Show More
@@ -1,961 +1,947 b'' | |||||
1 | # manifest.py - manifest revision class for mercurial |
|
1 | # manifest.py - manifest revision class for mercurial | |
2 | # |
|
2 | # | |
3 | # Copyright 2005-2007 Matt Mackall <mpm@selenic.com> |
|
3 | # Copyright 2005-2007 Matt Mackall <mpm@selenic.com> | |
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 i18n import _ |
|
8 | from i18n import _ | |
9 | import mdiff, parsers, error, revlog, util |
|
9 | import mdiff, parsers, error, revlog, util | |
10 | import array, struct |
|
10 | import array, struct | |
11 | import os |
|
11 | import os | |
12 |
|
12 | |||
13 | propertycache = util.propertycache |
|
13 | propertycache = util.propertycache | |
14 |
|
14 | |||
15 | def _parsev1(data): |
|
15 | def _parsev1(data): | |
16 | # This method does a little bit of excessive-looking |
|
16 | # This method does a little bit of excessive-looking | |
17 | # precondition checking. This is so that the behavior of this |
|
17 | # precondition checking. This is so that the behavior of this | |
18 | # class exactly matches its C counterpart to try and help |
|
18 | # class exactly matches its C counterpart to try and help | |
19 | # prevent surprise breakage for anyone that develops against |
|
19 | # prevent surprise breakage for anyone that develops against | |
20 | # the pure version. |
|
20 | # the pure version. | |
21 | if data and data[-1] != '\n': |
|
21 | if data and data[-1] != '\n': | |
22 | raise ValueError('Manifest did not end in a newline.') |
|
22 | raise ValueError('Manifest did not end in a newline.') | |
23 | prev = None |
|
23 | prev = None | |
24 | for l in data.splitlines(): |
|
24 | for l in data.splitlines(): | |
25 | if prev is not None and prev > l: |
|
25 | if prev is not None and prev > l: | |
26 | raise ValueError('Manifest lines not in sorted order.') |
|
26 | raise ValueError('Manifest lines not in sorted order.') | |
27 | prev = l |
|
27 | prev = l | |
28 | f, n = l.split('\0') |
|
28 | f, n = l.split('\0') | |
29 | if len(n) > 40: |
|
29 | if len(n) > 40: | |
30 | yield f, revlog.bin(n[:40]), n[40:] |
|
30 | yield f, revlog.bin(n[:40]), n[40:] | |
31 | else: |
|
31 | else: | |
32 | yield f, revlog.bin(n), '' |
|
32 | yield f, revlog.bin(n), '' | |
33 |
|
33 | |||
34 | def _parsev2(data): |
|
34 | def _parsev2(data): | |
35 | metadataend = data.find('\n') |
|
35 | metadataend = data.find('\n') | |
36 | # Just ignore metadata for now |
|
36 | # Just ignore metadata for now | |
37 | pos = metadataend + 1 |
|
37 | pos = metadataend + 1 | |
38 | prevf = '' |
|
38 | prevf = '' | |
39 | while pos < len(data): |
|
39 | while pos < len(data): | |
40 | end = data.find('\n', pos + 1) # +1 to skip stem length byte |
|
40 | end = data.find('\n', pos + 1) # +1 to skip stem length byte | |
41 | if end == -1: |
|
41 | if end == -1: | |
42 | raise ValueError('Manifest ended with incomplete file entry.') |
|
42 | raise ValueError('Manifest ended with incomplete file entry.') | |
43 | stemlen = ord(data[pos]) |
|
43 | stemlen = ord(data[pos]) | |
44 | items = data[pos + 1:end].split('\0') |
|
44 | items = data[pos + 1:end].split('\0') | |
45 | f = prevf[:stemlen] + items[0] |
|
45 | f = prevf[:stemlen] + items[0] | |
46 | if prevf > f: |
|
46 | if prevf > f: | |
47 | raise ValueError('Manifest entries not in sorted order.') |
|
47 | raise ValueError('Manifest entries not in sorted order.') | |
48 | fl = items[1] |
|
48 | fl = items[1] | |
49 | # Just ignore metadata (items[2:] for now) |
|
49 | # Just ignore metadata (items[2:] for now) | |
50 | n = data[end + 1:end + 21] |
|
50 | n = data[end + 1:end + 21] | |
51 | yield f, n, fl |
|
51 | yield f, n, fl | |
52 | pos = end + 22 |
|
52 | pos = end + 22 | |
53 | prevf = f |
|
53 | prevf = f | |
54 |
|
54 | |||
55 | def _parse(data): |
|
55 | def _parse(data): | |
56 | """Generates (path, node, flags) tuples from a manifest text""" |
|
56 | """Generates (path, node, flags) tuples from a manifest text""" | |
57 | if data.startswith('\0'): |
|
57 | if data.startswith('\0'): | |
58 | return iter(_parsev2(data)) |
|
58 | return iter(_parsev2(data)) | |
59 | else: |
|
59 | else: | |
60 | return iter(_parsev1(data)) |
|
60 | return iter(_parsev1(data)) | |
61 |
|
61 | |||
62 | def _text(it, usemanifestv2): |
|
62 | def _text(it, usemanifestv2): | |
63 | """Given an iterator over (path, node, flags) tuples, returns a manifest |
|
63 | """Given an iterator over (path, node, flags) tuples, returns a manifest | |
64 | text""" |
|
64 | text""" | |
65 | if usemanifestv2: |
|
65 | if usemanifestv2: | |
66 | return _textv2(it) |
|
66 | return _textv2(it) | |
67 | else: |
|
67 | else: | |
68 | return _textv1(it) |
|
68 | return _textv1(it) | |
69 |
|
69 | |||
70 | def _textv1(it): |
|
70 | def _textv1(it): | |
71 | files = [] |
|
71 | files = [] | |
72 | lines = [] |
|
72 | lines = [] | |
73 | _hex = revlog.hex |
|
73 | _hex = revlog.hex | |
74 | for f, n, fl in it: |
|
74 | for f, n, fl in it: | |
75 | files.append(f) |
|
75 | files.append(f) | |
76 | # if this is changed to support newlines in filenames, |
|
76 | # if this is changed to support newlines in filenames, | |
77 | # be sure to check the templates/ dir again (especially *-raw.tmpl) |
|
77 | # be sure to check the templates/ dir again (especially *-raw.tmpl) | |
78 | lines.append("%s\0%s%s\n" % (f, _hex(n), fl)) |
|
78 | lines.append("%s\0%s%s\n" % (f, _hex(n), fl)) | |
79 |
|
79 | |||
80 | _checkforbidden(files) |
|
80 | _checkforbidden(files) | |
81 | return ''.join(lines) |
|
81 | return ''.join(lines) | |
82 |
|
82 | |||
83 | def _textv2(it): |
|
83 | def _textv2(it): | |
84 | files = [] |
|
84 | files = [] | |
85 | lines = ['\0\n'] |
|
85 | lines = ['\0\n'] | |
86 | prevf = '' |
|
86 | prevf = '' | |
87 | for f, n, fl in it: |
|
87 | for f, n, fl in it: | |
88 | files.append(f) |
|
88 | files.append(f) | |
89 | stem = os.path.commonprefix([prevf, f]) |
|
89 | stem = os.path.commonprefix([prevf, f]) | |
90 | stemlen = min(len(stem), 255) |
|
90 | stemlen = min(len(stem), 255) | |
91 | lines.append("%c%s\0%s\n%s\n" % (stemlen, f[stemlen:], fl, n)) |
|
91 | lines.append("%c%s\0%s\n%s\n" % (stemlen, f[stemlen:], fl, n)) | |
92 | prevf = f |
|
92 | prevf = f | |
93 | _checkforbidden(files) |
|
93 | _checkforbidden(files) | |
94 | return ''.join(lines) |
|
94 | return ''.join(lines) | |
95 |
|
95 | |||
96 | class _lazymanifest(dict): |
|
96 | class _lazymanifest(dict): | |
97 | """This is the pure implementation of lazymanifest. |
|
97 | """This is the pure implementation of lazymanifest. | |
98 |
|
98 | |||
99 | It has not been optimized *at all* and is not lazy. |
|
99 | It has not been optimized *at all* and is not lazy. | |
100 | """ |
|
100 | """ | |
101 |
|
101 | |||
102 | def __init__(self, data): |
|
102 | def __init__(self, data): | |
103 | dict.__init__(self) |
|
103 | dict.__init__(self) | |
104 | for f, n, fl in _parse(data): |
|
104 | for f, n, fl in _parse(data): | |
105 | self[f] = n, fl |
|
105 | self[f] = n, fl | |
106 |
|
106 | |||
107 | def __setitem__(self, k, v): |
|
107 | def __setitem__(self, k, v): | |
108 | node, flag = v |
|
108 | node, flag = v | |
109 | assert node is not None |
|
109 | assert node is not None | |
110 | if len(node) > 21: |
|
110 | if len(node) > 21: | |
111 | node = node[:21] # match c implementation behavior |
|
111 | node = node[:21] # match c implementation behavior | |
112 | dict.__setitem__(self, k, (node, flag)) |
|
112 | dict.__setitem__(self, k, (node, flag)) | |
113 |
|
113 | |||
114 | def __iter__(self): |
|
114 | def __iter__(self): | |
115 | return iter(sorted(dict.keys(self))) |
|
115 | return iter(sorted(dict.keys(self))) | |
116 |
|
116 | |||
117 | def iterkeys(self): |
|
117 | def iterkeys(self): | |
118 | return iter(sorted(dict.keys(self))) |
|
118 | return iter(sorted(dict.keys(self))) | |
119 |
|
119 | |||
120 | def iterentries(self): |
|
120 | def iterentries(self): | |
121 | return ((f, e[0], e[1]) for f, e in sorted(self.iteritems())) |
|
121 | return ((f, e[0], e[1]) for f, e in sorted(self.iteritems())) | |
122 |
|
122 | |||
123 | def copy(self): |
|
123 | def copy(self): | |
124 | c = _lazymanifest('') |
|
124 | c = _lazymanifest('') | |
125 | c.update(self) |
|
125 | c.update(self) | |
126 | return c |
|
126 | return c | |
127 |
|
127 | |||
128 | def diff(self, m2, clean=False): |
|
128 | def diff(self, m2, clean=False): | |
129 | '''Finds changes between the current manifest and m2.''' |
|
129 | '''Finds changes between the current manifest and m2.''' | |
130 | diff = {} |
|
130 | diff = {} | |
131 |
|
131 | |||
132 | for fn, e1 in self.iteritems(): |
|
132 | for fn, e1 in self.iteritems(): | |
133 | if fn not in m2: |
|
133 | if fn not in m2: | |
134 | diff[fn] = e1, (None, '') |
|
134 | diff[fn] = e1, (None, '') | |
135 | else: |
|
135 | else: | |
136 | e2 = m2[fn] |
|
136 | e2 = m2[fn] | |
137 | if e1 != e2: |
|
137 | if e1 != e2: | |
138 | diff[fn] = e1, e2 |
|
138 | diff[fn] = e1, e2 | |
139 | elif clean: |
|
139 | elif clean: | |
140 | diff[fn] = None |
|
140 | diff[fn] = None | |
141 |
|
141 | |||
142 | for fn, e2 in m2.iteritems(): |
|
142 | for fn, e2 in m2.iteritems(): | |
143 | if fn not in self: |
|
143 | if fn not in self: | |
144 | diff[fn] = (None, ''), e2 |
|
144 | diff[fn] = (None, ''), e2 | |
145 |
|
145 | |||
146 | return diff |
|
146 | return diff | |
147 |
|
147 | |||
148 | def filtercopy(self, filterfn): |
|
148 | def filtercopy(self, filterfn): | |
149 | c = _lazymanifest('') |
|
149 | c = _lazymanifest('') | |
150 | for f, n, fl in self.iterentries(): |
|
150 | for f, n, fl in self.iterentries(): | |
151 | if filterfn(f): |
|
151 | if filterfn(f): | |
152 | c[f] = n, fl |
|
152 | c[f] = n, fl | |
153 | return c |
|
153 | return c | |
154 |
|
154 | |||
155 | def text(self): |
|
155 | def text(self): | |
156 | """Get the full data of this manifest as a bytestring.""" |
|
156 | """Get the full data of this manifest as a bytestring.""" | |
157 | return _textv1(self.iterentries()) |
|
157 | return _textv1(self.iterentries()) | |
158 |
|
158 | |||
159 | try: |
|
159 | try: | |
160 | _lazymanifest = parsers.lazymanifest |
|
160 | _lazymanifest = parsers.lazymanifest | |
161 | except AttributeError: |
|
161 | except AttributeError: | |
162 | pass |
|
162 | pass | |
163 |
|
163 | |||
164 | class manifestdict(object): |
|
164 | class manifestdict(object): | |
165 | def __init__(self, data=''): |
|
165 | def __init__(self, data=''): | |
166 | if data.startswith('\0'): |
|
166 | if data.startswith('\0'): | |
167 | #_lazymanifest can not parse v2 |
|
167 | #_lazymanifest can not parse v2 | |
168 | self._lm = _lazymanifest('') |
|
168 | self._lm = _lazymanifest('') | |
169 | for f, n, fl in _parsev2(data): |
|
169 | for f, n, fl in _parsev2(data): | |
170 | self._lm[f] = n, fl |
|
170 | self._lm[f] = n, fl | |
171 | else: |
|
171 | else: | |
172 | self._lm = _lazymanifest(data) |
|
172 | self._lm = _lazymanifest(data) | |
173 |
|
173 | |||
174 | def __getitem__(self, key): |
|
174 | def __getitem__(self, key): | |
175 | return self._lm[key][0] |
|
175 | return self._lm[key][0] | |
176 |
|
176 | |||
177 | def find(self, key): |
|
177 | def find(self, key): | |
178 | return self._lm[key] |
|
178 | return self._lm[key] | |
179 |
|
179 | |||
180 | def __len__(self): |
|
180 | def __len__(self): | |
181 | return len(self._lm) |
|
181 | return len(self._lm) | |
182 |
|
182 | |||
183 | def __setitem__(self, key, node): |
|
183 | def __setitem__(self, key, node): | |
184 | self._lm[key] = node, self.flags(key, '') |
|
184 | self._lm[key] = node, self.flags(key, '') | |
185 |
|
185 | |||
186 | def __contains__(self, key): |
|
186 | def __contains__(self, key): | |
187 | return key in self._lm |
|
187 | return key in self._lm | |
188 |
|
188 | |||
189 | def __delitem__(self, key): |
|
189 | def __delitem__(self, key): | |
190 | del self._lm[key] |
|
190 | del self._lm[key] | |
191 |
|
191 | |||
192 | def __iter__(self): |
|
192 | def __iter__(self): | |
193 | return self._lm.__iter__() |
|
193 | return self._lm.__iter__() | |
194 |
|
194 | |||
195 | def iterkeys(self): |
|
195 | def iterkeys(self): | |
196 | return self._lm.iterkeys() |
|
196 | return self._lm.iterkeys() | |
197 |
|
197 | |||
198 | def keys(self): |
|
198 | def keys(self): | |
199 | return list(self.iterkeys()) |
|
199 | return list(self.iterkeys()) | |
200 |
|
200 | |||
201 | def filesnotin(self, m2): |
|
201 | def filesnotin(self, m2): | |
202 | '''Set of files in this manifest that are not in the other''' |
|
202 | '''Set of files in this manifest that are not in the other''' | |
203 | files = set(self) |
|
203 | files = set(self) | |
204 | files.difference_update(m2) |
|
204 | files.difference_update(m2) | |
205 | return files |
|
205 | return files | |
206 |
|
206 | |||
207 | @propertycache |
|
207 | @propertycache | |
208 | def _dirs(self): |
|
208 | def _dirs(self): | |
209 | return util.dirs(self) |
|
209 | return util.dirs(self) | |
210 |
|
210 | |||
211 | def dirs(self): |
|
211 | def dirs(self): | |
