Show More
@@ -1,848 +1,854 b'' | |||
|
1 | 1 | # Mercurial built-in replacement for cvsps. |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2008, Frank Kingswood <frank@kingswood-consulting.co.uk> |
|
4 | 4 | # |
|
5 | 5 | # This software may be used and distributed according to the terms of the |
|
6 | 6 | # GNU General Public License version 2 or any later version. |
|
7 | 7 | |
|
8 | 8 | import os |
|
9 | 9 | import re |
|
10 | 10 | import cPickle as pickle |
|
11 | 11 | from mercurial import util |
|
12 | 12 | from mercurial.i18n import _ |
|
13 | 13 | from mercurial import hook |
|
14 | 14 | from mercurial import util |
|
15 | 15 | |
|
16 | 16 | class logentry(object): |
|
17 | 17 | '''Class logentry has the following attributes: |
|
18 | 18 | .author - author name as CVS knows it |
|
19 | 19 | .branch - name of branch this revision is on |
|
20 | 20 | .branches - revision tuple of branches starting at this revision |
|
21 | 21 | .comment - commit message |
|
22 | 22 | .date - the commit date as a (time, tz) tuple |
|
23 | 23 | .dead - true if file revision is dead |
|
24 | 24 | .file - Name of file |
|
25 | 25 | .lines - a tuple (+lines, -lines) or None |
|
26 | 26 | .parent - Previous revision of this entry |
|
27 | 27 | .rcs - name of file as returned from CVS |
|
28 | 28 | .revision - revision number as tuple |
|
29 | 29 | .tags - list of tags on the file |
|
30 | 30 | .synthetic - is this a synthetic "file ... added on ..." revision? |
|
31 | 31 | .mergepoint- the branch that has been merged from |
|
32 | 32 | (if present in rlog output) |
|
33 | 33 | .branchpoints- the branches that start at the current entry |
|
34 | 34 | ''' |
|
35 | 35 | def __init__(self, **entries): |
|
36 | 36 | self.synthetic = False |
|
37 | 37 | self.__dict__.update(entries) |
|
38 | 38 | |
|
39 | 39 | def __repr__(self): |
|
40 | 40 | return "<%s at 0x%x: %s %s>" % (self.__class__.__name__, |
|
41 | 41 | id(self), |
|
42 | 42 | self.file, |
|
43 | 43 | ".".join(map(str, self.revision))) |
|
44 | 44 | |
|
45 | 45 | class logerror(Exception): |
|
46 | 46 | pass |
|
47 | 47 | |
|
48 | 48 | def getrepopath(cvspath): |
|
49 | 49 | """Return the repository path from a CVS path. |
|
50 | 50 | |
|
51 | 51 | >>> getrepopath('/foo/bar') |
|
52 | 52 | '/foo/bar' |
|
53 | 53 | >>> getrepopath('c:/foo/bar') |
|
54 | 54 | 'c:/foo/bar' |
|
55 | 55 | >>> getrepopath(':pserver:10/foo/bar') |
|
56 | 56 | '/foo/bar' |
|
57 | 57 | >>> getrepopath(':pserver:10c:/foo/bar') |
|
58 | 58 | '/foo/bar' |
|
59 | 59 | >>> getrepopath(':pserver:/foo/bar') |
|
60 | 60 | '/foo/bar' |
|
61 | 61 | >>> getrepopath(':pserver:c:/foo/bar') |
|
62 | 62 | 'c:/foo/bar' |
|
63 | 63 | >>> getrepopath(':pserver:truc@foo.bar:/foo/bar') |
|
64 | 64 | '/foo/bar' |
|
65 | 65 | >>> getrepopath(':pserver:truc@foo.bar:c:/foo/bar') |
|
66 | 66 | 'c:/foo/bar' |
|
67 | 67 | """ |
|
68 | 68 | # According to CVS manual, CVS paths are expressed like: |
|
69 | 69 | # [:method:][[user][:password]@]hostname[:[port]]/path/to/repository |
|
70 | 70 | # |
|
71 | 71 | # Unfortunately, Windows absolute paths start with a drive letter |
|
72 | 72 | # like 'c:' making it harder to parse. Here we assume that drive |
|
73 | 73 | # letters are only one character long and any CVS component before |
|
74 | 74 | # the repository path is at least 2 characters long, and use this |
|
75 | 75 | # to disambiguate. |
|
76 | 76 | parts = cvspath.split(':') |
|
77 | 77 | if len(parts) == 1: |
|
78 | 78 | return parts[0] |
|
79 | 79 | # Here there is an ambiguous case if we have a port number |
|
80 | 80 | # immediately followed by a Windows driver letter. We assume this |
|
81 | 81 | # never happens and decide it must be CVS path component, |
|
82 | 82 | # therefore ignoring it. |
|
83 | 83 | if len(parts[-2]) > 1: |
|
84 | 84 | return parts[-1].lstrip('0123456789') |
|
85 | 85 | return parts[-2] + ':' + parts[-1] |
|
86 | 86 | |
|
87 | 87 | def createlog(ui, directory=None, root="", rlog=True, cache=None): |
