _dirchildren( |
|
740 | return _dirchildren( | |
718 | itertools.chain(self._dirs, self._roots, self._parents), |
|
741 | itertools.chain(self._dirs, self._roots, self._parents), | |
719 | onlyinclude=self._parents, |
|
742 | onlyinclude=self._parents, | |
720 | ) |
|
743 | ) | |
721 |
|
744 | |||
722 | def visitchildrenset(self, dir): |
|
745 | def visitchildrenset(self, dir): | |
723 | if self._prefix and dir in self._roots: |
|
746 | if self._prefix and dir in self._roots: | |
724 | return b'all' |
|
747 | return b'all' | |
725 | # Note: this does *not* include the 'dir in self._parents' case from |
|
748 | # Note: this does *not* include the 'dir in self._parents' case from | |
726 | # visitdir, that's handled below. |
|
749 | # visitdir, that's handled below. | |
727 | if ( |
|
750 | if ( | |
728 | b'' in self._roots |
|
751 | b'' in self._roots | |
729 | or dir in self._roots |
|
|||
730 | or dir in self._dirs |
|
752 | or dir in self._dirs | |
731 | or any( |
|
753 | or path_or_parents_in_set(dir, self._roots) | |
732 | parentdir in self._roots for parentdir in pathutil.finddirs(dir) |
|
|||
733 | ) |
|
|||
734 | ): |
|
754 | ): | |
735 | return b'this' |
|
755 | return b'this' | |
736 |
|
756 | |||
737 | if dir in self._parents: |
|
757 | if dir in self._parents: | |
738 | return self._allparentschildren.get(dir) or set() |
|
758 | return self._allparentschildren.get(dir) or set() | |
739 | return set() |
|
759 | return set() | |
740 |
|
760 | |||
741 | @encoding.strmethod |
|
761 | @encoding.strmethod | |
742 | def __repr__(self): |
|
762 | def __repr__(self): | |
743 | return b'<includematcher includes=%r>' % pycompat.bytestr(self._pats) |
|
763 | return b'<includematcher includes=%r>' % pycompat.bytestr(self._pats) | |
744 |
|
764 | |||
745 |
|
765 | |||
746 | class exactmatcher(basematcher): |
|
766 | class exactmatcher(basematcher): | |
747 | r"""Matches the input files exactly. They are interpreted as paths, not |
|
767 | r"""Matches the input files exactly. They are interpreted as paths, not | |
748 | patterns (so no kind-prefixes). |
|
768 | patterns (so no kind-prefixes). | |
749 |
|
769 | |||
750 | >>> m = exactmatcher([b'a.txt', br're:.*\.c$']) |
|
770 | >>> m = exactmatcher([b'a.txt', br're:.*\.c$']) | |
751 | >>> m(b'a.txt') |
|
771 | >>> m(b'a.txt') | |
752 | True |
|
772 | True | |
753 | >>> m(b'b.txt') |
|
773 | >>> m(b'b.txt') | |
754 | False |
|
774 | False | |
755 |
|
775 | |||
756 | Input files that would be matched are exactly those returned by .files() |
|
776 | Input files that would be matched are exactly those returned by .files() | |
757 | >>> m.files() |
|
777 | >>> m.files() | |
758 | ['a.txt', 're:.*\\.c$'] |
|
778 | ['a.txt', 're:.*\\.c$'] | |
759 |
|
779 | |||
760 | So pattern 're:.*\.c$' is not considered as a regex, but as a file name |
|
780 | So pattern 're:.*\.c$' is not considered as a regex, but as a file name | |
761 | >>> m(b'main.c') |
|
781 | >>> m(b'main.c') | |
762 | False |
|
782 | False | |
763 | >>> m(br're:.*\.c$') |
|
783 | >>> m(br're:.*\.c$') | |
764 | True |
|
784 | True | |
765 | """ |
|
785 | """ | |
766 |
|
786 | |||
767 | def __init__(self, files, badfn=None): |
|
787 | def __init__(self, files, badfn=None): | |
768 | super(exactmatcher, self).__init__(badfn) |
|
788 | super(exactmatcher, self).__init__(badfn) | |
769 |
|
789 | |||
770 | if isinstance(files, list): |
|
790 | if isinstance(files, list): | |
771 | self._files = files |
|
791 | self._files = files | |
772 | else: |
|
792 | else: | |
773 | self._files = list(files) |
|
793 | self._files = list(files) | |
774 |
|
794 | |||
775 | matchfn = basematcher.exact |
|
795 | matchfn = basematcher.exact | |
776 |
|
796 | |||
777 | @propertycache |
|
797 | @propertycache | |
778 | def _dirs(self): |
|
798 | def _dirs(self): | |
779 | return set(pathutil.dirs(self._fileset)) |
|
799 | return set(pathutil.dirs(self._fileset)) | |
780 |
|
800 | |||
781 | def visitdir(self, dir): |
|
801 | def visitdir(self, dir): | |
782 | return dir in self._dirs |
|
802 | return dir in self._dirs | |
783 |
|
803 | |||
784 | def visitchildrenset(self, dir): |
|
804 | def visitchildrenset(self, dir): | |
785 | if not self._fileset or dir not in self._dirs: |
|
805 | if not self._fileset or dir not in self._dirs: | |
786 | return set() |
|
806 | return set() | |
787 |
|
807 | |||
788 | candidates = self._fileset | self._dirs - {b''} |
|
808 | candidates = self._fileset | self._dirs - {b''} | |
789 | if dir != b'': |
|
809 | if dir != b'': | |
790 | d = dir + b'/' |
|
810 | d = dir + b'/' | |
791 | candidates = {c[len(d) :] for c in candidates if c.startswith(d)} |
|
811 | candidates = {c[len(d) :] for c in candidates if c.startswith(d)} | |
792 | # self._dirs includes all of the directories, recursively, so if |
|
812 | # self._dirs includes all of the directories, recursively, so if | |
793 | # we're attempting to match foo/bar/baz.txt, it'll have '', 'foo', |
|
813 | # we're attempting to match foo/bar/baz.txt, it'll have '', 'foo', | |
794 | # 'foo/bar' in it. Thus we can safely ignore a candidate that has a |
|
814 | # 'foo/bar' in it. Thus we can safely ignore a candidate that has a | |
795 | # '/' in it, indicating a it's for a subdir-of-a-subdir; the |
|
815 | # '/' in it, indicating a it's for a subdir-of-a-subdir; the | |
796 | # immediate subdir will be in there without a slash. |
|
816 | # immediate subdir will be in there without a slash. | |
797 | ret = {c for c in candidates if b'/' not in c} |
|
817 | ret = {c for c in candidates if b'/' not in c} | |
798 | # We really do not expect ret to be empty, since that would imply that |
|
818 | # We really do not expect ret to be empty, since that would imply that | |
799 | # there's something in _dirs that didn't have a file in _fileset. |
|
819 | # there's something in _dirs that didn't have a file in _fileset. | |
800 | assert ret |
|
820 | assert ret | |
801 | return ret |
|
821 | return ret | |
802 |
|
822 | |||
803 | def isexact(self): |
|
823 | def isexact(self): | |
804 | return True |
|
824 | return True | |
805 |
|
825 | |||
806 | @encoding.strmethod |
|
826 | @encoding.strmethod | |
807 | def __repr__(self): |
|
827 | def __repr__(self): | |
808 | return b'<exactmatcher files=%r>' % self._files |
|
828 | return b'<exactmatcher files=%r>' % self._files | |
809 |
|
829 | |||
810 |
|
830 | |||
811 | class differencematcher(basematcher): |
|
831 | class differencematcher(basematcher): | |
812 | """Composes two matchers by matching if the first matches and the second |
|
832 | """Composes two matchers by matching if the first matches and the second | |
813 | does not. |
|
833 | does not. | |
814 |
|
834 | |||
815 | The second matcher's non-matching-attributes (bad, traversedir) are ignored. |
|
835 | The second matcher's non-matching-attributes (bad, traversedir) are ignored. | |
816 | """ |
|
836 | """ | |
817 |
|
837 | |||
818 | def __init__(self, m1, m2): |
|
838 | def __init__(self, m1, m2): | |
819 | super(differencematcher, self).__init__() |
|
839 | super(differencematcher, self).__init__() | |
820 | self._m1 = m1 |
|
840 | self._m1 = m1 | |
821 | self._m2 = m2 |
|
841 | self._m2 = m2 | |
822 | self.bad = m1.bad |
|
842 | self.bad = m1.bad | |
823 | self.traversedir = m1.traversedir |
|
843 | self.traversedir = m1.traversedir | |
824 |
|
844 | |||
825 | def matchfn(self, f): |
|
845 | def matchfn(self, f): | |
826 | return self._m1(f) and not self._m2(f) |
|
846 | return self._m1(f) and not self._m2(f) | |
827 |
|
847 | |||
828 | @propertycache |
|
848 | @propertycache | |
829 | def _files(self): |
|
849 | def _files(self): | |
830 | if self.isexact(): |
|
850 | if self.isexact(): | |
831 | return [f for f in self._m1.files() if self(f)] |
|
851 | return [f for f in self._m1.files() if self(f)] | |
832 | # If m1 is not an exact matcher, we can't easily figure out the set of |
|