212 | return self._dirs |
|
212 | return self._dirs | |
213 |
|
213 | |||
214 | def hasdir(self, dir): |
|
214 | def hasdir(self, dir): | |
215 | return dir in self._dirs |
|
215 | return dir in self._dirs | |
216 |
|
216 | |||
217 | def _filesfastpath(self, match): |
|
217 | def _filesfastpath(self, match): | |
218 | '''Checks whether we can correctly and quickly iterate over matcher |
|
218 | '''Checks whether we can correctly and quickly iterate over matcher | |
219 | files instead of over manifest files.''' |
|
219 | files instead of over manifest files.''' | |
220 | files = match.files() |
|
220 | files = match.files() | |
221 | return (len(files) < 100 and (match.isexact() or |
|
221 | return (len(files) < 100 and (match.isexact() or | |
222 | (not match.anypats() and all(fn in self for fn in files)))) |
|
222 | (not match.anypats() and all(fn in self for fn in files)))) | |
223 |
|
223 | |||
224 | def walk(self, match): |
|
224 | def walk(self, match): | |
225 | '''Generates matching file names. |
|
225 | '''Generates matching file names. | |
226 |
|
226 | |||
227 | Equivalent to manifest.matches(match).iterkeys(), but without creating |
|
227 | Equivalent to manifest.matches(match).iterkeys(), but without creating | |
228 | an entirely new manifest. |
|
228 | an entirely new manifest. | |
229 |
|
229 | |||
230 | It also reports nonexistent files by marking them bad with match.bad(). |
|
230 | It also reports nonexistent files by marking them bad with match.bad(). | |
231 | ''' |
|
231 | ''' | |
232 | if match.always(): |
|
232 | if match.always(): | |
233 | for f in iter(self): |
|
233 | for f in iter(self): | |
234 | yield f |
|
234 | yield f | |
235 | return |
|
235 | return | |
236 |
|
236 | |||
237 | fset = set(match.files()) |
|
237 | fset = set(match.files()) | |
238 |
|
238 | |||
239 | # avoid the entire walk if we're only looking for specific files |
|
239 | # avoid the entire walk if we're only looking for specific files | |
240 | if self._filesfastpath(match): |
|
240 | if self._filesfastpath(match): | |
241 | for fn in sorted(fset): |
|
241 | for fn in sorted(fset): | |
242 | yield fn |
|
242 | yield fn | |
243 | return |
|
243 | return | |
244 |
|
244 | |||
245 | for fn in self: |
|
245 | for fn in self: | |
246 | if fn in fset: |
|
246 | if fn in fset: | |
247 | # specified pattern is the exact name |
|
247 | # specified pattern is the exact name | |
248 | fset.remove(fn) |
|
248 | fset.remove(fn) | |
249 | if match(fn): |
|
249 | if match(fn): | |
250 | yield fn |
|
250 | yield fn | |
251 |
|
251 | |||
252 | # for dirstate.walk, files=['.'] means "walk the whole tree". |
|
252 | # for dirstate.walk, files=['.'] means "walk the whole tree". | |
253 | # follow that here, too |
|
253 | # follow that here, too | |
254 | fset.discard('.') |
|
254 | fset.discard('.') | |
255 |
|
255 | |||
256 | for fn in sorted(fset): |
|
256 | for fn in sorted(fset): | |
257 | if not self.hasdir(fn): |
|
257 | if not self.hasdir(fn): | |
258 | match.bad(fn, None) |
|
258 | match.bad(fn, None) | |
259 |
|
259 | |||
260 | def matches(self, match): |
|
260 | def matches(self, match): | |
261 | '''generate a new manifest filtered by the match argument''' |
|
261 | '''generate a new manifest filtered by the match argument''' | |
262 | if match.always(): |
|
262 | if match.always(): | |
263 | return self.copy() |
|
263 | return self.copy() | |
264 |
|
264 | |||
265 | if self._filesfastpath(match): |
|
265 | if self._filesfastpath(match): | |
266 | m = manifestdict() |
|
266 | m = manifestdict() | |
267 | lm = self._lm |
|
267 | lm = self._lm | |
268 | for fn in match.files(): |
|
268 | for fn in match.files(): | |
269 | if fn in lm: |
|
269 | if fn in lm: | |
270 | m._lm[fn] = lm[fn] |
|
270 | m._lm[fn] = lm[fn] | |
271 | return m |
|
271 | return m | |
272 |
|
272 | |||
273 | m = manifestdict() |
|
273 | m = manifestdict() | |
274 | m._lm = self._lm.filtercopy(match) |
|
274 | m._lm = self._lm.filtercopy(match) | |
275 | return m |
|
275 | return m | |
276 |
|
276 | |||
277 | def diff(self, m2, clean=False): |
|
277 | def diff(self, m2, clean=False): | |
278 | '''Finds changes between the current manifest and m2. |
|
278 | '''Finds changes between the current manifest and m2. | |
279 |
|
279 | |||
280 | Args: |
|
280 | Args: | |
281 | m2: the manifest to which this manifest should be compared. |
|
281 | m2: the manifest to which this manifest should be compared. | |
282 | clean: if true, include files unchanged between these manifests |
|
282 | clean: if true, include files unchanged between these manifests | |
283 | with a None value in the returned dictionary. |
|
283 | with a None value in the returned dictionary. | |
284 |
|
284 | |||
285 | The result is returned as a dict with filename as key and |
|
285 | The result is returned as a dict with filename as key and | |
286 | values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the |
|
286 | values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the | |
287 | nodeid in the current/other manifest and fl1/fl2 is the flag |
|
287 | nodeid in the current/other manifest and fl1/fl2 is the flag | |
288 | in the current/other manifest. Where the file does not exist, |
|
288 | in the current/other manifest. Where the file does not exist, | |
289 | the nodeid will be None and the flags will be the empty |
|
289 | the nodeid will be None and the flags will be the empty | |
290 | string. |
|
290 | string. | |
291 | ''' |
|
291 | ''' | |
292 | return self._lm.diff(m2._lm, clean) |
|
292 | return self._lm.diff(m2._lm, clean) | |
293 |
|
293 | |||
294 | def setflag(self, key, flag): |
|
294 | def setflag(self, key, flag): | |
295 | self._lm[key] = self[key], flag |
|
295 | self._lm[key] = self[key], flag | |
296 |
|
296 | |||
297 | def get(self, key, default=None): |
|
297 | def get(self, key, default=None): | |
298 | try: |
|
298 | try: | |
299 | return self._lm[key][0] |
|
299 | return self._lm[key][0] | |
300 | except KeyError: |
|
300 | except KeyError: | |
301 | return default |
|
301 | return default | |
302 |
|
302 | |||
303 | def flags(self, key, default=''): |
|
303 | def flags(self, key, default=''): | |
304 | try: |
|
304 | try: | |
305 | return self._lm[key][1] |
|
305 | return self._lm[key][1] | |
306 | except KeyError: |
|
306 | except KeyError: | |
307 | return default |
|
307 | return default | |
308 |
|
308 | |||
309 | def copy(self): |
|
309 | def copy(self): | |
310 | c = manifestdict() |
|
310 | c = manifestdict() | |
311 | c._lm = self._lm.copy() |
|
311 | c._lm = self._lm.copy() | |
312 | return c |
|
312 | return c | |
313 |
|
313 | |||
314 | def iteritems(self): |
|
314 | def iteritems(self): | |
315 | return (x[:2] for x in self._lm.iterentries()) |
|
315 | return (x[:2] for x in self._lm.iterentries()) | |
316 |
|
316 | |||
317 | def text(self, usemanifestv2=False): |
|
317 | def text(self, usemanifestv2=False): | |
318 | if usemanifestv2: |
|
318 | if usemanifestv2: | |
319 | return _textv2(self._lm.iterentries()) |
|
319 | return _textv2(self._lm.iterentries()) | |
320 | else: |
|
320 | else: | |
321 | # use (probably) native version for v1 |
|
321 | # use (probably) native version for v1 | |
322 | return self._lm.text() |
|
322 | return self._lm.text() | |
323 |
|
323 | |||
324 | def fastdelta(self, base, changes): |
|
324 | def fastdelta(self, base, changes): | |
325 | """Given a base manifest text as an array.array and a list of changes |
|
325 | """Given a base manifest text as an array.array and a list of changes | |
326 | relative to that text, compute a delta that can be used by revlog. |
|
326 | relative to that text, compute a delta that can be used by revlog. | |
327 | """ |
|
327 | """ | |
328 | delta = [] |
|
328 | delta = [] | |
329 | dstart = None |
|
329 | dstart = None | |
330 | dend = None |
|
330 | dend = None | |
331 | dline = [""] |
|
331 | dline = [""] | |
332 | start = 0 |
|
332 | start = 0 | |
333 | # zero copy representation of base as a buffer |
|
333 | # zero copy representation of base as a buffer | |
334 | addbuf = util.buffer(base) |
|
334 | addbuf = util.buffer(base) | |
335 |
|
335 | |||
336 | # start with a readonly loop that finds the offset of |
|
336 | # start with a readonly loop that finds the offset of | |
337 | # each line and creates the deltas |
|
337 | # each line and creates the deltas | |
338 | for f, todelete in changes: |
|
338 | for f, todelete in changes: | |
339 | # bs will either be the index of the item or the insert point |
|
339 | # bs will either be the index of the item or the insert point | |
340 | start, end = _msearch(addbuf, f, start) |
|
340 | start, end = _msearch(addbuf, f, start) | |
341 | if not todelete: |
|
341 | if not todelete: | |
342 | h, fl = self._lm[f] |
|
342 | h, fl = self._lm[f] | |
343 | l = "%s\0%s%s\n" % (f, revlog.hex(h), fl) |
|
343 | l = "%s\0%s%s\n" % (f, revlog.hex(h), fl) | |
344 | else: |
|
344 | else: | |
345 | if start == end: |
|
345 | if start == end: | |
346 | # item we want to delete was not found, error out |
|
346 | # item we want to delete was not found, error out | |
347 | raise AssertionError( |
|
347 | raise AssertionError( | |
348 | _("failed to remove %s from manifest") % f) |
|
348 | _("failed to remove %s from manifest") % f) | |
349 | l = "" |
|
349 | l = "" | |
350 | if dstart is not None and dstart <= start and dend >= start: |
|
350 | if dstart is not None and dstart <= start and dend >= start: | |
351 | if dend < end: |
|
351 | if dend < end: | |
352 | dend = end |
|
352 | dend = end | |
353 | if l: |
|
353 | if l: | |
354 | dline.append(l) |
|
354 | dline.append(l) | |
355 | else: |
|
355 | else: | |
356 | if dstart is not None: |
|
356 | if dstart is not None: | |
357 | delta.append([dstart, dend, "".join(dline)]) |
|
357 | delta.append([dstart, dend, "".join(dline)]) | |
358 | dstart = start |
|
358 | dstart = start | |
359 | dend = end |
|
359 | dend = end | |
360 | dline = [l] |
|
360 | dline = [l] | |
361 |
|
361 | |||
362 | if dstart is not None: |
|
362 | if dstart is not None: | |
363 | delta.append([dstart, dend, "".join(dline)]) |
|
363 | delta.append([dstart, dend, "".join(dline)]) | |
364 | # apply the delta to the base, and get a delta for addrevision |
|
364 | # apply the delta to the base, and get a delta for addrevision | |
365 | deltatext, arraytext = _addlistdelta(base, delta) |
|
365 | deltatext, arraytext = _addlistdelta(base, delta) | |
366 | return arraytext, deltatext |
|
366 | return arraytext, deltatext | |
367 |
|
367 | |||
368 | def _msearch(m, s, lo=0, hi=None): |
|
368 | def _msearch(m, s, lo=0, hi=None): | |
369 | '''return a tuple (start, end) that says where to find s within m. |
|
369 | '''return a tuple (start, end) that says where to find s within m. | |
370 |
|
370 | |||
371 | If the string is found m[start:end] are the line containing |
|
371 | If the string is found m[start:end] are the line containing | |
372 | that string. If start == end the string was not found and |
|
372 | that string. If start == end the string was not found and | |
373 | they indicate the proper sorted insertion point. |
|
373 | they indicate the proper sorted insertion point. | |
374 |
|
374 | |||
375 | m should be a buffer or a string |
|
375 | m should be a buffer or a string | |
376 | s is a string''' |
|
376 | s is a string''' | |
377 | def advance(i, c): |
|
377 | def advance(i, c): | |
378 | while i < lenm and m[i] != c: |
|
378 | while i < lenm and m[i] != c: | |
379 | i += 1 |
|
379 | i += 1 | |
380 | return i |
|
380 | return i | |
381 | if not s: |
|
381 | if not s: | |
382 | return (lo, lo) |
|
382 | return (lo, lo) | |
383 | lenm = len(m) |
|
383 | lenm = len(m) | |
384 | if not hi: |
|
384 | if not hi: | |
385 | hi = lenm |
|
385 | hi = lenm | |
386 | while lo < hi: |
|
386 | while lo < hi: | |
387 | mid = (lo + hi) // 2 |
|
387 | mid = (lo + hi) // 2 | |
388 | start = mid |
|
388 | start = mid | |
389 | while start > 0 and m[start - 1] != '\n': |
|
389 | while start > 0 and m[start - 1] != '\n': | |
390 | start -= 1 |
|
390 | start -= 1 | |
391 | end = advance(start, '\0') |
|
391 | end = advance(start, '\0') | |
392 | if m[start:end] < s: |
|
392 | if m[start:end] < s: | |
393 | # we know that after the null there are 40 bytes of sha1 |
|
393 | # we know that after the null there are 40 bytes of sha1 | |
394 | # this translates to the bisect lo = mid + 1 |
|
394 | # this translates to the bisect lo = mid + 1 | |
395 | lo = advance(end + 40, '\n') + 1 |
|
395 | lo = advance(end + 40, '\n') + 1 | |
396 | else: |
|
396 | else: | |
397 | # this translates to the bisect hi = mid |
|
397 | # this translates to the bisect hi = mid | |
398 | hi = start |
|
398 | hi = start | |
399 | end = advance(lo, '\0') |
|
399 | end = advance(lo, '\0') | |
400 | found = m[lo:end] |
|
400 | found = m[lo:end] | |
401 | if s == found: |
|
401 | if s == found: | |
402 | # we know that after the null there are 40 bytes of sha1 |
|
402 | # we know that after the null there are 40 bytes of sha1 | |
403 | end = advance(end + 40, '\n') |