|
88 | 88 | '''Collect the CVS rlog''' |
|
89 | 89 | |
|
90 | 90 | # Because we store many duplicate commit log messages, reusing strings |
|
91 | 91 | # saves a lot of memory and pickle storage space. |
|
92 | 92 | _scache = {} |
|
93 | 93 | def scache(s): |
|
94 | 94 | "return a shared version of a string" |
|
95 | 95 | return _scache.setdefault(s, s) |
|
96 | 96 | |
|
97 | 97 | ui.status(_('collecting CVS rlog\n')) |
|
98 | 98 | |
|
99 | 99 | log = [] # list of logentry objects containing the CVS state |
|
100 | 100 | |
|
101 | 101 | # patterns to match in CVS (r)log output, by state of use |
|
102 | 102 | re_00 = re.compile('RCS file: (.+)$') |
|
103 | 103 | re_01 = re.compile('cvs \\[r?log aborted\\]: (.+)$') |
|
104 | 104 | re_02 = re.compile('cvs (r?log|server): (.+)\n$') |
|
105 | 105 | re_03 = re.compile("(Cannot access.+CVSROOT)|" |
|
106 | 106 | "(can't create temporary directory.+)$") |
|
107 | 107 | re_10 = re.compile('Working file: (.+)$') |
|
108 | 108 | re_20 = re.compile('symbolic names:') |
|
109 | 109 | re_30 = re.compile('\t(.+): ([\\d.]+)$') |
|
110 | 110 | re_31 = re.compile('----------------------------$') |
|
111 | 111 | re_32 = re.compile('=======================================' |
|
112 | 112 | '======================================$') |
|
113 | 113 | re_50 = re.compile('revision ([\\d.]+)(\s+locked by:\s+.+;)?$') |
|
114 | 114 | re_60 = re.compile(r'date:\s+(.+);\s+author:\s+(.+);\s+state:\s+(.+?);' |
|
115 | 115 | r'(\s+lines:\s+(\+\d+)?\s+(-\d+)?;)?' |
|
116 | 116 | r'(.*mergepoint:\s+([^;]+);)?') |
|
117 | 117 | re_70 = re.compile('branches: (.+);$') |
|
118 | 118 | |
|
119 | 119 | file_added_re = re.compile(r'file [^/]+ was (initially )?added on branch') |
|
120 | 120 | |
|
121 | 121 | prefix = '' # leading path to strip of what we get from CVS |
|
122 | 122 | |
|
123 | 123 | if directory is None: |
|
124 | 124 | # Current working directory |
|
125 | 125 | |
|
126 | 126 | # Get the real directory in the repository |
|
127 | 127 | try: |
|
128 | 128 | prefix = open(os.path.join('CVS','Repository')).read().strip() |
|
129 | 129 | directory = prefix |
|
130 | 130 | if prefix == ".": |
|
131 | 131 | prefix = "" |
|
132 | 132 | except IOError: |
|
133 | 133 | raise logerror(_('not a CVS sandbox')) |
|
134 | 134 | |
|
135 | 135 | if prefix and not prefix.endswith(os.sep): |
|
136 | 136 | prefix += os.sep |
|
137 | 137 | |
|
138 | 138 | # Use the Root file in the sandbox, if it exists |
|
139 | 139 | try: |
|
140 | 140 | root = open(os.path.join('CVS','Root')).read().strip() |
|
141 | 141 | except IOError: |
|
142 | 142 | pass |
|
143 | 143 | |
|
144 | 144 | if not root: |
|
145 | 145 | root = os.environ.get('CVSROOT', '') |
|
146 | 146 | |
|
147 | 147 | # read log cache if one exists |
|
148 | 148 | oldlog = [] |
|
149 | 149 | date = None |
|
150 | 150 | |
|
151 | 151 | if cache: |
|
152 | 152 | cachedir = os.path.expanduser('~/.hg.cvsps') |
|
153 | 153 | if not os.path.exists(cachedir): |
|
154 | 154 | os.mkdir(cachedir) |
|
155 | 155 | |
|
156 | 156 | # The cvsps cache pickle needs a uniquified name, based on the |
|
157 | 157 | # repository location. The address may have all sort of nasties |
|
158 | 158 | # in it, slashes, colons and such. So here we take just the |
|
159 | 159 | # alphanumerics, concatenated in a way that does not mix up the |
|
160 | 160 | # various components, so that |
|
161 | 161 | # :pserver:user@server:/path |
|
162 | 162 | # and |
|
163 | 163 | # /pserver/user/server/path |
|
164 | 164 | # are mapped to different cache file names. |
|
165 | 165 | cachefile = root.split(":") + [directory, "cache"] |
|
166 | 166 | cachefile = ['-'.join(re.findall(r'\w+', s)) for s in cachefile if s] |
|
167 | 167 | cachefile = os.path.join(cachedir, |
|
168 | 168 | '.'.join([s for s in cachefile if s])) |
|
169 | 169 | |
|
170 | 170 | if cache == 'update': |
|
171 | 171 | try: |
|
172 | 172 | ui.note(_('reading cvs log cache %s\n') % cachefile) |