852 | # If m1 is not an exact matcher, we can't easily figure out the set of | |
833 | # files, because its files() are not always files. For example, if |
|
853 | # files, because its files() are not always files. For example, if | |
834 | # m1 is "path:dir" and m2 is "rootfileins:.", we don't |
|
854 | # m1 is "path:dir" and m2 is "rootfileins:.", we don't | |
835 | # want to remove "dir" from the set even though it would match m2, |
|
855 | # want to remove "dir" from the set even though it would match m2, | |
836 | # because the "dir" in m1 may not be a file. |
|
856 | # because the "dir" in m1 may not be a file. | |
837 | return self._m1.files() |
|
857 | return self._m1.files() | |
838 |
|
858 | |||
839 | def visitdir(self, dir): |
|
859 | def visitdir(self, dir): | |
840 | if self._m2.visitdir(dir) == b'all': |
|
860 | if self._m2.visitdir(dir) == b'all': | |
841 | return False |
|
861 | return False | |
842 | elif not self._m2.visitdir(dir): |
|
862 | elif not self._m2.visitdir(dir): | |
843 | # m2 does not match dir, we can return 'all' here if possible |
|
863 | # m2 does not match dir, we can return 'all' here if possible | |
844 | return self._m1.visitdir(dir) |
|
864 | return self._m1.visitdir(dir) | |
845 | return bool(self._m1.visitdir(dir)) |
|
865 | return bool(self._m1.visitdir(dir)) | |
846 |
|
866 | |||
847 | def visitchildrenset(self, dir): |
|
867 | def visitchildrenset(self, dir): | |
848 | m2_set = self._m2.visitchildrenset(dir) |
|
868 | m2_set = self._m2.visitchildrenset(dir) | |
849 | if m2_set == b'all': |
|
869 | if m2_set == b'all': | |
850 | return set() |
|
870 | return set() | |
851 | m1_set = self._m1.visitchildrenset(dir) |
|
871 | m1_set = self._m1.visitchildrenset(dir) | |
852 | # Possible values for m1: 'all', 'this', set(...), set() |
|
872 | # Possible values for m1: 'all', 'this', set(...), set() | |
853 | # Possible values for m2: 'this', set(...), set() |
|
873 | # Possible values for m2: 'this', set(...), set() | |
854 | # If m2 has nothing under here that we care about, return m1, even if |
|
874 | # If m2 has nothing under here that we care about, return m1, even if | |
855 | # it's 'all'. This is a change in behavior from visitdir, which would |
|
875 | # it's 'all'. This is a change in behavior from visitdir, which would | |
856 | # return True, not 'all', for some reason. |
|
876 | # return True, not 'all', for some reason. | |
857 | if not m2_set: |
|
877 | if not m2_set: | |
858 | return m1_set |
|
878 | return m1_set | |
859 | if m1_set in [b'all', b'this']: |
|
879 | if m1_set in [b'all', b'this']: | |
860 | # Never return 'all' here if m2_set is any kind of non-empty (either |
|
880 | # Never return 'all' here if m2_set is any kind of non-empty (either | |
861 | # 'this' or set(foo)), since m2 might return set() for a |
|
881 | # 'this' or set(foo)), since m2 might return set() for a | |
862 | # subdirectory. |
|
882 | # subdirectory. | |
863 | return b'this' |
|
883 | return b'this' | |
864 | # Possible values for m1: set(...), set() |
|
884 | # Possible values for m1: set(...), set() | |
865 | # Possible values for m2: 'this', set(...) |
|
885 | # Possible values for m2: 'this', set(...) | |
866 | # We ignore m2's set results. They're possibly incorrect: |
|
886 | # We ignore m2's set results. They're possibly incorrect: | |
867 | # m1 = path:dir/subdir, m2=rootfilesin:dir, visitchildrenset(''): |
|
887 | # m1 = path:dir/subdir, m2=rootfilesin:dir, visitchildrenset(''): | |
868 | # m1 returns {'dir'}, m2 returns {'dir'}, if we subtracted we'd |
|
888 | # m1 returns {'dir'}, m2 returns {'dir'}, if we subtracted we'd | |
869 | # return set(), which is *not* correct, we still need to visit 'dir'! |
|
889 | # return set(), which is *not* correct, we still need to visit 'dir'! | |
870 | return m1_set |
|
890 | return m1_set | |
871 |
|
891 | |||
872 | def isexact(self): |
|
892 | def isexact(self): | |
873 | return self._m1.isexact() |
|
893 | return self._m1.isexact() | |
874 |
|
894 | |||
875 | @encoding.strmethod |
|
895 | @encoding.strmethod | |
876 | def __repr__(self): |
|
896 | def __repr__(self): | |
877 | return b'<differencematcher m1=%r, m2=%r>' % (self._m1, self._m2) |
|
897 | return b'<differencematcher m1=%r, m2=%r>' % (self._m1, self._m2) | |
878 |
|
898 | |||
879 |
|
899 | |||
880 | def intersectmatchers(m1, m2): |
|
900 | def intersectmatchers(m1, m2): | |
881 | """Composes two matchers by matching if both of them match. |
|
901 | """Composes two matchers by matching if both of them match. | |
882 |
|
902 | |||
883 | The second matcher's non-matching-attributes (bad, traversedir) are ignored. |
|
903 | The second matcher's non-matching-attributes (bad, traversedir) are ignored. | |
884 | """ |
|
904 | """ | |
885 | if m1 is None or m2 is None: |
|
905 | if m1 is None or m2 is None: | |
886 | return m1 or m2 |
|
906 | return m1 or m2 | |
887 | if m1.always(): |
|
907 | if m1.always(): | |
888 | m = copy.copy(m2) |
|
908 | m = copy.copy(m2) | |
889 | # TODO: Consider encapsulating these things in a class so there's only |
|
909 | # TODO: Consider encapsulating these things in a class so there's only | |
890 | # one thing to copy from m1. |
|
910 | # one thing to copy from m1. | |
891 | m.bad = m1.bad |
|
911 | m.bad = m1.bad | |
892 | m.traversedir = m1.traversedir |
|
912 | m.traversedir = m1.traversedir | |
893 | return m |
|
913 | return m | |
894 | if m2.always(): |
|
914 | if m2.always(): | |
895 | m = copy.copy(m1) |
|
915 | m = copy.copy(m1) | |
896 | return m |
|
916 | return m | |
897 | return intersectionmatcher(m1, m2) |
|
917 | return intersectionmatcher(m1, m2) | |
898 |
|
918 | |||
899 |
|
919 | |||
900 | class intersectionmatcher(basematcher): |
|
920 | class intersectionmatcher(basematcher): | |
901 | def __init__(self, m1, m2): |
|
921 | def __init__(self, m1, m2): | |
902 | super(intersectionmatcher, self).__init__() |
|
922 | super(intersectionmatcher, self).__init__() | |
903 | self._m1 = m1 |
|
923 | self._m1 = m1 | |
904 | self._m2 = m2 |
|
924 | self._m2 = m2 | |
905 | self.bad = m1.bad |
|
925 | self.bad = m1.bad | |
906 | self.traversedir = m1.traversedir |
|
926 | self.traversedir = m1.traversedir | |
907 |
|
927 | |||
908 | @propertycache |
|
928 | @propertycache | |
909 | def _files(self): |
|
929 | def _files(self): | |
910 | if self.isexact(): |
|
930 | if self.isexact(): | |
911 | m1, m2 = self._m1, self._m2 |
|
931 | m1, m2 = self._m1, self._m2 | |
912 | if not m1.isexact(): |
|
932 | if not m1.isexact(): | |
913 | m1, m2 = m2, m1 |
|
933 | m1, m2 = m2, m1 | |
914 | return [f for f in m1.files() if m2(f)] |
|
934 | return [f for f in m1.files() if m2(f)] | |
915 | # It neither m1 nor m2 is an exact matcher, we can't easily intersect |
|
935 | # It neither m1 nor m2 is an exact matcher, we can't easily intersect | |
916 | # the set of files, because their files() are not always files. For |
|
936 | # the set of files, because their files() are not always files. For | |
917 | # example, if intersecting a matcher "-I glob:foo.txt" with matcher of |
|
937 | # example, if intersecting a matcher "-I glob:foo.txt" with matcher of | |
918 | # "path:dir2", we don't want to remove "dir2" from the set. |
|
938 | # "path:dir2", we don't want to remove "dir2" from the set. | |
919 | return self._m1.files() + self._m2.files() |
|
939 | return self._m1.files() + self._m2.files() | |
920 |
|
940 | |||
921 | def matchfn(self, f): |
|
941 | def matchfn(self, f): | |
922 | return self._m1(f) and self._m2(f) |
|
942 | return self._m1(f) and self._m2(f) | |
923 |
|
943 | |||
924 | def visitdir(self, dir): |
|
944 | def visitdir(self, dir): | |
925 | visit1 = self._m1.visitdir(dir) |
|
945 | visit1 = self._m1.visitdir(dir) | |
926 | if visit1 == b'all': |
|
946 | if visit1 == b'all': | |
927 | return self._m2.visitdir(dir) |
|
947 | return self._m2.visitdir(dir) | |