|
403 | end = advance(end + 40, '\n') | |
404 | return (lo, end + 1) |
|
404 | return (lo, end + 1) | |
405 | else: |
|
405 | else: | |
406 | return (lo, lo) |
|
406 | return (lo, lo) | |
407 |
|
407 | |||
408 | def _checkforbidden(l): |
|
408 | def _checkforbidden(l): | |
409 | """Check filenames for illegal characters.""" |
|
409 | """Check filenames for illegal characters.""" | |
410 | for f in l: |
|
410 | for f in l: | |
411 | if '\n' in f or '\r' in f: |
|
411 | if '\n' in f or '\r' in f: | |
412 | raise error.RevlogError( |
|
412 | raise error.RevlogError( | |
413 | _("'\\n' and '\\r' disallowed in filenames: %r") % f) |
|
413 | _("'\\n' and '\\r' disallowed in filenames: %r") % f) | |
414 |
|
414 | |||
415 |
|
415 | |||
416 | # apply the changes collected during the bisect loop to our addlist |
|
416 | # apply the changes collected during the bisect loop to our addlist | |
417 | # return a delta suitable for addrevision |
|
417 | # return a delta suitable for addrevision | |
418 | def _addlistdelta(addlist, x): |
|
418 | def _addlistdelta(addlist, x): | |
419 | # for large addlist arrays, building a new array is cheaper |
|
419 | # for large addlist arrays, building a new array is cheaper | |
420 | # than repeatedly modifying the existing one |
|
420 | # than repeatedly modifying the existing one | |
421 | currentposition = 0 |
|
421 | currentposition = 0 | |
422 | newaddlist = array.array('c') |
|
422 | newaddlist = array.array('c') | |
423 |
|
423 | |||
424 | for start, end, content in x: |
|
424 | for start, end, content in x: | |
425 | newaddlist += addlist[currentposition:start] |
|
425 | newaddlist += addlist[currentposition:start] | |
426 | if content: |
|
426 | if content: | |
427 | newaddlist += array.array('c', content) |
|
427 | newaddlist += array.array('c', content) | |
428 |
|
428 | |||
429 | currentposition = end |
|
429 | currentposition = end | |
430 |
|
430 | |||
431 | newaddlist += addlist[currentposition:] |
|
431 | newaddlist += addlist[currentposition:] | |
432 |
|
432 | |||
433 | deltatext = "".join(struct.pack(">lll", start, end, len(content)) |
|
433 | deltatext = "".join(struct.pack(">lll", start, end, len(content)) | |
434 | + content for start, end, content in x) |
|
434 | + content for start, end, content in x) | |
435 | return deltatext, newaddlist |
|
435 | return deltatext, newaddlist | |
436 |
|
436 | |||
437 | def _splittopdir(f): |
|
437 | def _splittopdir(f): | |
438 | if '/' in f: |
|
438 | if '/' in f: | |
439 | dir, subpath = f.split('/', 1) |
|
439 | dir, subpath = f.split('/', 1) | |
440 | return dir + '/', subpath |
|
440 | return dir + '/', subpath | |
441 | else: |
|
441 | else: | |
442 | return '', f |
|
442 | return '', f | |
443 |
|
443 | |||
444 | class treemanifest(object): |
|
444 | class treemanifest(object): | |
445 | def __init__(self, dir='', text=''): |
|
445 | def __init__(self, dir='', text=''): | |
446 | self._dir = dir |
|
446 | self._dir = dir | |
447 | self._node = revlog.nullid |
|
447 | self._node = revlog.nullid | |
448 | self._dirs = {} |
|
448 | self._dirs = {} | |
449 | # Using _lazymanifest here is a little slower than plain old dicts |
|
449 | # Using _lazymanifest here is a little slower than plain old dicts | |
450 | self._files = {} |
|
450 | self._files = {} | |
451 | self._flags = {} |
|
451 | self._flags = {} | |
452 | def readsubtree(subdir, subm): |
|
452 | def readsubtree(subdir, subm): | |
453 | raise AssertionError('treemanifest constructor only accepts ' |
|
453 | raise AssertionError('treemanifest constructor only accepts ' | |
454 | 'flat manifests') |
|
454 | 'flat manifests') | |
455 | self.parse(text, readsubtree) |
|
455 | self.parse(text, readsubtree) | |
456 |
|
456 | |||
457 | def _subpath(self, path): |
|
457 | def _subpath(self, path): | |
458 | return self._dir + path |
|
458 | return self._dir + path | |
459 |
|
459 | |||
460 | def __len__(self): |
|
460 | def __len__(self): | |
461 | size = len(self._files) |
|
461 | size = len(self._files) | |
462 | for m in self._dirs.values(): |
|
462 | for m in self._dirs.values(): | |
463 | size += m.__len__() |
|
463 | size += m.__len__() | |
464 | return size |
|
464 | return size | |
465 |
|
465 | |||
466 | def _isempty(self): |
|
466 | def _isempty(self): | |
467 | return (not self._files and (not self._dirs or |
|
467 | return (not self._files and (not self._dirs or | |
468 | all(m._isempty() for m in self._dirs.values()))) |
|
468 | all(m._isempty() for m in self._dirs.values()))) | |
469 |
|
469 | |||
470 | def __str__(self): |
|
470 | def __str__(self): | |
471 | return ('<treemanifest dir=%s, node=%s>' % |
|
471 | return ('<treemanifest dir=%s, node=%s>' % | |
472 | (self._dir, revlog.hex(self._node))) |
|
472 | (self._dir, revlog.hex(self._node))) | |
473 |
|
473 | |||
474 | def dir(self): |
|
474 | def dir(self): | |
475 | '''The directory that this tree manifest represents, including a |
|
475 | '''The directory that this tree manifest represents, including a | |
476 | trailing '/'. Empty string for the repo root directory.''' |
|
476 | trailing '/'. Empty string for the repo root directory.''' | |
477 | return self._dir |
|
477 | return self._dir | |
478 |
|
478 | |||
479 | def node(self): |
|
479 | def node(self): | |
480 | '''This node of this instance. nullid for unsaved instances. Should |
|
480 | '''This node of this instance. nullid for unsaved instances. Should | |
481 | be updated when the instance is read or written from a revlog. |
|
481 | be updated when the instance is read or written from a revlog. | |
482 | ''' |
|
482 | ''' | |
483 | return self._node |
|
483 | return self._node | |
484 |
|
484 | |||
485 | def setnode(self, node): |
|
485 | def setnode(self, node): | |
486 | self._node = node |
|
486 | self._node = node | |
487 |
|
487 | |||
488 | def iteritems(self): |
|
488 | def iteritems(self): | |
489 | for p, n in sorted(self._dirs.items() + self._files.items()): |
|
489 | for p, n in sorted(self._dirs.items() + self._files.items()): | |
490 | if p in self._files: |
|
490 | if p in self._files: | |
491 | yield self._subpath(p), n |
|
491 | yield self._subpath(p), n | |
492 | else: |
|
492 | else: | |
493 | for f, sn in n.iteritems(): |
|
493 | for f, sn in n.iteritems(): | |
494 | yield f, sn |
|
494 | yield f, sn | |
495 |
|
495 | |||
496 | def iterkeys(self): |
|
496 | def iterkeys(self): | |
497 | for p in sorted(self._dirs.keys() + self._files.keys()): |
|
497 | for p in sorted(self._dirs.keys() + self._files.keys()): | |
498 | if p in self._files: |
|
498 | if p in self._files: | |
499 | yield self._subpath(p) |
|
499 | yield self._subpath(p) | |
500 | else: |
|
500 | else: | |
501 | for f in self._dirs[p].iterkeys(): |
|
501 | for f in self._dirs[p].iterkeys(): | |
502 | yield f |
|
502 | yield f | |
503 |
|
503 | |||
504 | def keys(self): |
|
504 | def keys(self): | |
505 | return list(self.iterkeys()) |
|
505 | return list(self.iterkeys()) | |
506 |
|
506 | |||
507 | def __iter__(self): |
|
507 | def __iter__(self): | |
508 | return self.iterkeys() |
|
508 | return self.iterkeys() | |
509 |
|
509 | |||
510 | def __contains__(self, f): |
|
510 | def __contains__(self, f): | |
511 | if f is None: |
|
511 | if f is None: | |
512 | return False |
|
512 | return False | |
513 | dir, subpath = _splittopdir(f) |
|
513 | dir, subpath = _splittopdir(f) | |
514 | if dir: |
|
514 | if dir: | |
515 | if dir not in self._dirs: |
|
515 | if dir not in self._dirs: | |
516 | return False |
|
516 | return False | |
517 | return self._dirs[dir].__contains__(subpath) |
|
517 | return self._dirs[dir].__contains__(subpath) | |
518 | else: |
|
518 | else: | |
519 | return f in self._files |
|
519 | return f in self._files | |
520 |
|
520 | |||
521 | def get(self, f, default=None): |
|
521 | def get(self, f, default=None): | |
522 | dir, subpath = _splittopdir(f) |
|
522 | dir, subpath = _splittopdir(f) | |
523 | if dir: |
|
523 | if dir: | |
524 | if dir not in self._dirs: |
|
524 | if dir not in self._dirs: | |
525 | return default |
|
525 | return default | |
526 | return self._dirs[dir].get(subpath, default) |
|
526 | return self._dirs[dir].get(subpath, default) | |
527 | else: |
|
527 | else: | |
528 | return self._files.get(f, default) |
|
528 | return self._files.get(f, default) | |
529 |
|
529 | |||
530 | def __getitem__(self, f): |
|
530 | def __getitem__(self, f): | |
531 | dir, subpath = _splittopdir(f) |
|
531 | dir, subpath = _splittopdir(f) | |
532 | if dir: |
|
532 | if dir: | |
533 | return self._dirs[dir].__getitem__(subpath) |
|
533 | return self._dirs[dir].__getitem__(subpath) | |
534 | else: |
|
534 | else: | |
535 | return self._files[f] |
|
535 | return self._files[f] | |
536 |
|
536 | |||
537 | def flags(self, f): |
|
537 | def flags(self, f): | |
538 | dir, subpath = _splittopdir(f) |
|
538 | dir, subpath = _splittopdir(f) | |
539 | if dir: |
|
539 | if dir: | |
540 | if dir not in self._dirs: |
|
540 | if dir not in self._dirs: | |
541 | return '' |
|
541 | return '' | |
542 | return self._dirs[dir].flags(subpath) |
|
542 | return self._dirs[dir].flags(subpath) | |
543 | else: |
|
543 | else: | |
544 | if f in self._dirs: |
|
544 | if f in self._dirs: | |
545 | return '' |
|
545 | return '' | |
546 | return self._flags.get(f, '') |
|
546 | return self._flags.get(f, '') | |
547 |
|
547 | |||
548 | def find(self, f): |
|
548 | def find(self, f): | |
549 | dir, subpath = _splittopdir(f) |
|
549 | dir, subpath = _splittopdir(f) | |
550 | if dir: |
|
550 | if dir: | |
551 | return self._dirs[dir].find(subpath) |
|
551 | return self._dirs[dir].find(subpath) | |
552 | else: |
|
552 | else: | |
553 | return self._files[f], self._flags.get(f, '') |
|
553 | return self._files[f], self._flags.get(f, '') | |
554 |
|
554 | |||
555 | def __delitem__(self, f): |
|
555 | def __delitem__(self, f): | |
556 | dir, subpath = _splittopdir(f) |
|
556 | dir, subpath = _splittopdir(f) | |
557 | if dir: |
|
557 | if dir: | |
558 | self._dirs[dir].__delitem__(subpath) |
|
558 | self._dirs[dir].__delitem__(subpath) | |
559 | # If the directory is now empty, remove it |
|
559 | # If the directory is now empty, remove it | |
560 | if self._dirs[dir]._isempty(): |
|
560 | if self._dirs[dir]._isempty(): | |
561 | del self._dirs[dir] |
|
561 | del self._dirs[dir] | |
562 | else: |
|
562 | else: | |
563 | del self._files[f] |
|
563 | del self._files[f] | |
564 | if f in self._flags: |
|
564 | if f in self._flags: | |
565 | del self._flags[f] |
|
565 | del self._flags[f] | |
566 |
|
566 | |||
567 | def __setitem__(self, f, n): |
|
567 | def __setitem__(self, f, n): | |
568 | assert n is not None |
|
568 | assert n is not None | |
569 | dir, subpath = _splittopdir(f) |
|
569 | dir, subpath = _splittopdir(f) | |
570 | if dir: |
|
570 | if dir: | |
571 | if dir not in self._dirs: |
|
571 | if dir not in self._dirs: | |
572 | self._dirs[dir] = treemanifest(self._subpath(dir)) |
|
572 | self._dirs[dir] = treemanifest(self._subpath(dir)) | |
573 | self._dirs[dir].__setitem__(subpath, n) |
|
573 | self._dirs[dir].__setitem__(subpath, n) | |
574 | else: |
|
574 | else: | |
575 | self._files[f] = n[:21] # to match manifestdict's behavior |
|
575 | self._files[f] = n[:21] # to match manifestdict's behavior | |
576 |
|
576 | |||
577 | def setflag(self, f, flags): |
|
577 | def setflag(self, f, flags): | |
578 | """Set the flags (symlink, executable) for path f.""" |
|
578 | """Set the flags (symlink, executable) for path f.""" | |
579 | assert 'd' not in flags |
|
579 | assert 'd' not in flags | |
580 | dir, subpath = _splittopdir(f) |
|
580 | dir, subpath = _splittopdir(f) | |
581 | if dir: |
|
581 | if dir: | |
582 | if dir not in self._dirs: |
|
582 | if dir not in self._dirs: | |
583 | self._dirs[dir] = treemanifest(self._subpath(dir)) |
|
583 | self._dirs[dir] = treemanifest(self._subpath(dir)) | |
584 | self._dirs[dir].setflag(subpath, flags) |
|
584 | self._dirs[dir].setflag(subpath, flags) | |
585 | else: |
|
585 | else: | |
586 | self._flags[f] = flags |
|
586 | self._flags[f] = flags | |
587 |
|
587 | |||
588 | def copy(self): |
|
588 | def copy(self): | |
589 | copy = treemanifest(self._dir) |
|
589 | copy = treemanifest(self._dir) | |
590 | copy._node = self._node |
|
590 | copy._node = self._node | |
591 | for d in self._dirs: |
|
591 | for d in self._dirs: | |
592 | copy._dirs[d] = self._dirs[d].copy() |
|
592 | copy._dirs[d] = self._dirs[d].copy() | |
593 | copy._files = dict.copy(self._files) |
|
593 | copy._files = dict.copy(self._files) | |
594 | copy._flags = dict.copy(self._flags) |
|
594 | copy._flags = dict.copy(self._flags) | |
595 | return copy |
|
595 | return copy | |
596 |
|
596 | |||
597 | def filesnotin(self, m2): |
|
597 | def filesnotin(self, m2): | |
598 | '''Set of files in this manifest that are not in the other''' |
|
598 | '''Set of files in this manifest that are not in the other''' | |
599 | files = set() |
|
599 | files = set() | |
600 | def _filesnotin(t1, t2): |
|
600 | def _filesnotin(t1, t2): | |
601 | for d, m1 in t1._dirs.iteritems(): |