928 | # bool() because visit1=True + visit2='all' should not be 'all' |
|
948 | # bool() because visit1=True + visit2='all' should not be 'all' | |
929 | return bool(visit1 and self._m2.visitdir(dir)) |
|
949 | return bool(visit1 and self._m2.visitdir(dir)) | |
930 |
|
950 | |||
931 | def visitchildrenset(self, dir): |
|
951 | def visitchildrenset(self, dir): | |
932 | m1_set = self._m1.visitchildrenset(dir) |
|
952 | m1_set = self._m1.visitchildrenset(dir) | |
933 | if not m1_set: |
|
953 | if not m1_set: | |
934 | return set() |
|
954 | return set() | |
935 | m2_set = self._m2.visitchildrenset(dir) |
|
955 | m2_set = self._m2.visitchildrenset(dir) | |
936 | if not m2_set: |
|
956 | if not m2_set: | |
937 | return set() |
|
957 | return set() | |
938 |
|
958 | |||
939 | if m1_set == b'all': |
|
959 | if m1_set == b'all': | |
940 | return m2_set |
|
960 | return m2_set | |
941 | elif m2_set == b'all': |
|
961 | elif m2_set == b'all': | |
942 | return m1_set |
|
962 | return m1_set | |
943 |
|
963 | |||
944 | if m1_set == b'this' or m2_set == b'this': |
|
964 | if m1_set == b'this' or m2_set == b'this': | |
945 | return b'this' |
|
965 | return b'this' | |
946 |
|
966 | |||
947 | assert isinstance(m1_set, set) and isinstance(m2_set, set) |
|
967 | assert isinstance(m1_set, set) and isinstance(m2_set, set) | |
948 | return m1_set.intersection(m2_set) |
|
968 | return m1_set.intersection(m2_set) | |
949 |
|
969 | |||
950 | def always(self): |
|
970 | def always(self): | |
951 | return self._m1.always() and self._m2.always() |
|
971 | return self._m1.always() and self._m2.always() | |
952 |
|
972 | |||
953 | def isexact(self): |
|
973 | def isexact(self): | |
954 | return self._m1.isexact() or self._m2.isexact() |
|
974 | return self._m1.isexact() or self._m2.isexact() | |
955 |
|
975 | |||
956 | @encoding.strmethod |
|
976 | @encoding.strmethod | |
957 | def __repr__(self): |
|
977 | def __repr__(self): | |
958 | return b'<intersectionmatcher m1=%r, m2=%r>' % (self._m1, self._m2) |
|
978 | return b'<intersectionmatcher m1=%r, m2=%r>' % (self._m1, self._m2) | |
959 |
|
979 | |||
960 |
|
980 | |||
961 | class subdirmatcher(basematcher): |
|
981 | class subdirmatcher(basematcher): | |
962 | """Adapt a matcher to work on a subdirectory only. |
|
982 | """Adapt a matcher to work on a subdirectory only. | |
963 |
|
983 | |||
964 | The paths are remapped to remove/insert the path as needed: |
|
984 | The paths are remapped to remove/insert the path as needed: | |
965 |
|
985 | |||
966 | >>> from . import pycompat |
|
986 | >>> from . import pycompat | |
967 | >>> m1 = match(util.localpath(b'/root'), b'', [b'a.txt', b'sub/b.txt'], auditor=lambda name: None) |
|
987 | >>> m1 = match(util.localpath(b'/root'), b'', [b'a.txt', b'sub/b.txt'], auditor=lambda name: None) | |
968 | >>> m2 = subdirmatcher(b'sub', m1) |
|
988 | >>> m2 = subdirmatcher(b'sub', m1) | |
969 | >>> m2(b'a.txt') |
|
989 | >>> m2(b'a.txt') | |
970 | False |
|
990 | False | |
971 | >>> m2(b'b.txt') |
|
991 | >>> m2(b'b.txt') | |
972 | True |
|
992 | True | |
973 | >>> m2.matchfn(b'a.txt') |
|
993 | >>> m2.matchfn(b'a.txt') | |
974 | False |
|
994 | False | |
975 | >>> m2.matchfn(b'b.txt') |
|
995 | >>> m2.matchfn(b'b.txt') | |
976 | True |
|
996 | True | |
977 | >>> m2.files() |
|
997 | >>> m2.files() | |
978 | ['b.txt'] |
|
998 | ['b.txt'] | |
979 | >>> m2.exact(b'b.txt') |
|
999 | >>> m2.exact(b'b.txt') | |
980 | True |
|
1000 | True | |
981 | >>> def bad(f, msg): |
|
1001 | >>> def bad(f, msg): | |
982 | ... print(pycompat.sysstr(b"%s: %s" % (f, msg))) |
|
1002 | ... print(pycompat.sysstr(b"%s: %s" % (f, msg))) | |
983 | >>> m1.bad = bad |
|
1003 | >>> m1.bad = bad | |
984 | >>> m2.bad(b'x.txt', b'No such file') |
|
1004 | >>> m2.bad(b'x.txt', b'No such file') | |
985 | sub/x.txt: No such file |
|
1005 | sub/x.txt: No such file | |
986 | """ |
|
1006 | """ | |
987 |
|
1007 | |||
988 | def __init__(self, path, matcher): |
|
1008 | def __init__(self, path, matcher): | |
989 | super(subdirmatcher, self).__init__() |
|
1009 | super(subdirmatcher, self).__init__() | |
990 | self._path = path |
|
1010 | self._path = path | |
991 | self._matcher = matcher |
|
1011 | self._matcher = matcher | |
992 | self._always = matcher.always() |
|
1012 | self._always = matcher.always() | |
993 |
|
1013 | |||
994 | self._files = [ |
|
1014 | self._files = [ | |
995 | f[len(path) + 1 :] |
|
1015 | f[len(path) + 1 :] | |
996 | for f in matcher._files |
|
1016 | for f in matcher._files | |
997 | if f.startswith(path + b"/") |
|
1017 | if f.startswith(path + b"/") | |
998 | ] |
|
1018 | ] | |
999 |
|
1019 | |||
1000 | # If the parent repo had a path to this subrepo and the matcher is |
|
1020 | # If the parent repo had a path to this subrepo and the matcher is | |
1001 | # a prefix matcher, this submatcher always matches. |
|
1021 | # a prefix matcher, this submatcher always matches. | |
1002 | if matcher.prefix(): |
|
1022 | if matcher.prefix(): | |
1003 | self._always = any(f == path for f in matcher._files) |
|
1023 | self._always = any(f == path for f in matcher._files) | |
1004 |
|
1024 | |||
1005 | def bad(self, f, msg): |
|
1025 | def bad(self, f, msg): | |
1006 | self._matcher.bad(self._path + b"/" + f, msg) |
|
1026 | self._matcher.bad(self._path + b"/" + f, msg) | |
1007 |
|
1027 | |||
1008 | def matchfn(self, f): |
|
1028 | def matchfn(self, f): | |
1009 | # Some information is lost in the superclass's constructor, so we |
|
1029 | # Some information is lost in the superclass's constructor, so we | |
1010 | # can not accurately create the matching function for the subdirectory |
|
1030 | # can not accurately create the matching function for the subdirectory | |
1011 | # from the inputs. Instead, we override matchfn() and visitdir() to |
|
1031 | # from the inputs. Instead, we override matchfn() and visitdir() to | |
1012 | # call the original matcher with the subdirectory path prepended. |
|
1032 | # call the original matcher with the subdirectory path prepended. | |
1013 | return self._matcher.matchfn(self._path + b"/" + f) |
|
1033 | return self._matcher.matchfn(self._path + b"/" + f) | |
1014 |
|
1034 | |||
1015 | def visitdir(self, dir): |
|
1035 | def visitdir(self, dir): | |
1016 | if dir == b'': |
|
1036 | if dir == b'': | |
1017 | dir = self._path |
|
1037 | dir = self._path | |
1018 | else: |
|
1038 | else: | |
1019 | dir = self._path + b"/" + dir |
|
1039 | dir = self._path + b"/" + dir | |
1020 | return self._matcher.visitdir(dir) |
|
1040 | return self._matcher.visitdir(dir) | |
1021 |
|
1041 | |||
1022 | def visitchildrenset(self, dir): |
|
1042 | def visitchildrenset(self, dir): | |
1023 | if dir == b'': |
|
1043 | if dir == b'': | |
1024 | dir = self._path |
|
1044 | dir = self._path | |
1025 | else: |
|
1045 | else: | |
1026 | dir = self._path + b"/" + dir |
|
1046 | dir = self._path + b"/" + dir | |
1027 | return self._matcher.visitchildrenset(dir) |
|
1047 | return self._matcher.visitchildrenset(dir) | |
1028 |
|
1048 | |||
1029 | def always(self): |
|
1049 | def always(self): | |
1030 | return self._always |
|
1050 | return self._always | |
1031 |
|
1051 | |||
1032 | def prefix(self): |
|
1052 | def prefix(self): | |
1033 | return self._matcher.prefix() and not self._always |
|
1053 | return self._matcher.prefix() and not self._always | |
1034 |
|
1054 | |||
1035 | @encoding.strmethod |
|
1055 | @encoding.strmethod | |
1036 | def __repr__(self): |
|
1056 | def __repr__(self): | |
1037 | return b'<subdirmatcher path=%r, matcher=%r>' % ( |
|
1057 | return b'<subdirmatcher path=%r, matcher=%r>' % ( | |
1038 | self._path, |
|
1058 | self._path, | |
1039 | self._matcher, |
|
1059 | self._matcher, | |
1040 | ) |
|
1060 | ) | |
1041 |
|
1061 | |||
1042 |
|
1062 | |||
1043 | class prefixdirmatcher(basematcher): |
|
1063 | class prefixdirmatcher(basematcher): | |
1044 | """Adapt a matcher to work on a parent directory. |
|
1064 | """Adapt a matcher to work on a parent directory. | |
1045 |
|
1065 | |||
1046 | The matcher's non-matching-attributes (bad, traversedir) are ignored. |
|
1066 | The matcher's non-matching-attributes (bad, traversedir) are ignored. | |
1047 |
|
1067 | |||
1048 | The prefix path should usually be the relative path from the root of |
|
1068 | The prefix path should usually be the relative path from the root of | |
1049 | this matcher to the root of the wrapped matcher. |
|
1069 | this matcher to the root of the wrapped matcher. | |
1050 |
|
1070 | |||
1051 | >>> m1 = match(util.localpath(b'/root/d/e'), b'f', [b'../a.txt', b'b.txt'], auditor=lambda name: None) |
|
1071 | >>> m1 = match(util.localpath(b'/root/d/e'), b'f', [b'../a.txt', b'b.txt'], auditor=lambda name: None) | |
1052 | >>> m2 = prefixdirmatcher(b'd/e', m1) |
|
1072 | >>> m2 = prefixdirmatcher(b'd/e', m1) | |
1053 | >>> m2(b'a.txt') |
|
1073 | >>> m2(b'a.txt') | |
1054 | False |
|
1074 | False | |
1055 | >>> m2(b'd/e/a.txt') |
|
1075 | >>> m2(b'd/e/a.txt') | |
1056 | True |
|
1076 | True | |
1057 | >>> m2(b'd/e/b.txt') |
|
1077 | >>> m2(b'd/e/b.txt') | |
1058 | False |
|
1078 | False | |
1059 | >>> m2.files() |
|
1079 | >>> m2.files() | |
1060 | ['d/e/a.txt', 'd/e/f/b.txt'] |
|
1080 | ['d/e/a.txt', 'd/e/f/b.txt'] | |
1061 | >>> m2.exact(b'd/e/a.txt') |
|
1081 | >>> m2.exact(b'd/e/a.txt') | |
1062 | True |
|
1082 | True | |
1063 | >>> m2.visitdir(b'd') |
|
1083 | >>> m2.visitdir(b'd') | |
1064 | True |
|
1084 | True | |
1065 | >>> m2.visitdir(b'd/e') |
|
1085 | >>> m2.visitdir(b'd/e') | |
1066 | True |
|
1086 | True | |
1067 | >>> m2.visitdir(b'd/e/f') |
|
1087 | >>> m2.visitdir(b'd/e/f') | |
1068 | True |
|
1088 | True | |
1069 | >>> m2.visitdir(b'd/e/g') |
|
1089 | >>> m2.visitdir(b'd/e/g') | |
1070 | False |
|
1090 | False | |
1071 | >>> m2.visitdir(b'd/ef') |
|
1091 | >>> m2.visitdir(b'd/ef') | |
1072 | False |
|
1092 | False | |
1073 | """ |
|
1093 | """ | |
1074 |
|
1094 | |||
1075 | def __init__(self, path, matcher, badfn=None): |
|
1095 | def __init__(self, path, matcher, badfn=None): | |
1076 | super(prefixdirmatcher, self).__init__(badfn) |
|
1096 | super(prefixdirmatcher, self).__init__(badfn) | |
1077 | if not path: |
|
1097 | if not path: | |
1078 | raise error.ProgrammingError(b'prefix path must not be empty') |
|
1098 | raise error.ProgrammingError(b'prefix path must not be empty') | |
1079 | self._path = path |
|
1099 | self._path = path | |
1080 | self._pathprefix = path + b'/' |
|
1100 | self._pathprefix = path + b'/' | |
1081 | self._matcher = matcher |
|
1101 | self._matcher = matcher | |
1082 |
|
1102 | |||
1083 | @propertycache |
|
1103 | @propertycache | |
1084 | def _files(self): |
|
1104 | def _files(self): | |
1085 | return [self._pathprefix + f for f in self._matcher._files] |
|
1105 | return [self._pathprefix + f for f in self._matcher._files] | |
1086 |
|
1106 | |||
1087 | def matchfn(self, f): |
|
1107 | def matchfn(self, f): | |
1088 | if not f.startswith(self._pathprefix): |
|
1108 | if not f.startswith(self._pathprefix): | |
1089 | return False |
|
1109 | return False | |
1090 | return self._matcher.matchfn(f[len(self._pathprefix) :]) |
|
1110 | return self._matcher.matchfn(f[len(self._pathprefix) :]) | |
1091 |
|
1111 | |||
1092 | @propertycache |
|
1112 | @propertycache | |
1093 | def _pathdirs(self): |
|
1113 | def _pathdirs(self): | |
1094 | return set(pathutil.finddirs(self._path)) |
|
1114 | return set(pathutil.finddirs(self._path)) | |
1095 |
|
1115 | |||
1096 | def visitdir(self, dir): |
|
1116 | def visitdir(self, dir): | |
1097 | if dir == self._path: |
|
1117 | if dir == self._path: | |
1098 | return self._matcher.visitdir(b'') |
|
1118 | return self._matcher.visitdir(b'') | |
1099 | if dir.startswith(self._pathprefix): |
|
1119 | if dir.startswith(self._pathprefix): | |
1100 | return self._matcher.visitdir(dir[len(self._pathprefix) :]) |
|
1120 | return self._matcher.visitdir(dir[len(self._pathprefix) :]) | |
1101 | return dir in self._pathdirs |
|
1121 | return dir in self._pathdirs | |
1102 |
|
1122 | |||
1103 | def visitchildrenset(self, dir): |
|
1123 | def visitchildrenset(self, dir): | |
1104 | if dir == self._path: |
|
1124 | if dir == self._path: | |
1105 | return self._matcher.visitchildrenset(b'') |
|
1125 | return self._matcher.visitchildrenset(b'') | |
1106 | if dir.startswith(self._pathprefix): |
|
1126 | if dir.startswith(self._pathprefix): | |
1107 | return self._matcher.visitchildrenset(dir[len(self._pathprefix) :]) |
|
1127 | return self._matcher.visitchildrenset(dir[len(self._pathprefix) :]) | |
1108 | if dir in self._pathdirs: |
|
1128 | if dir in self._pathdirs: | |
1109 | return b'this' |
|
1129 | return b'this' | |
1110 | return set() |
|
1130 | return set() | |
1111 |
|
1131 | |||
1112 | def isexact(self): |
|
1132 | def isexact(self): | |
1113 | return self._matcher.isexact() |
|
1133 | return self._matcher.isexact() | |
1114 |
|
1134 | |||
1115 | def prefix(self): |
|
1135 | def prefix(self): | |
1116 | return self._matcher.prefix() |
|
1136 | return self._matcher.prefix() | |
1117 |
|
1137 | |||
1118 | @encoding.strmethod |
|
1138 | @encoding.strmethod | |
1119 | def __repr__(self): |
|
1139 | def __repr__(self): | |
1120 | return b'<prefixdirmatcher path=%r, matcher=%r>' % ( |
|
1140 | return b'<prefixdirmatcher path=%r, matcher=%r>' % ( | |
1121 | pycompat.bytestr(self._path), |
|
1141 | pycompat.bytestr(self._path), | |
1122 | self._matcher, |
|
1142 | self._matcher, | |
1123 | ) |
|
1143 | ) | |
1124 |
|
1144 | |||
1125 |
|
1145 | |||
1126 | class unionmatcher(basematcher): |
|
1146 | class unionmatcher(basematcher): | |
1127 | """A matcher that is the union of several matchers. |
|
1147 | """A matcher that is the union of several matchers. | |
1128 |
|
1148 | |||
1129 | The non-matching-attributes (bad, traversedir) are taken from the first |
|
1149 | The non-matching-attributes (bad, traversedir) are taken from the first | |
1130 | matcher. |
|
1150 | matcher. | |
1131 | """ |
|
1151 | """ | |
1132 |
|
1152 | |||
1133 | def __init__(self, matchers): |
|
1153 | def __init__(self, matchers): | |
1134 | m1 = matchers[0] |
|
1154 | m1 = matchers[0] | |
1135 | super(unionmatcher, self).__init__() |
|
1155 | super(unionmatcher, self).__init__() | |
1136 | self.traversedir = m1.traversedir |
|
1156 | self.traversedir = m1.traversedir | |
1137 | self._matchers = matchers |
|
1157 | self._matchers = matchers | |
1138 |
|
1158 | |||
1139 | def matchfn(self, f): |
|
1159 | def matchfn(self, f): | |
1140 | for match in self._matchers: |
|
1160 | for match in self._matchers: | |
1141 | if match(f): |
|
1161 | if match(f): | |
1142 | return True |
|
1162 | return True | |
1143 | return False |
|
1163 | return False | |
1144 |
|
1164 | |||
1145 | def visitdir(self, dir): |
|
1165 | def visitdir(self, dir): | |
1146 | r = False |
|
1166 | r = False | |
1147 | for m in self._matchers: |
|
1167 | for m in self._matchers: | |
1148 | v = m.visitdir(dir) |
|
1168 | v = m.visitdir(dir) | |
1149 | if v == b'all': |
|
1169 | if v == b'all': | |
1150 | return v |
|
1170 | return v | |
1151 | r |= v |
|
1171 | r |= v | |
1152 | return r |
|
1172 | return r | |
1153 |
|
1173 | |||
1154 | def visitchildrenset(self, dir): |
|
1174 | def visitchildrenset(self, dir): | |
1155 | r = set() |
|
1175 | r = set() | |
1156 | this = False |
|
1176 | this = False | |
1157 | for m in self._matchers: |
|
1177 | for m in self._matchers: | |
1158 | v = m.visitchildrenset(dir) |
|
1178 | v = m.visitchildrenset(dir) | |
1159 | if not v: |
|
1179 | if not v: | |
1160 | continue |
|
1180 | continue | |
1161 | if v == b'all': |
|
1181 | if v == b'all': | |
1162 | return v |
|
1182 | return v | |
1163 | if this or v == b'this': |
|
1183 | if this or v == b'this': | |
1164 | this = True |
|
1184 | this = True | |
1165 | # don't break, we might have an 'all' in here. |