|
601 | for d, m1 in t1._dirs.iteritems(): | |
602 | if d in t2._dirs: |
|
602 | if d in t2._dirs: | |
603 | m2 = t2._dirs[d] |
|
603 | m2 = t2._dirs[d] | |
604 | _filesnotin(m1, m2) |
|
604 | _filesnotin(m1, m2) | |
605 | else: |
|
605 | else: | |
606 | files.update(m1.iterkeys()) |
|
606 | files.update(m1.iterkeys()) | |
607 |
|
607 | |||
608 | for fn in t1._files.iterkeys(): |
|
608 | for fn in t1._files.iterkeys(): | |
609 | if fn not in t2._files: |
|
609 | if fn not in t2._files: | |
610 | files.add(t1._subpath(fn)) |
|
610 | files.add(t1._subpath(fn)) | |
611 |
|
611 | |||
612 | _filesnotin(self, m2) |
|
612 | _filesnotin(self, m2) | |
613 | return files |
|
613 | return files | |
614 |
|
614 | |||
615 | @propertycache |
|
615 | @propertycache | |
616 | def _alldirs(self): |
|
616 | def _alldirs(self): | |
617 | return util.dirs(self) |
|
617 | return util.dirs(self) | |
618 |
|
618 | |||
619 | def dirs(self): |
|
619 | def dirs(self): | |
620 | return self._alldirs |
|
620 | return self._alldirs | |
621 |
|
621 | |||
622 | def hasdir(self, dir): |
|
622 | def hasdir(self, dir): | |
623 | topdir, subdir = _splittopdir(dir) |
|
623 | topdir, subdir = _splittopdir(dir) | |
624 | if topdir: |
|
624 | if topdir: | |
625 | if topdir in self._dirs: |
|
625 | if topdir in self._dirs: | |
626 | return self._dirs[topdir].hasdir(subdir) |
|
626 | return self._dirs[topdir].hasdir(subdir) | |
627 | return False |
|
627 | return False | |
628 | return (dir + '/') in self._dirs |
|
628 | return (dir + '/') in self._dirs | |
629 |
|
629 | |||
630 | def walk(self, match): |
|
630 | def walk(self, match): | |
631 | '''Generates matching file names. |
|
631 | '''Generates matching file names. | |
632 |
|
632 | |||
633 | Equivalent to manifest.matches(match).iterkeys(), but without creating |
|
633 | Equivalent to manifest.matches(match).iterkeys(), but without creating | |
634 | an entirely new manifest. |
|
634 | an entirely new manifest. | |
635 |
|
635 | |||
636 | It also reports nonexistent files by marking them bad with match.bad(). |
|
636 | It also reports nonexistent files by marking them bad with match.bad(). | |
637 | ''' |
|
637 | ''' | |
638 | if match.always(): |
|
638 | if match.always(): | |
639 | for f in iter(self): |
|
639 | for f in iter(self): | |
640 | yield f |
|
640 | yield f | |
641 | return |
|
641 | return | |
642 |
|
642 | |||
643 | fset = set(match.files()) |
|
643 | fset = set(match.files()) | |
644 |
|
644 | |||
645 | for fn in self._walk(match): |
|
645 | for fn in self._walk(match): | |
646 | if fn in fset: |
|
646 | if fn in fset: | |
647 | # specified pattern is the exact name |
|
647 | # specified pattern is the exact name | |
648 | fset.remove(fn) |
|
648 | fset.remove(fn) | |
649 | yield fn |
|
649 | yield fn | |
650 |
|
650 | |||
651 | # for dirstate.walk, files=['.'] means "walk the whole tree". |
|
651 | # for dirstate.walk, files=['.'] means "walk the whole tree". | |
652 | # follow that here, too |
|
652 | # follow that here, too | |
653 | fset.discard('.') |
|
653 | fset.discard('.') | |
654 |
|
654 | |||
655 | for fn in sorted(fset): |
|
655 | for fn in sorted(fset): | |
656 | if not self.hasdir(fn): |
|
656 | if not self.hasdir(fn): | |
657 | match.bad(fn, None) |
|
657 | match.bad(fn, None) | |
658 |
|
658 | |||
659 |
def _walk(self, match |
|
659 | def _walk(self, match): | |
660 | '''Recursively generates matching file names for walk(). |
|
660 | '''Recursively generates matching file names for walk().''' | |
661 |
|
661 | if not match.visitdir(self._dir[:-1] or '.'): | ||
662 | Will visit all subdirectories if alldirs is True, otherwise it will |
|
|||
663 | only visit subdirectories for which match.visitdir is True.''' |
|
|||
664 |
|
||||
665 | if not alldirs: |
|
|||
666 | # substring to strip trailing slash |
|
|||
667 | visit = match.visitdir(self._dir[:-1] or '.') |
|
|||
668 | if not visit: |
|
|||
669 |
|
|
662 | return | |
670 | alldirs = (visit == 'all') |
|
|||
671 |
|
663 | |||
672 | # yield this dir's files and walk its submanifests |
|
664 | # yield this dir's files and walk its submanifests | |
673 | for p in sorted(self._dirs.keys() + self._files.keys()): |
|
665 | for p in sorted(self._dirs.keys() + self._files.keys()): | |
674 | if p in self._files: |
|
666 | if p in self._files: | |
675 | fullp = self._subpath(p) |
|
667 | fullp = self._subpath(p) | |
676 | if match(fullp): |
|
668 | if match(fullp): | |
677 | yield fullp |
|
669 | yield fullp | |
678 | else: |
|
670 | else: | |
679 |
for f in self._dirs[p]._walk(match |
|
671 | for f in self._dirs[p]._walk(match): | |
680 | yield f |
|
672 | yield f | |
681 |
|
673 | |||
682 | def matches(self, match): |
|
674 | def matches(self, match): | |
683 | '''generate a new manifest filtered by the match argument''' |
|
675 | '''generate a new manifest filtered by the match argument''' | |
684 | if match.always(): |
|
676 | if match.always(): | |
685 | return self.copy() |
|
677 | return self.copy() | |
686 |
|
678 | |||
687 | return self._matches(match) |
|
679 | return self._matches(match) | |
688 |
|
680 | |||
689 |
def _matches(self, match |
|
681 | def _matches(self, match): | |
690 | '''recursively generate a new manifest filtered by the match argument. |
|
682 | '''recursively generate a new manifest filtered by the match argument. | |
691 |
|
683 | ''' | ||
692 | Will visit all subdirectories if alldirs is True, otherwise it will |
|
|||
693 | only visit subdirectories for which match.visitdir is True.''' |
|
|||
694 |
|
||||
695 | ret = treemanifest(self._dir) |
|
684 | ret = treemanifest(self._dir) | |
696 | if not alldirs: |
|
685 | ||
697 | # substring to strip trailing slash |
|
686 | if not match.visitdir(self._dir[:-1] or '.'): | |
698 | visit = match.visitdir(self._dir[:-1] or '.') |
|
|||
699 | if not visit: |
|
|||
700 |
|
|
687 | return ret | |
701 | alldirs = (visit == 'all') |
|
|||
702 |
|
688 | |||
703 | for fn in self._files: |
|
689 | for fn in self._files: | |
704 | fullp = self._subpath(fn) |
|
690 | fullp = self._subpath(fn) | |
705 | if not match(fullp): |
|
691 | if not match(fullp): | |
706 | continue |
|
692 | continue | |
707 | ret._files[fn] = self._files[fn] |
|
693 | ret._files[fn] = self._files[fn] | |
708 | if fn in self._flags: |
|
694 | if fn in self._flags: | |
709 | ret._flags[fn] = self._flags[fn] |
|
695 | ret._flags[fn] = self._flags[fn] | |
710 |
|
696 | |||
711 | for dir, subm in self._dirs.iteritems(): |
|
697 | for dir, subm in self._dirs.iteritems(): | |
712 |
m = subm._matches(match |
|
698 | m = subm._matches(match) | |
713 | if not m._isempty(): |
|
699 | if not m._isempty(): | |
714 | ret._dirs[dir] = m |
|
700 | ret._dirs[dir] = m | |
715 |
|
701 | |||
716 | return ret |
|
702 | return ret | |
717 |
|
703 | |||
718 | def diff(self, m2, clean=False): |
|
704 | def diff(self, m2, clean=False): | |
719 | '''Finds changes between the current manifest and m2. |
|
705 | '''Finds changes between the current manifest and m2. | |
720 |
|
706 | |||
721 | Args: |
|
707 | Args: | |
722 | m2: the manifest to which this manifest should be compared. |
|
708 | m2: the manifest to which this manifest should be compared. | |
723 | clean: if true, include files unchanged between these manifests |
|
709 | clean: if true, include files unchanged between these manifests | |
724 | with a None value in the returned dictionary. |
|
710 | with a None value in the returned dictionary. | |
725 |
|
711 | |||
726 | The result is returned as a dict with filename as key and |
|
712 | The result is returned as a dict with filename as key and | |
727 | values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the |
|
713 | values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the | |
728 | nodeid in the current/other manifest and fl1/fl2 is the flag |
|
714 | nodeid in the current/other manifest and fl1/fl2 is the flag | |
729 | in the current/other manifest. Where the file does not exist, |
|
715 | in the current/other manifest. Where the file does not exist, | |
730 | the nodeid will be None and the flags will be the empty |
|
716 | the nodeid will be None and the flags will be the empty | |
731 | string. |
|
717 | string. | |
732 | ''' |
|
718 | ''' | |
733 | result = {} |
|
719 | result = {} | |
734 | emptytree = treemanifest() |
|
720 | emptytree = treemanifest() | |
735 | def _diff(t1, t2): |
|
721 | def _diff(t1, t2): | |
736 | for d, m1 in t1._dirs.iteritems(): |
|
722 | for d, m1 in t1._dirs.iteritems(): | |
737 | m2 = t2._dirs.get(d, emptytree) |
|
723 | m2 = t2._dirs.get(d, emptytree) | |
738 | _diff(m1, m2) |
|
724 | _diff(m1, m2) | |
739 |
|
725 | |||
740 | for d, m2 in t2._dirs.iteritems(): |
|
726 | for d, m2 in t2._dirs.iteritems(): | |
741 | if d not in t1._dirs: |
|
727 | if d not in t1._dirs: | |
742 | _diff(emptytree, m2) |
|
728 | _diff(emptytree, m2) | |
743 |
|
729 | |||
744 | for fn, n1 in t1._files.iteritems(): |
|
730 | for fn, n1 in t1._files.iteritems(): | |
745 | fl1 = t1._flags.get(fn, '') |
|
731 | fl1 = t1._flags.get(fn, '') | |
746 | n2 = t2._files.get(fn, None) |
|
732 | n2 = t2._files.get(fn, None) | |
747 | fl2 = t2._flags.get(fn, '') |
|
733 | fl2 = t2._flags.get(fn, '') | |
748 | if n1 != n2 or fl1 != fl2: |
|
734 | if n1 != n2 or fl1 != fl2: | |
749 | result[t1._subpath(fn)] = ((n1, fl1), (n2, fl2)) |
|
735 | result[t1._subpath(fn)] = ((n1, fl1), (n2, fl2)) | |
750 | elif clean: |
|
736 | elif clean: | |
751 | result[t1._subpath(fn)] = None |
|
737 | result[t1._subpath(fn)] = None | |
752 |
|
738 | |||
753 | for fn, n2 in t2._files.iteritems(): |
|
739 | for fn, n2 in t2._files.iteritems(): | |
754 | if fn not in t1._files: |
|
740 | if fn not in t1._files: | |
755 | fl2 = t2._flags.get(fn, '') |
|
741 | fl2 = t2._flags.get(fn, '') | |
756 | result[t2._subpath(fn)] = ((None, ''), (n2, fl2)) |
|
742 | result[t2._subpath(fn)] = ((None, ''), (n2, fl2)) | |
757 |
|
743 | |||
758 | _diff(self, m2) |
|
744 | _diff(self, m2) | |
759 | return result |
|
745 | return result | |
760 |
|
746 | |||
761 | def parse(self, text, readsubtree): |
|
747 | def parse(self, text, readsubtree): | |
762 | for f, n, fl in _parse(text): |
|
748 | for f, n, fl in _parse(text): | |
763 | if fl == 'd': |
|
749 | if fl == 'd': | |
764 | f = f + '/' |
|
750 | f = f + '/' | |
765 | self._dirs[f] = readsubtree(self._subpath(f), n) |
|
751 | self._dirs[f] = readsubtree(self._subpath(f), n) | |
766 | else: |
|
752 | else: | |
767 | # Use __setitem__ and setflag rather than assigning directly |
|
753 | # Use __setitem__ and setflag rather than assigning directly | |
768 | # to _files and _flags, thereby letting us parse flat manifests |
|
754 | # to _files and _flags, thereby letting us parse flat manifests | |
769 | # as well as tree manifests. |
|
755 | # as well as tree manifests. | |
770 | self[f] = n |
|
756 | self[f] = n | |
771 | if fl: |
|
757 | if fl: | |
772 | self.setflag(f, fl) |
|
758 | self.setflag(f, fl) | |
773 |
|
759 | |||
774 | def text(self, usemanifestv2=False): |
|
760 | def text(self, usemanifestv2=False): | |
775 | """Get the full data of this manifest as a bytestring.""" |
|
761 | """Get the full data of this manifest as a bytestring.""" | |
776 | flags = self.flags |
|
762 | flags = self.flags | |
777 | return _text(((f, self[f], flags(f)) for f in self.keys()), |
|
763 | return _text(((f, self[f], flags(f)) for f in self.keys()), | |
778 | usemanifestv2) |
|
764 | usemanifestv2) | |
779 |
|
765 | |||
780 | def dirtext(self, usemanifestv2=False): |
|
766 | def dirtext(self, usemanifestv2=False): | |
781 | """Get the full data of this directory as a bytestring. Make sure that |
|
767 | """Get the full data of this directory as a bytestring. Make sure that | |
782 | any submanifests have been written first, so their nodeids are correct. |
|
768 | any submanifests have been written first, so their nodeids are correct. | |
783 | """ |
|
769 | """ | |
784 | flags = self.flags |
|
770 | flags = self.flags | |
785 | dirs = [(d[:-1], self._dirs[d]._node, 'd') for d in self._dirs] |
|
771 | dirs = [(d[:-1], self._dirs[d]._node, 'd') for d in self._dirs] | |
786 | files = [(f, self._files[f], flags(f)) for f in self._files] |
|
772 | files = [(f, self._files[f], flags(f)) for f in self._files] | |
787 | return _text(sorted(dirs + files), usemanifestv2) |
|
773 | return _text(sorted(dirs + files), usemanifestv2) | |
788 |
|
774 | |||
789 | def writesubtrees(self, m1, m2, writesubtree): |
|
775 | def writesubtrees(self, m1, m2, writesubtree): | |
790 | emptytree = treemanifest() |
|
776 | emptytree = treemanifest() | |
791 | for d, subm in self._dirs.iteritems(): |
|
777 | for d, subm in self._dirs.iteritems(): | |
792 | subp1 = m1._dirs.get(d, emptytree)._node |
|
778 | subp1 = m1._dirs.get(d, emptytree)._node | |
793 | subp2 = m2._dirs.get(d, emptytree)._node |
|
779 | subp2 = m2._dirs.get(d, emptytree)._node | |
794 | if subp1 == revlog.nullid: |
|
780 | if subp1 == revlog.nullid: | |
795 | subp1, subp2 = subp2, subp1 |
|
781 | subp1, subp2 = subp2, subp1 | |
796 | writesubtree(subm, subp1, subp2) |
|
782 | writesubtree(subm, subp1, subp2) | |
797 |
|
783 | |||
798 | class manifest(revlog.revlog): |
|
784 | class manifest(revlog.revlog): | |
799 | def __init__(self, opener, dir='', dirlogcache=None): |
|
785 | def __init__(self, opener, dir='', dirlogcache=None): | |
800 | '''The 'dir' and 'dirlogcache' arguments are for internal use by |
|
786 | '''The 'dir' and 'dirlogcache' arguments are for internal use by | |
801 | manifest.manifest only. External users should create a root manifest |
|
787 | manifest.manifest only. External users should create a root manifest | |
802 | log with manifest.manifest(opener) and call dirlog() on it. |
|
788 | log with manifest.manifest(opener) and call dirlog() on it. | |
803 | ''' |
|
789 | ''' | |
804 | # During normal operations, we expect to deal with not more than four |
|
790 | # During normal operations, we expect to deal with not more than four | |
805 | # revs at a time (such as during commit --amend). When rebasing large |
|
791 | # revs at a time (such as during commit --amend). When rebasing large | |
806 | # stacks of commits, the number can go up, hence the config knob below. |
|
792 | # stacks of commits, the number can go up, hence the config knob below. | |
807 | cachesize = 4 |
|
793 | cachesize = 4 | |
808 | usetreemanifest = False |
|
794 | usetreemanifest = False | |
809 | usemanifestv2 = False |
|
795 | usemanifestv2 = False | |
810 | opts = getattr(opener, 'options', None) |
|
796 | opts = getattr(opener, 'options', None) | |
811 | if opts is not None: |
|
797 | if opts is not None: | |
812 | cachesize = opts.get('manifestcachesize', cachesize) |
|
798 | cachesize = opts.get('manifestcachesize', cachesize) | |
813 | usetreemanifest = opts.get('treemanifest', usetreemanifest) |
|
799 | usetreemanifest = opts.get('treemanifest', usetreemanifest) | |
814 | usemanifestv2 = opts.get('manifestv2', usemanifestv2) |
|
800 | usemanifestv2 = opts.get('manifestv2', usemanifestv2) | |
815 | self._mancache = util.lrucachedict(cachesize) |
|
801 | self._mancache = util.lrucachedict(cachesize) | |
816 | self._treeinmem = usetreemanifest |
|
802 | self._treeinmem = usetreemanifest | |
817 | self._treeondisk = usetreemanifest |
|
803 | self._treeondisk = usetreemanifest | |
818 | self._usemanifestv2 = usemanifestv2 |
|
804 | self._usemanifestv2 = usemanifestv2 | |
819 | indexfile = "00manifest.i" |
|
805 | indexfile = "00manifest.i" | |
820 | if dir: |
|
806 | if dir: | |
821 | assert self._treeondisk |
|
807 | assert self._treeondisk | |
822 | if not dir.endswith('/'): |
|
808 | if not dir.endswith('/'): | |
823 | dir = dir + '/' |
|
809 | dir = dir + '/' | |
824 | indexfile = "meta/" + dir + "00manifest.i" |
|
810 | indexfile = "meta/" + dir + "00manifest.i" | |
825 | revlog.revlog.__init__(self, opener, indexfile) |
|
811 | revlog.revlog.__init__(self, opener, indexfile) | |
826 | self._dir = dir |
|
812 | self._dir = dir | |
827 | # The dirlogcache is kept on the root manifest log |
|
813 | # The dirlogcache is kept on the root manifest log | |
828 | if dir: |
|
814 | if dir: | |
829 | self._dirlogcache = dirlogcache |
|
815 | self._dirlogcache = dirlogcache | |
830 | else: |
|
816 | else: | |
831 | self._dirlogcache = {'': self} |
|
817 | self._dirlogcache = {'': self} | |
832 |
|
818 | |||
833 | def _newmanifest(self, data=''): |
|
819 | def _newmanifest(self, data=''): | |
834 | if self._treeinmem: |
|
820 | if self._treeinmem: | |
835 | return treemanifest(self._dir, data) |
|
821 | return treemanifest(self._dir, data) | |
836 | return manifestdict(data) |
|
822 | return manifestdict(data) | |
837 |
|
823 | |||
838 | def dirlog(self, dir): |
|
824 | def dirlog(self, dir): | |
839 | assert self._treeondisk |
|
825 | assert self._treeondisk | |
840 | if dir not in self._dirlogcache: |
|
826 | if dir not in self._dirlogcache: | |
841 | self._dirlogcache[dir] = manifest(self.opener, dir, |
|
827 | self._dirlogcache[dir] = manifest(self.opener, dir, | |
842 | self._dirlogcache) |
|
828 | self._dirlogcache) | |
843 | return self._dirlogcache[dir] |
|
829 | return self._dirlogcache[dir] | |
844 |
|
830 | |||
845 | def _slowreaddelta(self, node): |
|
831 | def _slowreaddelta(self, node): | |
846 | r0 = self.deltaparent(self.rev(node)) |
|
832 | r0 = self.deltaparent(self.rev(node)) | |
847 | m0 = self.read(self.node(r0)) |
|
833 | m0 = self.read(self.node(r0)) | |
848 | m1 = self.read(node) |
|
834 | m1 = self.read(node) | |
849 | md = self._newmanifest() |
|
835 | md = self._newmanifest() | |
850 | for f, ((n0, fl0), (n1, fl1)) in m0.diff(m1).iteritems(): |
|
836 | for f, ((n0, fl0), (n1, fl1)) in m0.diff(m1).iteritems(): | |
851 | if n1: |
|
837 | if n1: | |
852 | md[f] = n1 |
|
838 | md[f] = n1 | |
853 | if fl1: |
|
839 | if fl1: | |
854 | md.setflag(f, fl1) |
|
840 | md.setflag(f, fl1) | |
855 | return md |
|
841 | return md | |
856 |
|
842 | |||
857 | def readdelta(self, node): |
|
843 | def readdelta(self, node): | |
858 | if self._usemanifestv2 or self._treeondisk: |
|
844 | if self._usemanifestv2 or self._treeondisk: | |
859 | return self._slowreaddelta(node) |
|
845 | return self._slowreaddelta(node) | |
860 | r = self.rev(node) |
|
846 | r = self.rev(node) | |
861 | d = mdiff.patchtext(self.revdiff(self.deltaparent(r), r)) |
|
847 | d = mdiff.patchtext(self.revdiff(self.deltaparent(r), r)) | |
862 | return self._newmanifest(d) |
|
848 | return self._newmanifest(d) | |
863 |
|
849 | |||
864 | def readfast(self, node): |
|
850 | def readfast(self, node): | |
865 | '''use the faster of readdelta or read |
|
851 | '''use the faster of readdelta or read | |
866 |
|
852 | |||
867 | This will return a manifest which is either only the files |
|
853 | This will return a manifest which is either only the files | |
868 | added/modified relative to p1, or all files in the |
|
854 | added/modified relative to p1, or all files in the | |
869 | manifest. Which one is returned depends on the codepath used |
|
855 | manifest. Which one is returned depends on the codepath used | |
870 | to retrieve the data. |
|
856 | to retrieve the data. | |
871 | ''' |
|
857 | ''' | |
872 | r = self.rev(node) |
|
858 | r = self.rev(node) | |
873 | deltaparent = self.deltaparent(r) |
|
859 | deltaparent = self.deltaparent(r) | |
874 | if deltaparent != revlog.nullrev and deltaparent in self.parentrevs(r): |
|
860 | if deltaparent != revlog.nullrev and deltaparent in self.parentrevs(r): | |
875 | return self.readdelta(node) |
|
861 | return self.readdelta(node) | |
876 | return self.read(node) |
|
862 | return self.read(node) | |
877 |
|
863 | |||
878 | def read(self, node): |
|
864 | def read(self, node): | |
879 | if node == revlog.nullid: |
|
865 | if node == revlog.nullid: | |
880 | return self._newmanifest() # don't upset local cache |
|
866 | return self._newmanifest() # don't upset local cache | |
881 | if node in self._mancache: |
|
867 | if node in self._mancache: | |
882 | return self._mancache[node][0] |
|
868 | return self._mancache[node][0] | |
883 | text = self.revision(node) |
|
869 | text = self.revision(node) | |
884 | if self._treeondisk: |
|
870 | if self._treeondisk: | |
885 | def readsubtree(dir, subm): |
|
871 | def readsubtree(dir, subm): | |
886 | return self.dirlog(dir).read(subm) |
|
872 | return self.dirlog(dir).read(subm) | |
887 | m = self._newmanifest() |
|
873 | m = self._newmanifest() | |
888 | m.parse(text, readsubtree) |
|
874 | m.parse(text, readsubtree) | |
889 | m.setnode(node) |
|
875 | m.setnode(node) | |
890 | arraytext = None |
|
876 | arraytext = None | |
891 | else: |
|
877 | else: | |
892 | m = self._newmanifest(text) |
|
878 | m = self._newmanifest(text) | |
893 | arraytext = array.array('c', text) |
|
879 | arraytext = array.array('c', text) | |
894 | self._mancache[node] = (m, arraytext) |
|
880 | self._mancache[node] = (m, arraytext) | |
895 | return m |
|
881 | return m | |
896 |
|
882 | |||
897 | def find(self, node, f): |
|
883 | def find(self, node, f): | |
898 | '''look up entry for a single file efficiently. |
|
884 | '''look up entry for a single file efficiently. | |
899 | return (node, flags) pair if found, (None, None) if not.''' |
|
885 | return (node, flags) pair if found, (None, None) if not.''' | |
900 | m = self.read(node) |
|
886 | m = self.read(node) | |
901 | try: |
|
887 | try: | |
902 | return m.find(f) |
|
888 | return m.find(f) | |
903 | except KeyError: |
|
889 | except KeyError: | |
904 | return None, None |
|
890 | return None, None | |
905 |
|
891 | |||
906 | def add(self, m, transaction, link, p1, p2, added, removed): |
|
892 | def add(self, m, transaction, link, p1, p2, added, removed): | |
907 | if (p1 in self._mancache and not self._treeinmem |
|
893 | if (p1 in self._mancache and not self._treeinmem | |
908 | and not self._usemanifestv2): |
|
894 | and not self._usemanifestv2): | |
909 | # If our first parent is in the manifest cache, we can |
|
895 | # If our first parent is in the manifest cache, we can | |
910 | # compute a delta here using properties we know about the |
|
896 | # compute a delta here using properties we know about the | |
911 | # manifest up-front, which may save time later for the |
|
897 | # manifest up-front, which may save time later for the | |
912 | # revlog layer. |
|
898 | # revlog layer. | |
913 |
|
899 | |||
914 | _checkforbidden(added) |
|
900 | _checkforbidden(added) | |
915 | # combine the changed lists into one list for sorting |
|
901 | # combine the changed lists into one list for sorting | |
916 | work = [(x, False) for x in added] |
|
902 | work = [(x, False) for x in added] | |
917 | work.extend((x, True) for x in removed) |
|
903 | work.extend((x, True) for x in removed) | |
918 | # this could use heapq.merge() (from Python 2.6+) or equivalent |
|
904 | # this could use heapq.merge() (from Python 2.6+) or equivalent | |
919 | # since the lists are already sorted |
|
905 | # since the lists are already sorted | |
920 | work.sort() |
|
906 | work.sort() | |
921 |
|
907 | |||
922 | arraytext, deltatext = m.fastdelta(self._mancache[p1][1], work) |
|
908 | arraytext, deltatext = m.fastdelta(self._mancache[p1][1], work) | |
923 | cachedelta = self.rev(p1), deltatext |
|
909 | cachedelta = self.rev(p1), deltatext | |
924 | text = util.buffer(arraytext) |
|
910 | text = util.buffer(arraytext) | |
925 | n = self.addrevision(text, transaction, link, p1, p2, cachedelta) |
|
911 | n = self.addrevision(text, transaction, link, p1, p2, cachedelta) | |
926 | else: |
|
912 | else: | |
927 | # The first parent manifest isn't already loaded, so we'll |
|
913 | # The first parent manifest isn't already loaded, so we'll | |
928 | # just encode a fulltext of the manifest and pass that |
|
914 | # just encode a fulltext of the manifest and pass that | |
929 | # through to the revlog layer, and let it handle the delta |
|
915 | # through to the revlog layer, and let it handle the delta | |
930 | # process. |
|
916 | # process. | |
931 | if self._treeondisk: |
|
917 | if self._treeondisk: | |
932 | m1 = self.read(p1) |
|
918 | m1 = self.read(p1) | |
933 | m2 = self.read(p2) |
|
919 | m2 = self.read(p2) | |
934 | n = self._addtree(m, transaction, link, m1, m2) |
|
920 | n = self._addtree(m, transaction, link, m1, m2) | |
935 | arraytext = None |
|
921 | arraytext = None | |
936 | else: |
|
922 | else: | |
937 | text = m.text(self._usemanifestv2) |
|
923 | text = m.text(self._usemanifestv2) | |
938 | n = self.addrevision(text, transaction, link, p1, p2) |
|
924 | n = self.addrevision(text, transaction, link, p1, p2) | |
939 | arraytext = array.array('c', text) |
|
925 | arraytext = array.array('c', text) | |
940 |
|
926 | |||
941 | self._mancache[n] = (m, arraytext) |
|
927 | self._mancache[n] = (m, arraytext) | |
942 |
|
928 | |||
943 | return n |
|
929 | return n | |
944 |
|
930 | |||
945 | def _addtree(self, m, transaction, link, m1, m2): |
|
931 | def _addtree(self, m, transaction, link, m1, m2): | |
946 | def writesubtree(subm, subp1, subp2): |
|
932 | def writesubtree(subm, subp1, subp2): | |
947 | sublog = self.dirlog(subm.dir()) |
|
933 | sublog = self.dirlog(subm.dir()) | |
948 | sublog.add(subm, transaction, link, subp1, subp2, None, None) |
|
934 | sublog.add(subm, transaction, link, subp1, subp2, None, None) | |
949 | m.writesubtrees(m1, m2, writesubtree) |
|
935 | m.writesubtrees(m1, m2, writesubtree) | |
950 | text = m.dirtext(self._usemanifestv2) |
|
936 | text = m.dirtext(self._usemanifestv2) | |
951 | # If the manifest is unchanged compared to one parent, |
|
937 | # If the manifest is unchanged compared to one parent, | |
952 | # don't write a new revision |
|
938 | # don't write a new revision | |
953 | if text == m1.dirtext(self._usemanifestv2): |
|
939 | if text == m1.dirtext(self._usemanifestv2): | |
954 | n = m1.node() |
|
940 | n = m1.node() | |
955 | elif text == m2.dirtext(self._usemanifestv2): |
|
941 | elif text == m2.dirtext(self._usemanifestv2): | |
956 | n = m2.node() |
|
942 | n = m2.node() | |