|
1185 | # don't break, we might have an 'all' in here. | |
1166 | continue |
|
1186 | continue | |
1167 | assert isinstance(v, set) |
|
1187 | assert isinstance(v, set) | |
1168 | r = r.union(v) |
|
1188 | r = r.union(v) | |
1169 | if this: |
|
1189 | if this: | |
1170 | return b'this' |
|
1190 | return b'this' | |
1171 | return r |
|
1191 | return r | |
1172 |
|
1192 | |||
1173 | @encoding.strmethod |
|
1193 | @encoding.strmethod | |
1174 | def __repr__(self): |
|
1194 | def __repr__(self): | |
1175 | return b'<unionmatcher matchers=%r>' % self._matchers |
|
1195 | return b'<unionmatcher matchers=%r>' % self._matchers | |
1176 |
|
1196 | |||
1177 |
|
1197 | |||
1178 | def patkind(pattern, default=None): |
|
1198 | def patkind(pattern, default=None): | |
1179 | r"""If pattern is 'kind:pat' with a known kind, return kind. |
|
1199 | r"""If pattern is 'kind:pat' with a known kind, return kind. | |
1180 |
|
1200 | |||
1181 | >>> patkind(br're:.*\.c$') |
|
1201 | >>> patkind(br're:.*\.c$') | |
1182 | 're' |
|
1202 | 're' | |
1183 | >>> patkind(b'glob:*.c') |
|
1203 | >>> patkind(b'glob:*.c') | |
1184 | 'glob' |
|
1204 | 'glob' | |
1185 | >>> patkind(b'relpath:test.py') |
|
1205 | >>> patkind(b'relpath:test.py') | |
1186 | 'relpath' |
|
1206 | 'relpath' | |
1187 | >>> patkind(b'main.py') |
|
1207 | >>> patkind(b'main.py') | |
1188 | >>> patkind(b'main.py', default=b're') |
|
1208 | >>> patkind(b'main.py', default=b're') | |
1189 | 're' |
|
1209 | 're' | |
1190 | """ |
|
1210 | """ | |
1191 | return _patsplit(pattern, default)[0] |
|
1211 | return _patsplit(pattern, default)[0] | |
1192 |
|
1212 | |||
1193 |
|
1213 | |||
1194 | def _patsplit(pattern, default): |
|
1214 | def _patsplit(pattern, default): | |
1195 | """Split a string into the optional pattern kind prefix and the actual |
|
1215 | """Split a string into the optional pattern kind prefix and the actual | |
1196 | pattern.""" |
|
1216 | pattern.""" | |
1197 | if b':' in pattern: |
|
1217 | if b':' in pattern: | |
1198 | kind, pat = pattern.split(b':', 1) |
|
1218 | kind, pat = pattern.split(b':', 1) | |
1199 | if kind in allpatternkinds: |
|
1219 | if kind in allpatternkinds: | |
1200 | return kind, pat |
|
1220 | return kind, pat | |
1201 | return default, pattern |
|
1221 | return default, pattern | |
1202 |
|
1222 | |||
1203 |
|
1223 | |||
1204 | def _globre(pat): |
|
1224 | def _globre(pat): | |
1205 | r"""Convert an extended glob string to a regexp string. |
|
1225 | r"""Convert an extended glob string to a regexp string. | |
1206 |
|
1226 | |||
1207 | >>> from . import pycompat |
|
1227 | >>> from . import pycompat | |
1208 | >>> def bprint(s): |
|
1228 | >>> def bprint(s): | |
1209 | ... print(pycompat.sysstr(s)) |
|
1229 | ... print(pycompat.sysstr(s)) | |
1210 | >>> bprint(_globre(br'?')) |
|
1230 | >>> bprint(_globre(br'?')) | |
1211 | . |
|
1231 | . | |
1212 | >>> bprint(_globre(br'*')) |
|
1232 | >>> bprint(_globre(br'*')) | |
1213 | [^/]* |
|
1233 | [^/]* | |
1214 | >>> bprint(_globre(br'**')) |
|
1234 | >>> bprint(_globre(br'**')) | |
1215 | .* |
|
1235 | .* | |
1216 | >>> bprint(_globre(br'**/a')) |
|
1236 | >>> bprint(_globre(br'**/a')) | |
1217 | (?:.*/)?a |
|
1237 | (?:.*/)?a | |
1218 | >>> bprint(_globre(br'a/**/b')) |
|
1238 | >>> bprint(_globre(br'a/**/b')) | |
1219 | a/(?:.*/)?b |
|
1239 | a/(?:.*/)?b | |
1220 | >>> bprint(_globre(br'[a*?!^][^b][!c]')) |
|
1240 | >>> bprint(_globre(br'[a*?!^][^b][!c]')) | |
1221 | [a*?!^][\^b][^c] |
|
1241 | [a*?!^][\^b][^c] | |
1222 | >>> bprint(_globre(br'{a,b}')) |
|
1242 | >>> bprint(_globre(br'{a,b}')) | |
1223 | (?:a|b) |
|
1243 | (?:a|b) | |
1224 | >>> bprint(_globre(br'.\*\?')) |
|
1244 | >>> bprint(_globre(br'.\*\?')) | |
1225 | \.\*\? |
|
1245 | \.\*\? | |
1226 | """ |
|
1246 | """ | |
1227 | i, n = 0, len(pat) |
|
1247 | i, n = 0, len(pat) | |
1228 | res = b'' |
|
1248 | res = b'' | |
1229 | group = 0 |
|
1249 | group = 0 | |
1230 | escape = util.stringutil.regexbytesescapemap.get |
|
1250 | escape = util.stringutil.regexbytesescapemap.get | |
1231 |
|
1251 | |||
1232 | def peek(): |
|
1252 | def peek(): | |
1233 | return i < n and pat[i : i + 1] |
|
1253 | return i < n and pat[i : i + 1] | |
1234 |
|
1254 | |||
1235 | while i < n: |
|
1255 | while i < n: | |
1236 | c = pat[i : i + 1] |
|
1256 | c = pat[i : i + 1] | |
1237 | i += 1 |
|
1257 | i += 1 | |
1238 | if c not in b'*?[{},\\': |
|
1258 | if c not in b'*?[{},\\': | |
1239 | res += escape(c, c) |
|
1259 | res += escape(c, c) | |
1240 | elif c == b'*': |
|
1260 | elif c == b'*': | |
1241 | if peek() == b'*': |
|
1261 | if peek() == b'*': | |
1242 | i += 1 |
|
1262 | i += 1 | |
1243 | if peek() == b'/': |
|
1263 | if peek() == b'/': | |
1244 | i += 1 |
|
1264 | i += 1 | |
1245 | res += b'(?:.*/)?' |
|
1265 | res += b'(?:.*/)?' | |
1246 | else: |
|
1266 | else: | |
1247 | res += b'.*' |
|
1267 | res += b'.*' | |
1248 | else: |
|
1268 | else: | |
1249 | res += b'[^/]*' |
|
1269 | res += b'[^/]*' | |
1250 | elif c == b'?': |
|
1270 | elif c == b'?': | |
1251 | res += b'.' |
|
1271 | res += b'.' | |
1252 | elif c == b'[': |
|
1272 | elif c == b'[': | |
1253 | j = i |
|
1273 | j = i | |
1254 | if j < n and pat[j : j + 1] in b'!]': |
|
1274 | if j < n and pat[j : j + 1] in b'!]': | |
1255 | j += 1 |
|
1275 | j += 1 | |
1256 | while j < n and pat[j : j + 1] != b']': |
|
1276 | while j < n and pat[j : j + 1] != b']': | |
1257 | j += 1 |
|
1277 | j += 1 | |
1258 | if j >= n: |
|
1278 | if j >= n: | |
1259 | res += b'\\[' |
|
1279 | res += b'\\[' | |
1260 | else: |
|
1280 | else: | |
1261 | stuff = pat[i:j].replace(b'\\', b'\\\\') |
|
1281 | stuff = pat[i:j].replace(b'\\', b'\\\\') | |
1262 | i = j + 1 |
|
1282 | i = j + 1 | |
1263 | if stuff[0:1] == b'!': |
|
1283 | if stuff[0:1] == b'!': | |
1264 | stuff = b'^' + stuff[1:] |
|
1284 | stuff = b'^' + stuff[1:] | |
1265 | elif stuff[0:1] == b'^': |
|
1285 | elif stuff[0:1] == b'^': | |
1266 | stuff = b'\\' + stuff |
|
1286 | stuff = b'\\' + stuff | |
1267 | res = b'%s[%s]' % (res, stuff) |
|
1287 | res = b'%s[%s]' % (res, stuff) | |
1268 | elif c == b'{': |
|
1288 | elif c == b'{': | |
1269 | group += 1 |
|
1289 | group += 1 | |
1270 | res += b'(?:' |
|
1290 | res += b'(?:' | |
1271 | elif c == b'}' and group: |
|
1291 | elif c == b'}' and group: | |
1272 | res += b')' |
|
1292 | res += b')' | |
1273 | group -= 1 |
|
1293 | group -= 1 | |
1274 | elif c == b',' and group: |
|
1294 | elif c == b',' and group: | |
1275 | res += b'|' |
|
1295 | res += b'|' | |
1276 | elif c == b'\\': |
|
1296 | elif c == b'\\': | |
1277 | p = peek() |
|
1297 | p = peek() | |
1278 | if p: |
|
1298 | if p: | |
1279 | i += 1 |
|
1299 | i += 1 | |
1280 | res += escape(p, p) |
|
1300 | res += escape(p, p) | |
1281 | else: |
|
1301 | else: | |
1282 | res += escape(c, c) |
|
1302 | res += escape(c, c) | |
1283 | else: |
|
1303 | else: | |
1284 | res += escape(c, c) |
|
1304 | res += escape(c, c) | |
1285 | return res |
|
1305 | return res | |
1286 |
|
1306 | |||
1287 |
|
1307 | |||
1288 | def _regex(kind, pat, globsuffix): |
|
1308 | def _regex(kind, pat, globsuffix): | |
1289 | """Convert a (normalized) pattern of any kind into a |
|
1309 | """Convert a (normalized) pattern of any kind into a | |
1290 | regular expression. |
|
1310 | regular expression. | |
1291 | globsuffix is appended to the regexp of globs.""" |
|
1311 | globsuffix is appended to the regexp of globs.""" | |
1292 | if not pat and kind in (b'glob', b'relpath'): |
|
1312 | if not pat and kind in (b'glob', b'relpath'): | |
1293 | return b'' |
|
1313 | return b'' | |
1294 | if kind == b're': |
|
1314 | if kind == b're': | |
1295 | return pat |
|
1315 | return pat | |
1296 | if kind in (b'path', b'relpath'): |
|
1316 | if kind in (b'path', b'relpath'): | |
1297 | if pat == b'.': |
|
1317 | if pat == b'.': | |
1298 | return b'' |
|
1318 | return b'' | |
1299 | return util.stringutil.reescape(pat) + b'(?:/|$)' |
|
1319 | return util.stringutil.reescape(pat) + b'(?:/|$)' | |
1300 | if kind == b'rootfilesin': |
|
1320 | if kind == b'rootfilesin': | |
1301 | if pat == b'.': |
|
1321 | if pat == b'.': | |
1302 | escaped = b'' |
|
1322 | escaped = b'' | |
1303 | else: |
|
1323 | else: | |
1304 | # Pattern is a directory name. |
|
1324 | # Pattern is a directory name. | |
1305 | escaped = util.stringutil.reescape(pat) + b'/' |
|
1325 | escaped = util.stringutil.reescape(pat) + b'/' | |
1306 | # Anything after the pattern must be a non-directory. |
|
1326 | # Anything after the pattern must be a non-directory. | |
1307 | return escaped + b'[^/]+$' |
|
1327 | return escaped + b'[^/]+$' | |
1308 | if kind == b'relglob': |
|
1328 | if kind == b'relglob': | |
1309 | globre = _globre(pat) |
|
1329 | globre = _globre(pat) | |
1310 | if globre.startswith(b'[^/]*'): |
|
1330 | if globre.startswith(b'[^/]*'): | |
1311 | # When pat has the form *XYZ (common), make the returned regex more |
|
1331 | # When pat has the form *XYZ (common), make the returned regex more | |
1312 | # legible by returning the regex for **XYZ instead of **/*XYZ. |
|
1332 | # legible by returning the regex for **XYZ instead of **/*XYZ. | |
1313 | return b'.*' + globre[len(b'[^/]*') :] + globsuffix |
|
1333 | return b'.*' + globre[len(b'[^/]*') :] + globsuffix | |
1314 | return b'(?:|.*/)' + globre + globsuffix |
|
1334 | return b'(?:|.*/)' + globre + globsuffix | |
1315 | if kind == b'relre': |
|
1335 | if kind == b'relre': | |
1316 | if pat.startswith(b'^'): |
|
1336 | if pat.startswith(b'^'): | |
1317 | return pat |
|
1337 | return pat | |
1318 | return b'.*' + pat |
|
1338 | return b'.*' + pat | |
1319 | if kind in (b'glob', b'rootglob'): |
|
1339 | if kind in (b'glob', b'rootglob'): | |
1320 | return _globre(pat) + globsuffix |
|
1340 | return _globre(pat) + globsuffix | |
1321 | raise error.ProgrammingError(b'not a regex pattern: %s:%s' % (kind, pat)) |
|
1341 | raise error.ProgrammingError(b'not a regex pattern: %s:%s' % (kind, pat)) | |
1322 |
|
1342 | |||
1323 |
|
1343 | |||
1324 | def _buildmatch(kindpats, globsuffix, root): |
|
1344 | def _buildmatch(kindpats, globsuffix, root): | |
1325 | """Return regexp string and a matcher function for kindpats. |
|
1345 | """Return regexp string and a matcher function for kindpats. | |
1326 | globsuffix is appended to the regexp of globs.""" |
|
1346 | globsuffix is appended to the regexp of globs.""" | |
1327 | matchfuncs = [] |
|
1347 | matchfuncs = [] | |
1328 |
|
1348 | |||
1329 | subincludes, kindpats = _expandsubinclude(kindpats, root) |
|
1349 | subincludes, kindpats = _expandsubinclude(kindpats, root) | |
1330 | if subincludes: |
|
1350 | if subincludes: | |
1331 | submatchers = {} |
|
1351 | submatchers = {} | |
1332 |
|
1352 | |||
1333 | def matchsubinclude(f): |
|
1353 | def matchsubinclude(f): | |
1334 | for prefix, matcherargs in subincludes: |
|
1354 | for prefix, matcherargs in subincludes: | |
1335 | if f.startswith(prefix): |
|
1355 | if f.startswith(prefix): | |
1336 | mf = submatchers.get(prefix) |
|
1356 | mf = submatchers.get(prefix) | |
1337 | if mf is None: |
|
1357 | if mf is None: | |
1338 | mf = match(*matcherargs) |
|
1358 | mf = match(*matcherargs) | |
1339 | submatchers[prefix] = mf |
|
1359 | submatchers[prefix] = mf | |
1340 |
|
1360 | |||
1341 | if mf(f[len(prefix) :]): |
|
1361 | if mf(f[len(prefix) :]): | |
1342 | return True |
|
1362 | return True | |
1343 | return False |
|
1363 | return False | |
1344 |
|
1364 | |||
1345 | matchfuncs.append(matchsubinclude) |
|
1365 | matchfuncs.append(matchsubinclude) | |
1346 |
|
1366 | |||
1347 | regex = b'' |
|
1367 | regex = b'' | |
1348 | if kindpats: |
|
1368 | if kindpats: | |
1349 | if all(k == b'rootfilesin' for k, p, s in kindpats): |
|
1369 | if all(k == b'rootfilesin' for k, p, s in kindpats): | |
1350 | dirs = {p for k, p, s in kindpats} |
|
1370 | dirs = {p for k, p, s in kindpats} | |
1351 |
|
1371 | |||
1352 | def mf(f): |
|
1372 | def mf(f): | |
1353 | i = f.rfind(b'/') |
|
1373 | i = f.rfind(b'/') | |
1354 | if i >= 0: |
|
1374 | if i >= 0: | |
1355 | dir = f[:i] |
|
1375 | dir = f[:i] | |
1356 | else: |
|
1376 | else: | |
1357 | dir = b'.' |
|
1377 | dir = b'.' | |
1358 | return dir in dirs |
|
1378 | return dir in dirs | |
1359 |
|
1379 | |||
1360 | regex = b'rootfilesin: %s' % stringutil.pprint(list(sorted(dirs))) |
|
1380 | regex = b'rootfilesin: %s' % stringutil.pprint(list(sorted(dirs))) | |
1361 | matchfuncs.append(mf) |
|
1381 | matchfuncs.append(mf) | |
1362 | else: |
|
1382 | else: | |
1363 | regex, mf = _buildregexmatch(kindpats, globsuffix) |
|
1383 | regex, mf = _buildregexmatch(kindpats, globsuffix) | |
1364 | matchfuncs.append(mf) |
|
1384 | matchfuncs.append(mf) | |
1365 |
|
1385 | |||
1366 | if len(matchfuncs) == 1: |
|
1386 | if len(matchfuncs) == 1: | |
1367 | return regex, matchfuncs[0] |
|
1387 | return regex, matchfuncs[0] | |
1368 | else: |
|
1388 | else: | |
1369 | return regex, lambda f: any(mf(f) for mf in matchfuncs) |
|
1389 | return regex, lambda f: any(mf(f) for mf in matchfuncs) | |
1370 |
|
1390 | |||
1371 |
|
1391 | |||
1372 | MAX_RE_SIZE = 20000 |
|
1392 | MAX_RE_SIZE = 20000 | |
1373 |
|
1393 | |||
1374 |
|
1394 | |||
1375 | def _joinregexes(regexps): |
|
1395 | def _joinregexes(regexps): | |
1376 | """gather multiple regular expressions into a single one""" |
|
1396 | """gather multiple regular expressions into a single one""" | |
1377 | return b'|'.join(regexps) |
|
1397 | return b'|'.join(regexps) | |
1378 |
|
1398 | |||
1379 |
|
1399 | |||
1380 | def _buildregexmatch(kindpats, globsuffix): |
|
1400 | def _buildregexmatch(kindpats, globsuffix): | |
1381 | """Build a match function from a list of kinds and kindpats, |
|
1401 | """Build a match function from a list of kinds and kindpats, | |
1382 | return regexp string and a matcher function. |
|
1402 | return regexp string and a matcher function. | |
1383 |
|
1403 | |||
1384 | Test too large input |
|
1404 | Test too large input | |
1385 | >>> _buildregexmatch([ |
|
1405 | >>> _buildregexmatch([ | |
1386 | ... (b'relglob', b'?' * MAX_RE_SIZE, b'') |
|
1406 | ... (b'relglob', b'?' * MAX_RE_SIZE, b'') | |
1387 | ... ], b'$') |
|
1407 | ... ], b'$') | |
1388 | Traceback (most recent call last): |
|
1408 | Traceback (most recent call last): | |
1389 | ... |
|
1409 | ... | |
1390 | Abort: matcher pattern is too long (20009 bytes) |
|
1410 | Abort: matcher pattern is too long (20009 bytes) | |
1391 | """ |
|
1411 | """ | |
1392 | try: |
|
1412 | try: | |
1393 | allgroups = [] |
|
1413 | allgroups = [] | |
1394 | regexps = [_regex(k, p, globsuffix) for (k, p, s) in kindpats] |
|
1414 | regexps = [_regex(k, p, globsuffix) for (k, p, s) in kindpats] | |
1395 | fullregexp = _joinregexes(regexps) |
|
1415 | fullregexp = _joinregexes(regexps) | |
1396 |
|
1416 | |||
1397 | startidx = 0 |
|
1417 | startidx = 0 | |
1398 | groupsize = 0 |
|
1418 | groupsize = 0 | |
1399 | for idx, r in enumerate(regexps): |
|
1419 | for idx, r in enumerate(regexps): | |
1400 | piecesize = len(r) |
|
1420 | piecesize = len(r) | |
1401 | if piecesize > MAX_RE_SIZE: |
|
1421 | if piecesize > MAX_RE_SIZE: | |
1402 | msg = _(b"matcher pattern is too long (%d bytes)") % piecesize |
|
1422 | msg = _(b"matcher pattern is too long (%d bytes)") % piecesize | |
1403 | raise error.Abort(msg) |
|
1423 | raise error.Abort(msg) | |
1404 | elif (groupsize + piecesize) > MAX_RE_SIZE: |
|
1424 | elif (groupsize + piecesize) > MAX_RE_SIZE: | |
1405 | group = regexps[startidx:idx] |
|
1425 | group = regexps[startidx:idx] | |
1406 | allgroups.append(_joinregexes(group)) |
|
1426 | allgroups.append(_joinregexes(group)) | |
1407 | startidx = idx |
|
1427 | startidx = idx | |
1408 | groupsize = 0 |
|
1428 | groupsize = 0 | |
1409 | groupsize += piecesize + 1 |
|
1429 | groupsize += piecesize + 1 | |
1410 |
|