957 | else: |
|
943 | else: | |
958 | n = self.addrevision(text, transaction, link, m1.node(), m2.node()) |
|
944 | n = self.addrevision(text, transaction, link, m1.node(), m2.node()) | |
959 | # Save nodeid so parent manifest can calculate its nodeid |
|
945 | # Save nodeid so parent manifest can calculate its nodeid | |
960 | m.setnode(n) |
|
946 | m.setnode(n) | |
961 | return n |
|
947 | return n |
@@ -1,546 +1,542 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 | import re |
|
8 | import re | |
9 | import util, pathutil |
|
9 | import util, pathutil | |
10 | from i18n import _ |
|
10 | from i18n import _ | |
11 |
|
11 | |||
12 | propertycache = util.propertycache |
|
12 | propertycache = util.propertycache | |
13 |
|
13 | |||
14 | def _rematcher(regex): |
|
14 | def _rematcher(regex): | |
15 | '''compile the regexp with the best available regexp engine and return a |
|
15 | '''compile the regexp with the best available regexp engine and return a | |
16 | matcher function''' |
|
16 | matcher function''' | |
17 | m = util.re.compile(regex) |
|
17 | m = util.re.compile(regex) | |
18 | try: |
|
18 | try: | |
19 | # slightly faster, provided by facebook's re2 bindings |
|
19 | # slightly faster, provided by facebook's re2 bindings | |
20 | return m.test_match |
|
20 | return m.test_match | |
21 | except AttributeError: |
|
21 | except AttributeError: | |
22 | return m.match |
|
22 | return m.match | |
23 |
|
23 | |||
24 | def _expandsets(kindpats, ctx, listsubrepos): |
|
24 | def _expandsets(kindpats, ctx, listsubrepos): | |
25 | '''Returns the kindpats list with the 'set' patterns expanded.''' |
|
25 | '''Returns the kindpats list with the 'set' patterns expanded.''' | |
26 | fset = set() |
|
26 | fset = set() | |
27 | other = [] |
|
27 | other = [] | |
28 |
|
28 | |||
29 | for kind, pat in kindpats: |
|
29 | for kind, pat in kindpats: | |
30 | if kind == 'set': |
|
30 | if kind == 'set': | |
31 | if not ctx: |
|
31 | if not ctx: | |
32 | raise util.Abort("fileset expression with no context") |
|
32 | raise util.Abort("fileset expression with no context") | |
33 | s = ctx.getfileset(pat) |
|
33 | s = ctx.getfileset(pat) | |
34 | fset.update(s) |
|
34 | fset.update(s) | |
35 |
|
35 | |||
36 | if listsubrepos: |
|
36 | if listsubrepos: | |
37 | for subpath in ctx.substate: |
|
37 | for subpath in ctx.substate: | |
38 | s = ctx.sub(subpath).getfileset(pat) |
|
38 | s = ctx.sub(subpath).getfileset(pat) | |
39 | fset.update(subpath + '/' + f for f in s) |
|
39 | fset.update(subpath + '/' + f for f in s) | |
40 |
|
40 | |||
41 | continue |
|
41 | continue | |
42 | other.append((kind, pat)) |
|
42 | other.append((kind, pat)) | |
43 | return fset, other |
|
43 | return fset, other | |
44 |
|
44 | |||
45 | def _kindpatsalwaysmatch(kindpats): |
|
45 | def _kindpatsalwaysmatch(kindpats): | |
46 | """"Checks whether the kindspats match everything, as e.g. |
|
46 | """"Checks whether the kindspats match everything, as e.g. | |
47 | 'relpath:.' does. |
|
47 | 'relpath:.' does. | |
48 | """ |
|
48 | """ | |
49 | for kind, pat in kindpats: |
|
49 | for kind, pat in kindpats: | |
50 | if pat != '' or kind not in ['relpath', 'glob']: |
|
50 | if pat != '' or kind not in ['relpath', 'glob']: | |
51 | return False |
|
51 | return False | |
52 | return True |
|
52 | return True | |
53 |
|
53 | |||
54 | class match(object): |
|
54 | class match(object): | |
55 | def __init__(self, root, cwd, patterns, include=[], exclude=[], |
|
55 | def __init__(self, root, cwd, patterns, include=[], exclude=[], | |
56 | default='glob', exact=False, auditor=None, ctx=None, |
|
56 | default='glob', exact=False, auditor=None, ctx=None, | |
57 | listsubrepos=False): |
|
57 | listsubrepos=False): | |
58 | """build an object to match a set of file patterns |
|
58 | """build an object to match a set of file patterns | |
59 |
|
59 | |||
60 | arguments: |
|
60 | arguments: | |
61 | root - the canonical root of the tree you're matching against |
|
61 | root - the canonical root of the tree you're matching against | |
62 | cwd - the current working directory, if relevant |
|
62 | cwd - the current working directory, if relevant | |
63 | patterns - patterns to find |
|
63 | patterns - patterns to find | |
64 | include - patterns to include (unless they are excluded) |
|
64 | include - patterns to include (unless they are excluded) | |
65 | exclude - patterns to exclude (even if they are included) |
|
65 | exclude - patterns to exclude (even if they are included) | |
66 | default - if a pattern in patterns has no explicit type, assume this one |
|
66 | default - if a pattern in patterns has no explicit type, assume this one | |
67 | exact - patterns are actually filenames (include/exclude still apply) |
|
67 | exact - patterns are actually filenames (include/exclude still apply) | |
68 |
|
68 | |||
69 | a pattern is one of: |
|
69 | a pattern is one of: | |
70 | 'glob:<glob>' - a glob relative to cwd |
|
70 | 'glob:<glob>' - a glob relative to cwd | |
71 | 're:<regexp>' - a regular expression |
|
71 | 're:<regexp>' - a regular expression | |
72 | 'path:<path>' - a path relative to repository root |
|
72 | 'path:<path>' - a path relative to repository root | |
73 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
73 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
74 | 'relpath:<path>' - a path relative to cwd |
|
74 | 'relpath:<path>' - a path relative to cwd | |
75 | 'relre:<regexp>' - a regexp that needn't match the start of a name |
|
75 | 'relre:<regexp>' - a regexp that needn't match the start of a name | |
76 | 'set:<fileset>' - a fileset expression |
|
76 | 'set:<fileset>' - a fileset expression | |
77 | '<something>' - a pattern of the specified default type |
|
77 | '<something>' - a pattern of the specified default type | |
78 | """ |
|
78 | """ | |
79 |
|
79 | |||
80 | self._root = root |
|
80 | self._root = root | |
81 | self._cwd = cwd |
|
81 | self._cwd = cwd | |
82 | self._files = [] # exact files and roots of patterns |
|
82 | self._files = [] # exact files and roots of patterns | |
83 | self._anypats = bool(include or exclude) |
|
83 | self._anypats = bool(include or exclude) | |
84 | self._always = False |
|
84 | self._always = False | |
85 | self._pathrestricted = bool(include or exclude or patterns) |
|
85 | self._pathrestricted = bool(include or exclude or patterns) | |
86 |
|
86 | |||
87 | matchfns = [] |
|
87 | matchfns = [] | |
88 | if include: |
|
88 | if include: | |
89 | kindpats = self._normalize(include, 'glob', root, cwd, auditor) |
|
89 | kindpats = self._normalize(include, 'glob', root, cwd, auditor) | |
90 | self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)', |
|
90 | self.includepat, im = _buildmatch(ctx, kindpats, '(?:/|$)', | |
91 | listsubrepos) |
|
91 | listsubrepos) | |
92 | matchfns.append(im) |
|
92 | matchfns.append(im) | |
93 | if exclude: |
|
93 | if exclude: | |
94 | kindpats = self._normalize(exclude, 'glob', root, cwd, auditor) |
|
94 | kindpats = self._normalize(exclude, 'glob', root, cwd, auditor) | |
95 | self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)', |
|
95 | self.excludepat, em = _buildmatch(ctx, kindpats, '(?:/|$)', | |
96 | listsubrepos) |
|
96 | listsubrepos) | |
97 | matchfns.append(lambda f: not em(f)) |
|
97 | matchfns.append(lambda f: not em(f)) | |
98 | if exact: |
|
98 | if exact: | |
99 | if isinstance(patterns, list): |
|
99 | if isinstance(patterns, list): | |
100 | self._files = patterns |
|
100 | self._files = patterns | |
101 | else: |
|
101 | else: | |
102 | self._files = list(patterns) |
|
102 | self._files = list(patterns) | |
103 | matchfns.append(self.exact) |
|
103 | matchfns.append(self.exact) | |
104 | elif patterns: |
|
104 | elif patterns: | |
105 | kindpats = self._normalize(patterns, default, root, cwd, auditor) |
|
105 | kindpats = self._normalize(patterns, default, root, cwd, auditor) | |
106 | if not _kindpatsalwaysmatch(kindpats): |
|
106 | if not _kindpatsalwaysmatch(kindpats): | |
107 | self._files = _roots(kindpats) |
|
107 | self._files = _roots(kindpats) | |
108 | self._anypats = self._anypats or _anypats(kindpats) |
|
108 | self._anypats = self._anypats or _anypats(kindpats) | |
109 | self.patternspat, pm = _buildmatch(ctx, kindpats, '$', |
|
109 | self.patternspat, pm = _buildmatch(ctx, kindpats, '$', | |
110 | listsubrepos) |
|
110 | listsubrepos) | |
111 | matchfns.append(pm) |
|
111 | matchfns.append(pm) | |
112 |
|
112 | |||
113 | if not matchfns: |
|
113 | if not matchfns: | |
114 | m = util.always |
|
114 | m = util.always | |
115 | self._always = True |
|
115 | self._always = True | |
116 | elif len(matchfns) == 1: |
|
116 | elif len(matchfns) == 1: | |
117 | m = matchfns[0] |
|
117 | m = matchfns[0] | |
118 | else: |
|
118 | else: | |
119 | def m(f): |
|
119 | def m(f): | |
120 | for matchfn in matchfns: |
|
120 | for matchfn in matchfns: | |
121 | if not matchfn(f): |
|
121 | if not matchfn(f): | |
122 | return False |
|
122 | return False | |
123 | return True |
|
123 | return True | |
124 |
|
124 | |||
125 | self.matchfn = m |
|
125 | self.matchfn = m | |
126 | self._fmap = set(self._files) |
|
126 | self._fmap = set(self._files) | |
127 |
|
127 | |||
128 | def __call__(self, fn): |
|
128 | def __call__(self, fn): | |
129 | return self.matchfn(fn) |
|
129 | return self.matchfn(fn) | |
130 | def __iter__(self): |
|
130 | def __iter__(self): | |
131 | for f in self._files: |
|
131 | for f in self._files: | |
132 | yield f |
|
132 | yield f | |
133 |
|
133 | |||
134 | # Callbacks related to how the matcher is used by dirstate.walk. |
|
134 | # Callbacks related to how the matcher is used by dirstate.walk. | |
135 | # Subscribers to these events must monkeypatch the matcher object. |
|
135 | # Subscribers to these events must monkeypatch the matcher object. | |
136 | def bad(self, f, msg): |
|
136 | def bad(self, f, msg): | |
137 | '''Callback from dirstate.walk for each explicit file that can't be |
|
137 | '''Callback from dirstate.walk for each explicit file that can't be | |
138 | found/accessed, with an error message.''' |
|
138 | found/accessed, with an error message.''' | |
139 | pass |
|
139 | pass | |
140 |
|
140 | |||
141 | # If an explicitdir is set, it will be called when an explicitly listed |
|
141 | # If an explicitdir is set, it will be called when an explicitly listed | |
142 | # directory is visited. |
|
142 | # directory is visited. | |
143 | explicitdir = None |
|
143 | explicitdir = None | |
144 |
|
144 | |||
145 | # If an traversedir is set, it will be called when a directory discovered |
|
145 | # If an traversedir is set, it will be called when a directory discovered | |
146 | # by recursive traversal is visited. |
|
146 | # by recursive traversal is visited. | |
147 | traversedir = None |
|
147 | traversedir = None | |
148 |
|
148 | |||
149 | def abs(self, f): |
|
149 | def abs(self, f): | |
150 | '''Convert a repo path back to path that is relative to the root of the |
|
150 | '''Convert a repo path back to path that is relative to the root of the | |
151 | matcher.''' |
|
151 | matcher.''' | |
152 | return f |
|
152 | return f | |
153 |
|
153 | |||
154 | def rel(self, f): |
|
154 | def rel(self, f): | |
155 | '''Convert repo path back to path that is relative to cwd of matcher.''' |
|
155 | '''Convert repo path back to path that is relative to cwd of matcher.''' | |
156 | return util.pathto(self._root, self._cwd, f) |
|
156 | return util.pathto(self._root, self._cwd, f) | |
157 |
|
157 | |||
158 | def uipath(self, f): |
|
158 | def uipath(self, f): | |
159 | '''Convert repo path to a display path. If patterns or -I/-X were used |
|
159 | '''Convert repo path to a display path. If patterns or -I/-X were used | |
160 | to create this matcher, the display path will be relative to cwd. |
|
160 | to create this matcher, the display path will be relative to cwd. | |
161 | Otherwise it is relative to the root of the repo.''' |
|
161 | Otherwise it is relative to the root of the repo.''' | |
162 | return (self._pathrestricted and self.rel(f)) or self.abs(f) |
|
162 | return (self._pathrestricted and self.rel(f)) or self.abs(f) | |
163 |
|
163 | |||
164 | def files(self): |
|
164 | def files(self): | |
165 | '''Explicitly listed files or patterns or roots: |
|
165 | '''Explicitly listed files or patterns or roots: | |
166 | if no patterns or .always(): empty list, |
|
166 | if no patterns or .always(): empty list, | |
167 | if exact: list exact files, |
|
167 | if exact: list exact files, | |
168 | if not .anypats(): list all files and dirs, |
|
168 | if not .anypats(): list all files and dirs, | |
169 | else: optimal roots''' |
|
169 | else: optimal roots''' | |
170 | return self._files |
|
170 | return self._files | |
171 |
|
171 | |||
172 | @propertycache |
|
172 | @propertycache | |
173 | def _dirs(self): |
|
173 | def _dirs(self): | |
174 | return set(util.dirs(self._fmap)) | set(['.']) |
|
174 | return set(util.dirs(self._fmap)) | set(['.']) | |
175 |
|
175 | |||
176 | def visitdir(self, dir): |
|