1430 | |||
1411 | if startidx == 0: |
|
1431 | if startidx == 0: | |
1412 | matcher = _rematcher(fullregexp) |
|
1432 | matcher = _rematcher(fullregexp) | |
1413 | func = lambda s: bool(matcher(s)) |
|
1433 | func = lambda s: bool(matcher(s)) | |
1414 | else: |
|
1434 | else: | |
1415 | group = regexps[startidx:] |
|
1435 | group = regexps[startidx:] | |
1416 | allgroups.append(_joinregexes(group)) |
|
1436 | allgroups.append(_joinregexes(group)) | |
1417 | allmatchers = [_rematcher(g) for g in allgroups] |
|
1437 | allmatchers = [_rematcher(g) for g in allgroups] | |
1418 | func = lambda s: any(m(s) for m in allmatchers) |
|
1438 | func = lambda s: any(m(s) for m in allmatchers) | |
1419 | return fullregexp, func |
|
1439 | return fullregexp, func | |
1420 | except re.error: |
|
1440 | except re.error: | |
1421 | for k, p, s in kindpats: |
|
1441 | for k, p, s in kindpats: | |
1422 | try: |
|
1442 | try: | |
1423 | _rematcher(_regex(k, p, globsuffix)) |
|
1443 | _rematcher(_regex(k, p, globsuffix)) | |
1424 | except re.error: |
|
1444 | except re.error: | |
1425 | if s: |
|
1445 | if s: | |
1426 | raise error.Abort( |
|
1446 | raise error.Abort( | |
1427 | _(b"%s: invalid pattern (%s): %s") % (s, k, p) |
|
1447 | _(b"%s: invalid pattern (%s): %s") % (s, k, p) | |
1428 | ) |
|
1448 | ) | |
1429 | else: |
|
1449 | else: | |
1430 | raise error.Abort(_(b"invalid pattern (%s): %s") % (k, p)) |
|
1450 | raise error.Abort(_(b"invalid pattern (%s): %s") % (k, p)) | |
1431 | raise error.Abort(_(b"invalid pattern")) |
|
1451 | raise error.Abort(_(b"invalid pattern")) | |
1432 |
|
1452 | |||
1433 |
|
1453 | |||
1434 | def _patternrootsanddirs(kindpats): |
|
1454 | def _patternrootsanddirs(kindpats): | |
1435 | """Returns roots and directories corresponding to each pattern. |
|
1455 | """Returns roots and directories corresponding to each pattern. | |
1436 |
|
1456 | |||
1437 | This calculates the roots and directories exactly matching the patterns and |
|
1457 | This calculates the roots and directories exactly matching the patterns and | |
1438 | returns a tuple of (roots, dirs) for each. It does not return other |
|
1458 | returns a tuple of (roots, dirs) for each. It does not return other | |
1439 | directories which may also need to be considered, like the parent |
|
1459 | directories which may also need to be considered, like the parent | |
1440 | directories. |
|
1460 | directories. | |
1441 | """ |
|
1461 | """ | |
1442 | r = [] |
|
1462 | r = [] | |
1443 | d = [] |
|
1463 | d = [] | |
1444 | for kind, pat, source in kindpats: |
|
1464 | for kind, pat, source in kindpats: | |
1445 | if kind in (b'glob', b'rootglob'): # find the non-glob prefix |
|
1465 | if kind in (b'glob', b'rootglob'): # find the non-glob prefix | |
1446 | root = [] |
|
1466 | root = [] | |
1447 | for p in pat.split(b'/'): |
|
1467 | for p in pat.split(b'/'): | |
1448 | if b'[' in p or b'{' in p or b'*' in p or b'?' in p: |
|
1468 | if b'[' in p or b'{' in p or b'*' in p or b'?' in p: | |
1449 | break |
|
1469 | break | |
1450 | root.append(p) |
|
1470 | root.append(p) | |
1451 | r.append(b'/'.join(root)) |
|
1471 | r.append(b'/'.join(root)) | |
1452 | elif kind in (b'relpath', b'path'): |
|
1472 | elif kind in (b'relpath', b'path'): | |
1453 | if pat == b'.': |
|
1473 | if pat == b'.': | |
1454 | pat = b'' |
|
1474 | pat = b'' | |
1455 | r.append(pat) |
|
1475 | r.append(pat) | |
1456 | elif kind in (b'rootfilesin',): |
|
1476 | elif kind in (b'rootfilesin',): | |
1457 | if pat == b'.': |
|
1477 | if pat == b'.': | |
1458 | pat = b'' |
|
1478 | pat = b'' | |
1459 | d.append(pat) |
|
1479 | d.append(pat) | |
1460 | else: # relglob, re, relre |
|
1480 | else: # relglob, re, relre | |
1461 | r.append(b'') |
|
1481 | r.append(b'') | |
1462 | return r, d |
|
1482 | return r, d | |
1463 |
|
1483 | |||
1464 |
|
1484 | |||
1465 | def _roots(kindpats): |
|
1485 | def _roots(kindpats): | |
1466 | '''Returns root directories to match recursively from the given patterns.''' |
|
1486 | '''Returns root directories to match recursively from the given patterns.''' | |
1467 | roots, dirs = _patternrootsanddirs(kindpats) |
|
1487 | roots, dirs = _patternrootsanddirs(kindpats) | |
1468 | return roots |
|
1488 | return roots | |
1469 |
|
1489 | |||
1470 |
|
1490 | |||
1471 | def _rootsdirsandparents(kindpats): |
|
1491 | def _rootsdirsandparents(kindpats): | |
1472 | """Returns roots and exact directories from patterns. |
|
1492 | """Returns roots and exact directories from patterns. | |
1473 |
|
1493 | |||
1474 | `roots` are directories to match recursively, `dirs` should |
|
1494 | `roots` are directories to match recursively, `dirs` should | |
1475 | be matched non-recursively, and `parents` are the implicitly required |
|
1495 | be matched non-recursively, and `parents` are the implicitly required | |
1476 | directories to walk to items in either roots or dirs. |
|
1496 | directories to walk to items in either roots or dirs. | |
1477 |
|
1497 | |||
1478 | Returns a tuple of (roots, dirs, parents). |
|
1498 | Returns a tuple of (roots, dirs, parents). | |
1479 |
|
1499 | |||
1480 | >>> r = _rootsdirsandparents( |
|
1500 | >>> r = _rootsdirsandparents( | |
1481 | ... [(b'glob', b'g/h/*', b''), (b'glob', b'g/h', b''), |
|
1501 | ... [(b'glob', b'g/h/*', b''), (b'glob', b'g/h', b''), | |
1482 | ... (b'glob', b'g*', b'')]) |
|
1502 | ... (b'glob', b'g*', b'')]) | |
1483 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1503 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1484 | (['g/h', 'g/h', ''], []) ['', 'g'] |
|
1504 | (['g/h', 'g/h', ''], []) ['', 'g'] | |
1485 | >>> r = _rootsdirsandparents( |
|
1505 | >>> r = _rootsdirsandparents( | |
1486 | ... [(b'rootfilesin', b'g/h', b''), (b'rootfilesin', b'', b'')]) |
|
1506 | ... [(b'rootfilesin', b'g/h', b''), (b'rootfilesin', b'', b'')]) | |
1487 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1507 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1488 | ([], ['g/h', '']) ['', 'g'] |
|
1508 | ([], ['g/h', '']) ['', 'g'] | |
1489 | >>> r = _rootsdirsandparents( |
|
1509 | >>> r = _rootsdirsandparents( | |
1490 | ... [(b'relpath', b'r', b''), (b'path', b'p/p', b''), |
|
1510 | ... [(b'relpath', b'r', b''), (b'path', b'p/p', b''), | |
1491 | ... (b'path', b'', b'')]) |
|
1511 | ... (b'path', b'', b'')]) | |
1492 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1512 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1493 | (['r', 'p/p', ''], []) ['', 'p'] |
|
1513 | (['r', 'p/p', ''], []) ['', 'p'] | |
1494 | >>> r = _rootsdirsandparents( |
|
1514 | >>> r = _rootsdirsandparents( | |
1495 | ... [(b'relglob', b'rg*', b''), (b're', b're/', b''), |
|
1515 | ... [(b'relglob', b'rg*', b''), (b're', b're/', b''), | |
1496 | ... (b'relre', b'rr', b'')]) |
|
1516 | ... (b'relre', b'rr', b'')]) | |
1497 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output |
|
1517 | >>> print(r[0:2], sorted(r[2])) # the set has an unstable output | |
1498 | (['', '', ''], []) [''] |
|
1518 | (['', '', ''], []) [''] | |
1499 | """ |
|
1519 | """ | |
1500 | r, d = _patternrootsanddirs(kindpats) |
|
1520 | r, d = _patternrootsanddirs(kindpats) | |
1501 |
|
1521 | |||
1502 | p = set() |
|
1522 | p = set() | |
1503 | # Add the parents as non-recursive/exact directories, since they must be |
|
1523 | # Add the parents as non-recursive/exact directories, since they must be | |
1504 | # scanned to get to either the roots or the other exact directories. |
|
1524 | # scanned to get to either the roots or the other exact directories. | |
1505 | p.update(pathutil.dirs(d)) |
|
1525 | p.update(pathutil.dirs(d)) | |
1506 | p.update(pathutil.dirs(r)) |
|
1526 | p.update(pathutil.dirs(r)) | |
1507 |
|
1527 | |||
1508 | # FIXME: all uses of this function convert these to sets, do so before |
|
1528 | # FIXME: all uses of this function convert these to sets, do so before | |
1509 | # returning. |
|
1529 | # returning. | |
1510 | # FIXME: all uses of this function do not need anything in 'roots' and |
|
1530 | # FIXME: all uses of this function do not need anything in 'roots' and | |
1511 | # 'dirs' to also be in 'parents', consider removing them before returning. |
|
1531 | # 'dirs' to also be in 'parents', consider removing them before returning. | |
1512 | return r, d, p |
|
1532 | return r, d, p | |
1513 |
|
1533 | |||
1514 |
|
1534 | |||
1515 | def _explicitfiles(kindpats): |
|
1535 | def _explicitfiles(kindpats): | |
1516 | """Returns the potential explicit filenames from the patterns. |
|
1536 | """Returns the potential explicit filenames from the patterns. | |
1517 |
|
1537 | |||
1518 | >>> _explicitfiles([(b'path', b'foo/bar', b'')]) |
|
1538 | >>> _explicitfiles([(b'path', b'foo/bar', b'')]) | |
1519 | ['foo/bar'] |
|
1539 | ['foo/bar'] | |
1520 | >>> _explicitfiles([(b'rootfilesin', b'foo/bar', b'')]) |
|
1540 | >>> _explicitfiles([(b'rootfilesin', b'foo/bar', b'')]) | |
1521 | [] |
|
1541 | [] | |
1522 | """ |
|
1542 | """ | |
1523 | # Keep only the pattern kinds where one can specify filenames (vs only |
|
1543 | # Keep only the pattern kinds where one can specify filenames (vs only | |
1524 | # directory names). |
|
1544 | # directory names). | |
1525 | filable = [kp for kp in kindpats if kp[0] not in (b'rootfilesin',)] |
|
1545 | filable = [kp for kp in kindpats if kp[0] not in (b'rootfilesin',)] | |
1526 | return _roots(filable) |
|
1546 | return _roots(filable) | |
1527 |
|
1547 | |||
1528 |
|
1548 | |||
1529 | def _prefix(kindpats): |
|
1549 | def _prefix(kindpats): | |
1530 | '''Whether all the patterns match a prefix (i.e. recursively)''' |
|
1550 | '''Whether all the patterns match a prefix (i.e. recursively)''' | |
1531 | for kind, pat, source in kindpats: |
|
1551 | for kind, pat, source in kindpats: | |
1532 | if kind not in (b'path', b'relpath'): |
|
1552 | if kind not in (b'path', b'relpath'): | |
1533 | return False |
|
1553 | return False | |
1534 | return True |
|
1554 | return True | |
1535 |
|
1555 | |||
1536 |
|
1556 | |||
1537 | _commentre = None |
|
1557 | _commentre = None | |
1538 |
|
1558 | |||
1539 |
|
1559 | |||
1540 | def readpatternfile(filepath, warn, sourceinfo=False): |
|
1560 | def readpatternfile(filepath, warn, sourceinfo=False): | |
1541 | """parse a pattern file, returning a list of |
|
1561 | """parse a pattern file, returning a list of | |
1542 | patterns. These patterns should be given to compile() |
|
1562 | patterns. These patterns should be given to compile() | |
1543 | to be validated and converted into a match function. |
|
1563 | to be validated and converted into a match function. | |
1544 |
|
1564 | |||
1545 | trailing white space is dropped. |
|
1565 | trailing white space is dropped. | |
1546 | the escape character is backslash. |
|
1566 | the escape character is backslash. | |
1547 | comments start with #. |
|
1567 | comments start with #. | |
1548 | empty lines are skipped. |
|
1568 | empty lines are skipped. | |
1549 |
|
1569 | |||
1550 | lines can be of the following formats: |
|
1570 | lines can be of the following formats: | |
1551 |
|
1571 | |||
1552 | syntax: regexp # defaults following lines to non-rooted regexps |
|
1572 | syntax: regexp # defaults following lines to non-rooted regexps | |
1553 | syntax: glob # defaults following lines to non-rooted globs |
|
1573 | syntax: glob # defaults following lines to non-rooted globs | |
1554 | re:pattern # non-rooted regular expression |
|
1574 | re:pattern # non-rooted regular expression | |
1555 | glob:pattern # non-rooted glob |
|
1575 | glob:pattern # non-rooted glob | |
1556 | rootglob:pat # rooted glob (same root as ^ in regexps) |
|
1576 | rootglob:pat # rooted glob (same root as ^ in regexps) | |
1557 | pattern # pattern of the current default type |
|
1577 | pattern # pattern of the current default type | |
1558 |
|
1578 | |||
1559 | if sourceinfo is set, returns a list of tuples: |
|
1579 | if sourceinfo is set, returns a list of tuples: | |
1560 | (pattern, lineno, originalline). |
|
1580 | (pattern, lineno, originalline). | |
1561 | This is useful to debug ignore patterns. |
|
1581 | This is useful to debug ignore patterns. | |
1562 | """ |
|
1582 | """ | |
1563 |
|
1583 | |||
1564 | syntaxes = { |
|
1584 | syntaxes = { | |
1565 | b're': b'relre:', |
|
1585 | b're': b'relre:', | |
1566 | b'regexp': b'relre:', |
|
1586 | b'regexp': b'relre:', | |
1567 | b'glob': b'relglob:', |
|
1587 | b'glob': b'relglob:', | |
1568 | b'rootglob': b'rootglob:', |
|
1588 | b'rootglob': b'rootglob:', | |
1569 | b'include': b'include', |
|
1589 | b'include': b'include', | |
1570 | b'subinclude': b'subinclude', |
|
1590 | b'subinclude': b'subinclude', | |
1571 | } |
|
1591 | } | |
1572 | syntax = b'relre:' |
|
1592 | syntax = b'relre:' | |
1573 | patterns = [] |
|
1593 | patterns = [] | |
1574 |
|
1594 | |||
1575 | fp = open(filepath, b'rb') |
|
1595 | fp = open(filepath, b'rb') | |
1576 | for lineno, line in enumerate(util.iterfile(fp), start=1): |
|
1596 | for lineno, line in enumerate(util.iterfile(fp), start=1): | |
1577 | if b"#" in line: |
|
1597 | if b"#" in line: | |
1578 | global _commentre |
|
1598 | global _commentre | |
1579 | if not _commentre: |
|
1599 | if not _commentre: | |
1580 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') |
|
1600 | _commentre = util.re.compile(br'((?:^|[^\\])(?:\\\\)*)#.*') | |
1581 | # remove comments prefixed by an even number of escapes |
|
1601 | # remove comments prefixed by an even number of escapes | |
1582 | m = _commentre.search(line) |
|
1602 | m = _commentre.search(line) | |
1583 | if m: |
|
1603 | if m: | |
1584 | line = line[: m.end(1)] |
|
1604 | line = line[: m.end(1)] | |
1585 | # fixup properly escaped comments that survived the above |
|
1605 | # fixup properly escaped comments that survived the above | |
1586 | line = line.replace(b"\\#", b"#") |
|
1606 | line = line.replace(b"\\#", b"#") | |
1587 | line = line.rstrip() |
|
1607 | line = line.rstrip() | |
1588 | if not line: |
|
1608 | if not line: | |
1589 | continue |
|
1609 | continue | |
1590 |
|
1610 | |||
1591 | if line.startswith(b'syntax:'): |
|
1611 | if line.startswith(b'syntax:'): | |
1592 | s = line[7:].strip() |
|
1612 | s = line[7:].strip() | |
1593 | try: |
|
1613 | try: | |
1594 | syntax = syntaxes[s] |
|
1614 | syntax = syntaxes[s] | |
1595 | except KeyError: |
|
1615 | except KeyError: | |
1596 | if warn: |
|
1616 | if warn: | |
1597 | warn( |
|
1617 | warn( | |
1598 | _(b"%s: ignoring invalid syntax '%s'\n") % (filepath, s) |
|
1618 | _(b"%s: ignoring invalid syntax '%s'\n") % (filepath, s) | |
1599 | ) |
|
1619 | ) | |
1600 | continue |
|
1620 | continue | |
1601 |
|
1621 | |||
1602 | linesyntax = syntax |
|
1622 | linesyntax = syntax | |
1603 | for s, rels in pycompat.iteritems(syntaxes): |
|
1623 | for s, rels in pycompat.iteritems(syntaxes): | |
1604 | if line.startswith(rels): |
|
1624 | if line.startswith(rels): | |
1605 | linesyntax = rels |
|
1625 | linesyntax = rels | |
1606 | line = line[len(rels) :] |
|
1626 | line = line[len(rels) :] | |
1607 | break |
|
1627 | break | |
1608 | elif line.startswith(s + b':'): |
|
1628 | elif line.startswith(s + b':'): | |
1609 | linesyntax = rels |
|
1629 | linesyntax = rels | |
1610 | line = line[len(s) + 1 :] |
|
1630 | line = line[len(s) + 1 :] | |
1611 | break |
|
1631 | break | |
1612 | if sourceinfo: |
|
1632 | if sourceinfo: | |
1613 | patterns.append((linesyntax + line, lineno, line)) |
|
1633 | patterns.append((linesyntax + line, lineno, line)) | |
1614 | else: |
|
1634 | else: | |
1615 | patterns.append(linesyntax + line) |
|
1635 | patterns.append(linesyntax + line) | |
1616 | fp.close() |
|
1636 | fp.close() | |
1617 | return patterns |
|
1637 | return patterns |
General Comments 0
You need to be logged in to leave comments.
Login now