176 | def visitdir(self, dir): | |
177 | '''Helps while traversing a directory tree. Returns the string 'all' if |
|
177 | return (not self._fmap or '.' in self._fmap or | |
178 | the given directory and all subdirectories should be visited. Otherwise |
|
178 | dir in self._fmap or dir in self._dirs or | |
179 | returns True or False indicating whether the given directory should be |
|
179 | any(parentdir in self._fmap | |
180 | visited. If 'all' is returned, calling this method on a subdirectory |
|
180 | for parentdir in util.finddirs(dir))) | |
181 | gives an undefined result.''' |
|
|||
182 | if not self._fmap or self.exact(dir): |
|
|||
183 | return 'all' |
|
|||
184 | return dir in self._dirs |
|
|||
185 |
|
181 | |||
186 | def exact(self, f): |
|
182 | def exact(self, f): | |
187 | '''Returns True if f is in .files().''' |
|
183 | '''Returns True if f is in .files().''' | |
188 | return f in self._fmap |
|
184 | return f in self._fmap | |
189 |
|
185 | |||
190 | def anypats(self): |
|
186 | def anypats(self): | |
191 | '''Matcher uses patterns or include/exclude.''' |
|
187 | '''Matcher uses patterns or include/exclude.''' | |
192 | return self._anypats |
|
188 | return self._anypats | |
193 |
|
189 | |||
194 | def always(self): |
|
190 | def always(self): | |
195 | '''Matcher will match everything and .files() will be empty |
|
191 | '''Matcher will match everything and .files() will be empty | |
196 | - optimization might be possible and necessary.''' |
|
192 | - optimization might be possible and necessary.''' | |
197 | return self._always |
|
193 | return self._always | |
198 |
|
194 | |||
199 | def ispartial(self): |
|
195 | def ispartial(self): | |
200 | '''True if the matcher won't always match. |
|
196 | '''True if the matcher won't always match. | |
201 |
|
197 | |||
202 | Although it's just the inverse of _always in this implementation, |
|
198 | Although it's just the inverse of _always in this implementation, | |
203 | an extenion such as narrowhg might make it return something |
|
199 | an extenion such as narrowhg might make it return something | |
204 | slightly different.''' |
|
200 | slightly different.''' | |
205 | return not self._always |
|
201 | return not self._always | |
206 |
|
202 | |||
207 | def isexact(self): |
|
203 | def isexact(self): | |
208 | return self.matchfn == self.exact |
|
204 | return self.matchfn == self.exact | |
209 |
|
205 | |||
210 | def _normalize(self, patterns, default, root, cwd, auditor): |
|
206 | def _normalize(self, patterns, default, root, cwd, auditor): | |
211 | '''Convert 'kind:pat' from the patterns list to tuples with kind and |
|
207 | '''Convert 'kind:pat' from the patterns list to tuples with kind and | |
212 | normalized and rooted patterns and with listfiles expanded.''' |
|
208 | normalized and rooted patterns and with listfiles expanded.''' | |
213 | kindpats = [] |
|
209 | kindpats = [] | |
214 | for kind, pat in [_patsplit(p, default) for p in patterns]: |
|
210 | for kind, pat in [_patsplit(p, default) for p in patterns]: | |
215 | if kind in ('glob', 'relpath'): |
|
211 | if kind in ('glob', 'relpath'): | |
216 | pat = pathutil.canonpath(root, cwd, pat, auditor) |
|
212 | pat = pathutil.canonpath(root, cwd, pat, auditor) | |
217 | elif kind in ('relglob', 'path'): |
|
213 | elif kind in ('relglob', 'path'): | |
218 | pat = util.normpath(pat) |
|
214 | pat = util.normpath(pat) | |
219 | elif kind in ('listfile', 'listfile0'): |
|
215 | elif kind in ('listfile', 'listfile0'): | |
220 | try: |
|
216 | try: | |
221 | files = util.readfile(pat) |
|
217 | files = util.readfile(pat) | |
222 | if kind == 'listfile0': |
|
218 | if kind == 'listfile0': | |
223 | files = files.split('\0') |
|
219 | files = files.split('\0') | |
224 | else: |
|
220 | else: | |
225 | files = files.splitlines() |
|
221 | files = files.splitlines() | |
226 | files = [f for f in files if f] |
|
222 | files = [f for f in files if f] | |
227 | except EnvironmentError: |
|
223 | except EnvironmentError: | |
228 | raise util.Abort(_("unable to read file list (%s)") % pat) |
|
224 | raise util.Abort(_("unable to read file list (%s)") % pat) | |
229 | kindpats += self._normalize(files, default, root, cwd, auditor) |
|
225 | kindpats += self._normalize(files, default, root, cwd, auditor) | |
230 | continue |
|
226 | continue | |
231 | # else: re or relre - which cannot be normalized |
|
227 | # else: re or relre - which cannot be normalized | |
232 | kindpats.append((kind, pat)) |
|
228 | kindpats.append((kind, pat)) | |
233 | return kindpats |
|
229 | return kindpats | |
234 |
|
230 | |||
235 | def exact(root, cwd, files): |
|
231 | def exact(root, cwd, files): | |
236 | return match(root, cwd, files, exact=True) |
|
232 | return match(root, cwd, files, exact=True) | |
237 |
|
233 | |||
238 | def always(root, cwd): |
|
234 | def always(root, cwd): | |
239 | return match(root, cwd, []) |
|
235 | return match(root, cwd, []) | |
240 |
|
236 | |||
241 | class narrowmatcher(match): |
|
237 | class narrowmatcher(match): | |
242 | """Adapt a matcher to work on a subdirectory only. |
|
238 | """Adapt a matcher to work on a subdirectory only. | |
243 |
|
239 | |||
244 | The paths are remapped to remove/insert the path as needed: |
|
240 | The paths are remapped to remove/insert the path as needed: | |
245 |
|
241 | |||
246 | >>> m1 = match('root', '', ['a.txt', 'sub/b.txt']) |
|
242 | >>> m1 = match('root', '', ['a.txt', 'sub/b.txt']) | |
247 | >>> m2 = narrowmatcher('sub', m1) |
|
243 | >>> m2 = narrowmatcher('sub', m1) | |
248 | >>> bool(m2('a.txt')) |
|
244 | >>> bool(m2('a.txt')) | |
249 | False |
|
245 | False | |
250 | >>> bool(m2('b.txt')) |
|
246 | >>> bool(m2('b.txt')) | |
251 | True |
|
247 | True | |
252 | >>> bool(m2.matchfn('a.txt')) |
|
248 | >>> bool(m2.matchfn('a.txt')) | |
253 | False |
|
249 | False | |
254 | >>> bool(m2.matchfn('b.txt')) |
|
250 | >>> bool(m2.matchfn('b.txt')) | |
255 | True |
|
251 | True | |
256 | >>> m2.files() |
|
252 | >>> m2.files() | |
257 | ['b.txt'] |
|
253 | ['b.txt'] | |
258 | >>> m2.exact('b.txt') |
|
254 | >>> m2.exact('b.txt') | |
259 | True |
|
255 | True | |
260 | >>> util.pconvert(m2.rel('b.txt')) |
|
256 | >>> util.pconvert(m2.rel('b.txt')) | |
261 | 'sub/b.txt' |
|
257 | 'sub/b.txt' | |
262 | >>> def bad(f, msg): |
|
258 | >>> def bad(f, msg): | |
263 | ... print "%s: %s" % (f, msg) |
|
259 | ... print "%s: %s" % (f, msg) | |
264 | >>> m1.bad = bad |
|
260 | >>> m1.bad = bad | |
265 | >>> m2.bad('x.txt', 'No such file') |
|
261 | >>> m2.bad('x.txt', 'No such file') | |
266 | sub/x.txt: No such file |
|
262 | sub/x.txt: No such file | |
267 | >>> m2.abs('c.txt') |
|
263 | >>> m2.abs('c.txt') | |
268 | 'sub/c.txt' |
|
264 | 'sub/c.txt' | |
269 | """ |
|
265 | """ | |
270 |
|
266 | |||
271 | def __init__(self, path, matcher): |
|
267 | def __init__(self, path, matcher): | |
272 | self._root = matcher._root |
|
268 | self._root = matcher._root | |
273 | self._cwd = matcher._cwd |
|
269 | self._cwd = matcher._cwd | |
274 | self._path = path |
|
270 | self._path = path | |
275 | self._matcher = matcher |
|
271 | self._matcher = matcher | |
276 | self._always = matcher._always |
|
272 | self._always = matcher._always | |
277 | self._pathrestricted = matcher._pathrestricted |
|
273 | self._pathrestricted = matcher._pathrestricted | |
278 |
|
274 | |||
279 | self._files = [f[len(path) + 1:] for f in matcher._files |
|
275 | self._files = [f[len(path) + 1:] for f in matcher._files | |
280 | if f.startswith(path + "/")] |
|
276 | if f.startswith(path + "/")] | |
281 | self._anypats = matcher._anypats |
|
277 | self._anypats = matcher._anypats | |
282 | self.matchfn = lambda fn: matcher.matchfn(self._path + "/" + fn) |
|
278 | self.matchfn = lambda fn: matcher.matchfn(self._path + "/" + fn) | |
283 | self._fmap = set(self._files) |
|
279 | self._fmap = set(self._files) | |
284 |
|
280 | |||
285 | def abs(self, f): |
|
281 | def abs(self, f): | |
286 | return self._matcher.abs(self._path + "/" + f) |
|
282 | return self._matcher.abs(self._path + "/" + f) | |
287 |
|
283 | |||
288 | def bad(self, f, msg): |
|
284 | def bad(self, f, msg): | |
289 | self._matcher.bad(self._path + "/" + f, msg) |
|
285 | self._matcher.bad(self._path + "/" + f, msg) | |
290 |
|
286 | |||
291 | def rel(self, f): |
|
287 | def rel(self, f): | |
292 | return self._matcher.rel(self._path + "/" + f) |
|
288 | return self._matcher.rel(self._path + "/" + f) | |
293 |
|
289 | |||
294 | class icasefsmatcher(match): |
|
290 | class icasefsmatcher(match): | |
295 | """A matcher for wdir on case insensitive filesystems, which normalizes the |
|
291 | """A matcher for wdir on case insensitive filesystems, which normalizes the | |
296 | given patterns to the case in the filesystem. |
|
292 | given patterns to the case in the filesystem. | |
297 | """ |
|
293 | """ | |
298 |
|
294 | |||
299 | def __init__(self, root, cwd, patterns, include, exclude, default, auditor, |
|
295 | def __init__(self, root, cwd, patterns, include, exclude, default, auditor, | |
300 | ctx, listsubrepos=False): |
|
296 | ctx, listsubrepos=False): | |
301 | init = super(icasefsmatcher, self).__init__ |
|
297 | init = super(icasefsmatcher, self).__init__ | |
302 | self._dsnormalize = ctx.repo().dirstate.normalize |
|
298 | self._dsnormalize = ctx.repo().dirstate.normalize | |
303 |
|
299 | |||
304 | init(root, cwd, patterns, include, exclude, default, auditor=auditor, |
|
300 | init(root, cwd, patterns, include, exclude, default, auditor=auditor, | |
305 | ctx=ctx, listsubrepos=listsubrepos) |
|
301 | ctx=ctx, listsubrepos=listsubrepos) | |
306 |
|
302 | |||
307 | # m.exact(file) must be based off of the actual user input, otherwise |
|
303 | # m.exact(file) must be based off of the actual user input, otherwise | |
308 | # inexact case matches are treated as exact, and not noted without -v. |
|
304 | # inexact case matches are treated as exact, and not noted without -v. | |
309 | if self._files: |
|
305 | if self._files: | |
310 | self._fmap = set(_roots(self._kp)) |
|
306 | self._fmap = set(_roots(self._kp)) | |
311 |
|
307 | |||
312 | def _normalize(self, patterns, default, root, cwd, auditor): |
|
308 | def _normalize(self, patterns, default, root, cwd, auditor): | |
313 | self._kp = super(icasefsmatcher, self)._normalize(patterns, default, |
|
309 | self._kp = super(icasefsmatcher, self)._normalize(patterns, default, | |
314 | root, cwd, auditor) |
|
310 | root, cwd, auditor) | |
315 | kindpats = [] |
|
311 | kindpats = [] | |
316 | for kind, pats in self._kp: |
|
312 | for kind, pats in self._kp: | |
317 | if kind not in ('re', 'relre'): # regex can't be normalized |
|
313 | if kind not in ('re', 'relre'): # regex can't be normalized | |
318 | pats = self._dsnormalize(pats) |
|
314 | pats = self._dsnormalize(pats) | |
319 | kindpats.append((kind, pats)) |
|
315 | kindpats.append((kind, pats)) | |
320 | return kindpats |
|
316 | return kindpats | |
321 |
|
317 | |||
322 | def patkind(pattern, default=None): |
|
318 | def patkind(pattern, default=None): | |
323 | '''If pattern is 'kind:pat' with a known kind, return kind.''' |
|
319 | '''If pattern is 'kind:pat' with a known kind, return kind.''' | |
324 | return _patsplit(pattern, default)[0] |
|
320 | return _patsplit(pattern, default)[0] | |
325 |
|
321 | |||
326 | def _patsplit(pattern, default): |
|
322 | def _patsplit(pattern, default): | |
327 | """Split a string into the optional pattern kind prefix and the actual |
|
323 | """Split a string into the optional pattern kind prefix and the actual | |
328 | pattern.""" |
|
324 | pattern.""" | |
329 | if ':' in pattern: |
|
325 | if ':' in pattern: | |
330 | kind, pat = pattern.split(':', 1) |
|
326 | kind, pat = pattern.split(':', 1) | |
331 | if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre', |
|
327 | if kind in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre', | |
332 | 'listfile', 'listfile0', 'set'): |
|
328 | 'listfile', 'listfile0', 'set'): | |
333 | return kind, pat |
|
329 | return kind, pat | |
334 | return default, pattern |
|
330 | return default, pattern | |
335 |
|
331 | |||
336 | def _globre(pat): |
|
332 | def _globre(pat): | |
337 | r'''Convert an extended glob string to a regexp string. |
|
333 | r'''Convert an extended glob string to a regexp string. | |
338 |
|
334 | |||
339 | >>> print _globre(r'?') |
|
335 | >>> print _globre(r'?') | |
340 | . |
|
336 | . | |
341 | >>> print _globre(r'*') |
|
337 | >>> print _globre(r'*') | |
342 | [^/]* |
|
338 | [^/]* | |
343 | >>> print _globre(r'**') |
|
339 | >>> print _globre(r'**') | |
344 | .* |
|
340 | .* | |
345 | >>> print _globre(r'**/a') |
|
341 | >>> print _globre(r'**/a') | |
346 | (?:.*/)?a |
|
342 | (?:.*/)?a | |
347 | >>> print _globre(r'a/**/b') |
|
343 | >>> print _globre(r'a/**/b') | |
348 | a\/(?:.*/)?b |
|
344 | a\/(?:.*/)?b | |
349 | >>> print _globre(r'[a*?!^][^b][!c]') |
|
345 | >>> print _globre(r'[a*?!^][^b][!c]') | |
350 | [a*?!^][\^b][^c] |
|
346 | [a*?!^][\^b][^c] | |
351 | >>> print _globre(r'{a,b}') |
|
347 | >>> print _globre(r'{a,b}') | |
352 | (?:a|b) |
|
348 | (?:a|b) | |
353 | >>> print _globre(r'.\*\?') |
|
349 | >>> print _globre(r'.\*\?') | |
354 | \.\*\? |
|
350 | \.\*\? | |
355 | ''' |
|
351 | ''' | |
356 | i, n = 0, len(pat) |
|
352 | i, n = 0, len(pat) | |
357 | res = '' |
|
353 | res = '' | |
358 | group = 0 |
|
354 | group = 0 | |
359 | escape = util.re.escape |
|
355 | escape = util.re.escape | |
360 | def peek(): |
|
356 | def peek(): | |
361 | return i < n and pat[i] |
|
357 | return i < n and pat[i] | |
362 | while i < n: |
|
358 | while i < n: | |
363 | c = pat[i] |
|
359 | c = pat[i] | |
364 | i += 1 |
|
360 | i += 1 | |
365 | if c not in '*?[{},\\': |
|
361 | if c not in '*?[{},\\': | |
366 | res += escape(c) |
|
362 | res += escape(c) | |
367 | elif c == '*': |
|
363 | elif c == '*': | |
368 | if peek() == '*': |
|
364 | if peek() == '*': | |
369 | i += 1 |
|
365 | i += 1 | |
370 | if peek() == '/': |
|
366 | if peek() == '/': | |
371 | i += 1 |
|
367 | i += 1 | |
372 | res += '(?:.*/)?' |
|
368 | res += '(?:.*/)?' | |
373 | else: |
|
369 | else: | |
374 | res += '.*' |
|
370 | res += '.*' | |
375 | else: |
|
371 | else: | |
376 | res += '[^/]*' |
|
372 | res += '[^/]*' | |
377 | elif c == '?': |
|
373 | elif c == '?': | |
378 | res += '.' |
|
374 | res += '.' | |
379 | elif c == '[': |
|
375 | elif c == '[': | |
380 | j = i |
|
376 | j = i | |
381 | if j < n and pat[j] in '!]': |
|
377 | if j < n and pat[j] in '!]': | |
382 | j += 1 |
|
378 | j += 1 | |
383 | while j < n and pat[j] != ']': |
|
379 | while j < n and pat[j] != ']': | |
384 | j += 1 |
|
380 | j += 1 | |
385 | if j >= n: |
|
381 | if j >= n: | |
386 | res += '\\[' |
|
382 | res += '\\[' | |
387 | else: |
|
383 | else: | |
388 | stuff = pat[i:j].replace('\\','\\\\') |
|
384 | stuff = pat[i:j].replace('\\','\\\\') | |
389 | i = j + 1 |
|
385 | i = j + 1 | |
390 | if stuff[0] == '!': |
|
386 | if stuff[0] == '!': | |
391 | stuff = '^' + stuff[1:] |
|
387 | stuff = '^' + stuff[1:] | |
392 | elif stuff[0] == '^': |
|
388 | elif stuff[0] == '^': | |
393 | stuff = '\\' + stuff |
|
389 | stuff = '\\' + stuff | |
394 | res = '%s[%s]' % (res, stuff) |
|
390 | res = '%s[%s]' % (res, stuff) | |
395 | elif c == '{': |
|
391 | elif c == '{': | |
396 | group += 1 |
|
392 | group += 1 | |
397 | res += '(?:' |
|
393 | res += '(?:' | |
398 | elif c == '}' and group: |
|
394 | elif c == '}' and group: | |
399 | res += ')' |
|
395 | res += ')' | |
400 | group -= 1 |
|
396 | group -= 1 | |
401 | elif c == ',' and group: |
|
397 | elif c == ',' and group: | |
402 | res += '|' |
|
398 | res += '|' | |
403 | elif c == '\\': |
|
399 | elif c == '\\': | |
404 | p = peek() |
|
400 | p = peek() | |
405 | if p: |
|
401 | if p: | |
406 | i += 1 |
|
402 | i += 1 | |
407 | res += escape(p) |
|
403 | res += escape(p) | |
408 | else: |
|
404 | else: | |
409 | res += escape(c) |
|
405 | res += escape(c) | |
410 | else: |
|
406 | else: | |
411 | res += escape(c) |
|
407 | res += escape(c) | |
412 | return res |
|
408 | return res | |
413 |
|
409 | |||
414 | def _regex(kind, pat, globsuffix): |
|
410 | def _regex(kind, pat, globsuffix): | |
415 | '''Convert a (normalized) pattern of any kind into a regular expression. |
|
411 | '''Convert a (normalized) pattern of any kind into a regular expression. | |
416 | globsuffix is appended to the regexp of globs.''' |
|
412 | globsuffix is appended to the regexp of globs.''' | |
417 | if not pat: |
|
413 | if not pat: | |
418 | return '' |
|
414 | return '' | |
419 | if kind == 're': |
|
415 | if kind == 're': | |
420 | return pat |
|
416 | return pat | |
421 | if kind == 'path': |
|
417 | if kind == 'path': | |
422 | return '^' + util.re.escape(pat) + '(?:/|$)' |
|
418 | return '^' + util.re.escape(pat) + '(?:/|$)' | |
423 | if kind == 'relglob': |
|
419 | if kind == 'relglob': | |
424 | return '(?:|.*/)' + _globre(pat) + globsuffix |
|
420 | return '(?:|.*/)' + _globre(pat) + globsuffix | |
425 | if kind == 'relpath': |
|
421 | if kind == 'relpath': | |
426 | return util.re.escape(pat) + '(?:/|$)' |
|
422 | return util.re.escape(pat) + '(?:/|$)' | |
427 | if kind == 'relre': |
|
423 | if kind == 'relre': | |
428 | if pat.startswith('^'): |
|
424 | if pat.startswith('^'): | |
429 | return pat |
|
425 | return pat | |
430 | return '.*' + pat |
|
426 | return '.*' + pat | |
431 | return _globre(pat) + globsuffix |
|
427 | return _globre(pat) + globsuffix | |
432 |
|
428 | |||
433 | def _buildmatch(ctx, kindpats, globsuffix, listsubrepos): |
|
429 | def _buildmatch(ctx, kindpats, globsuffix, listsubrepos): | |
434 | '''Return regexp string and a matcher function for kindpats. |
|
430 | '''Return regexp string and a matcher function for kindpats. | |
435 | globsuffix is appended to the regexp of globs.''' |
|
431 | globsuffix is appended to the regexp of globs.''' | |
436 | fset, kindpats = _expandsets(kindpats, ctx, listsubrepos) |
|
432 | fset, kindpats = _expandsets(kindpats, ctx, listsubrepos) | |
437 | if not kindpats: |
|
433 | if not kindpats: | |
438 | return "", fset.__contains__ |
|
434 | return "", fset.__contains__ | |
439 |
|
435 | |||
440 | regex, mf = _buildregexmatch(kindpats, globsuffix) |
|
436 | regex, mf = _buildregexmatch(kindpats, globsuffix) | |
441 | if fset: |
|
437 | if fset: | |
442 | return regex, lambda f: f in fset or mf(f) |
|
438 | return regex, lambda f: f in fset or mf(f) | |
443 | return regex, mf |
|
439 | return regex, mf | |
444 |
|
440 | |||
445 | def _buildregexmatch(kindpats, globsuffix): |
|
441 | def _buildregexmatch(kindpats, globsuffix): | |
446 | """Build a match function from a list of kinds and kindpats, |
|
442 | """Build a match function from a list of kinds and kindpats, | |
447 | return regexp string and a matcher function.""" |
|
443 | return regexp string and a matcher function.""" | |
448 | try: |
|
444 | try: | |
449 | regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix) |
|
445 | regex = '(?:%s)' % '|'.join([_regex(k, p, globsuffix) | |
450 | for (k, p) in kindpats]) |
|
446 | for (k, p) in kindpats]) | |
451 | if len(regex) > 20000: |
|
447 | if len(regex) > 20000: | |
452 | raise OverflowError |
|
448 | raise OverflowError | |
453 | return regex, _rematcher(regex) |
|
449 | return regex, _rematcher(regex) | |
454 | except OverflowError: |
|
450 | except OverflowError: | |
455 | # We're using a Python with a tiny regex engine and we |
|
451 | # We're using a Python with a tiny regex engine and we | |
456 | # made it explode, so we'll divide the pattern list in two |
|
452 | # made it explode, so we'll divide the pattern list in two | |
457 | # until it works |
|
453 | # until it works | |
458 | l = len(kindpats) |
|
454 | l = len(kindpats) | |
459 | if l < 2: |
|
455 | if l < 2: | |
460 | raise |
|
456 | raise | |
461 | regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix) |
|
457 | regexa, a = _buildregexmatch(kindpats[:l//2], globsuffix) | |
462 | regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix) |
|
458 | regexb, b = _buildregexmatch(kindpats[l//2:], globsuffix) | |
463 | return regex, lambda s: a(s) or b(s) |
|
459 | return regex, lambda s: a(s) or b(s) | |
464 | except re.error: |
|
460 | except re.error: | |
465 | for k, p in kindpats: |
|
461 | for k, p in kindpats: | |
466 | try: |
|
462 | try: | |
467 | _rematcher('(?:%s)' % _regex(k, p, globsuffix)) |
|
463 | _rematcher('(?:%s)' % _regex(k, p, globsuffix)) | |
468 | except re.error: |
|
464 | except re.error: | |
469 | raise util.Abort(_("invalid pattern (%s): %s") % (k, p)) |
|
465 | raise util.Abort(_("invalid pattern (%s): %s") % (k, p)) | |
470 | raise util.Abort(_("invalid pattern")) |
|
466 | raise util.Abort(_("invalid pattern")) | |
471 |
|
467 | |||
472 | def _roots(kindpats): |
|
468 | def _roots(kindpats): | |
473 | '''return roots and exact explicitly listed files from patterns |
|
469 | '''return roots and exact explicitly listed files from patterns | |
474 |
|
470 | |||
475 | >>> _roots([('glob', 'g/*'), ('glob', 'g'), ('glob', 'g*')]) |
|
471 | >>> _roots([('glob', 'g/*'), ('glob', 'g'), ('glob', 'g*')]) | |
476 | ['g', 'g', '.'] |
|
472 | ['g', 'g', '.'] | |
477 | >>> _roots([('relpath', 'r'), ('path', 'p/p'), ('path', '')]) |
|
473 | >>> _roots([('relpath', 'r'), ('path', 'p/p'), ('path', '')]) | |
478 | ['r', 'p/p', '.'] |
|
474 | ['r', 'p/p', '.'] | |
479 | >>> _roots([('relglob', 'rg*'), ('re', 're/'), ('relre', 'rr')]) |
|
475 | >>> _roots([('relglob', 'rg*'), ('re', 're/'), ('relre', 'rr')]) | |
480 | ['.', '.', '.'] |
|
476 | ['.', '.', '.'] | |
481 | ''' |
|
477 | ''' | |
482 | r = [] |
|
478 | r = [] | |
483 | for kind, pat in kindpats: |
|
479 | for kind, pat in kindpats: | |
484 | if kind == 'glob': # find the non-glob prefix |
|
480 | if kind == 'glob': # find the non-glob prefix | |
485 | root = [] |
|
481 | root = [] | |
486 | for p in pat.split('/'): |
|
482 | for p in pat.split('/'): | |
487 | if '[' in p or '{' in p or '*' in p or '?' in p: |
|
483 | if '[' in p or '{' in p or '*' in p or '?' in p: | |
488 | break |
|
484 | break | |
489 | root.append(p) |
|
485 | root.append(p) | |
490 | r.append('/'.join(root) or '.') |
|
486 | r.append('/'.join(root) or '.') | |
491 | elif kind in ('relpath', 'path'): |
|
487 | elif kind in ('relpath', 'path'): | |
492 | r.append(pat or '.') |
|
488 | r.append(pat or '.') | |
493 | else: # relglob, re, relre |
|
489 | else: # relglob, re, relre | |
494 | r.append('.') |
|
490 | r.append('.') | |
495 | return r |
|
491 | return r | |
496 |
|
492 | |||
497 | def _anypats(kindpats): |
|
493 | def _anypats(kindpats): | |
498 | for kind, pat in kindpats: |
|
494 | for kind, pat in kindpats: | |
499 | if kind in ('glob', 're', 'relglob', 'relre', 'set'): |
|
495 | if kind in ('glob', 're', 'relglob', 'relre', 'set'): | |
500 | return True |
|
496 | return True | |
501 |
|
497 | |||
502 | _commentre = None |
|
498 | _commentre = None | |
503 |
|
499 | |||
504 | def readpatternfile(filepath, warn): |
|
500 | def readpatternfile(filepath, warn): | |
505 | '''parse a pattern file, returning a list of |
|
501 | '''parse a pattern file, returning a list of | |
506 | patterns. These patterns should be given to compile() |
|
502 | patterns. These patterns should be given to compile() | |
507 | to be validated and converted into a match function.''' |
|
503 | to be validated and converted into a match function.''' | |
508 | syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:'} |
|
504 | syntaxes = {'re': 'relre:', 'regexp': 'relre:', 'glob': 'relglob:'} | |
509 | syntax = 'relre:' |
|
505 | syntax = 'relre:' | |
510 | patterns = [] |
|
506 | patterns = [] | |
511 |
|
507 | |||
512 | fp = open(filepath) |
|
508 | fp = open(filepath) | |
513 | for line in fp: |
|
509 | for line in fp: | |
514 | if "#" in line: |
|
510 | if "#" in line: | |
515 | global _commentre |
|
511 | global _commentre | |
516 | if not _commentre: |
|
512 | if not _commentre: | |
517 | _commentre = re.compile(r'((^|[^\\])(\\\\)*)#.*') |
|
513 | _commentre = re.compile(r'((^|[^\\])(\\\\)*)#.*') | |
518 | # remove comments prefixed by an even number of escapes |
|
514 | # remove comments prefixed by an even number of escapes | |
519 | line = _commentre.sub(r'\1', line) |
|
515 | line = _commentre.sub(r'\1', line) | |
520 | # fixup properly escaped comments that survived the above |
|
516 | # fixup properly escaped comments that survived the above | |
521 | line = line.replace("\\#", "#") |
|
517 | line = line.replace("\\#", "#") | |
522 | line = line.rstrip() |
|
518 | line = line.rstrip() | |
523 | if not line: |
|
519 | if not line: | |
524 | continue |
|
520 | continue | |
525 |
|
521 | |||
526 | if line.startswith('syntax:'): |
|
522 | if line.startswith('syntax:'): | |
527 | s = line[7:].strip() |
|
523 | s = line[7:].strip() | |
528 | try: |
|
524 | try: | |
529 | syntax = syntaxes[s] |
|
525 | syntax = syntaxes[s] | |
530 | except KeyError: |
|
526 | except KeyError: | |
531 | warn(_("%s: ignoring invalid syntax '%s'\n") % (filepath, s)) |
|
527 | warn(_("%s: ignoring invalid syntax '%s'\n") % (filepath, s)) | |
532 | continue |
|
528 | continue | |
533 |
|
529 | |||
534 | linesyntax = syntax |
|
530 | linesyntax = syntax | |
535 | for s, rels in syntaxes.iteritems(): |
|
531 | for s, rels in syntaxes.iteritems(): | |
536 | if line.startswith(rels): |
|
532 | if line.startswith(rels): | |
537 | linesyntax = rels |
|
533 | linesyntax = rels | |
538 | line = line[len(rels):] |
|
534 | line = line[len(rels):] | |
539 | break |
|
535 | break | |
540 | elif line.startswith(s+':'): |
|
536 | elif line.startswith(s+':'): | |
541 | linesyntax = rels |
|
537 | linesyntax = rels | |
542 | line = line[len(s) + 1:] |
|
538 | line = line[len(s) + 1:] | |
543 | break |
|
539 | break | |
544 | patterns.append(linesyntax + line) |
|
540 | patterns.append(linesyntax + line) | |
545 | fp.close() |
|
541 | fp.close() | |
546 | return patterns |
|
542 | return patterns |
General Comments 0
You need to be logged in to leave comments.
Login now