|
173 | 173 | oldlog = pickle.load(open(cachefile)) |
|
174 | 174 | ui.note(_('cache has %d log entries\n') % len(oldlog)) |
|
175 | 175 | except Exception, e: |
|
176 | 176 | ui.note(_('error reading cache: %r\n') % e) |
|
177 | 177 | |
|
178 | 178 | if oldlog: |
|
179 | 179 | date = oldlog[-1].date # last commit date as a (time,tz) tuple |
|
180 | 180 | date = util.datestr(date, '%Y/%m/%d %H:%M:%S %1%2') |
|
181 | 181 | |
|
182 | 182 | # build the CVS commandline |
|
183 | 183 | cmd = ['cvs', '-q'] |
|
184 | 184 | if root: |
|
185 | 185 | cmd.append('-d%s' % root) |
|
186 | 186 | p = util.normpath(getrepopath(root)) |
|
187 | 187 | if not p.endswith('/'): |
|
188 | 188 | p += '/' |
|
189 | 189 | if prefix: |
|
190 | 190 | # looks like normpath replaces "" by "." |
|
191 | 191 | prefix = p + util.normpath(prefix) |
|
192 | 192 | else: |
|
193 | 193 | prefix = p |
|
194 | 194 | cmd.append(['log', 'rlog'][rlog]) |
|
195 | 195 | if date: |
|
196 | 196 | # no space between option and date string |
|
197 | 197 | cmd.append('-d>%s' % date) |
|
198 | 198 | cmd.append(directory) |
|
199 | 199 | |
|
200 | 200 | # state machine begins here |
|
201 | 201 | tags = {} # dictionary of revisions on current file with their tags |
|
202 | 202 | branchmap = {} # mapping between branch names and revision numbers |
|
203 | 203 | state = 0 |
|
204 | 204 | store = False # set when a new record can be appended |
|
205 | 205 | |
|
206 | 206 | cmd = [util.shellquote(arg) for arg in cmd] |
|
207 | 207 | ui.note(_("running %s\n") % (' '.join(cmd))) |
|
208 | 208 | ui.debug("prefix=%r directory=%r root=%r\n" % (prefix, directory, root)) |
|
209 | 209 | |
|
210 | 210 | pfp = util.popen(' '.join(cmd)) |
|
211 | 211 | peek = pfp.readline() |
|
212 | 212 | while True: |
|
213 | 213 | line = peek |
|
214 | 214 | if line == '': |
|
215 | 215 | break |
|
216 | 216 | peek = pfp.readline() |
|
217 | 217 | if line.endswith('\n'): |
|
218 | 218 | line = line[:-1] |
|
219 | 219 | #ui.debug('state=%d line=%r\n' % (state, line)) |
|
220 | 220 | |
|
221 | 221 | if state == 0: |
|
222 | 222 | # initial state, consume input until we see 'RCS file' |
|
223 | 223 | match = re_00.match(line) |
|
224 | 224 | if match: |
|
225 | 225 | rcs = match.group(1) |
|
226 | 226 | tags = {} |
|
227 | 227 | if rlog: |
|
228 | 228 | filename = util.normpath(rcs[:-2]) |
|
229 | 229 | if filename.startswith(prefix): |
|
230 | 230 | filename = filename[len(prefix):] |
|
231 | 231 | if filename.startswith('/'): |
|
232 | 232 | filename = filename[1:] |
|
233 | 233 | if filename.startswith('Attic/'): |
|
234 | 234 | filename = filename[6:] |
|
235 | 235 | else: |
|
236 | 236 | filename = filename.replace('/Attic/', '/') |
|
237 | 237 | state = 2 |
|
238 | 238 | continue |
|
239 | 239 | state = 1 |
|
240 | 240 | continue |
|
241 | 241 | match = re_01.match(line) |
|
242 | 242 | if match: |
|
243 | 243 | raise logerror(match.group(1)) |
|
244 | 244 | match = re_02.match(line) |
|
245 | 245 | if match: |
|
246 | 246 | raise logerror(match.group(2)) |
|
247 | 247 | if re_03.match(line): |
|
248 | 248 | raise logerror(line) |
|
249 | 249 | |
|
250 | 250 | elif state == 1: |
|
251 | 251 | # expect 'Working file' (only when using log instead of rlog) |
|
252 | 252 | match = re_10.match(line) |
|
253 | 253 | assert match, _('RCS file must be followed by working file') |
|
254 | 254 | filename = util.normpath(match.group(1)) |
|
255 | 255 | state = 2 |
|
256 | 256 | |
|
257 | 257 | elif state == 2: |
|
258 | 258 | # expect 'symbolic names' |
|
259 | 259 | if re_20.match(line): |
|
260 | 260 | branchmap = {} |
|
261 | 261 | state = 3 |
|
262 | 262 | |
|
263 | 263 | elif state == 3: |
|
264 | 264 | # read the symbolic names and store as tags |
|
265 | 265 | match = re_30.match(line) |
|
266 | 266 | if match: |
|
267 | 267 | rev = [int(x) for x in match.group(2).split('.')] |
|
268 | 268 | |
|
269 | 269 | # Convert magic branch number to an odd-numbered one |
|
270 | 270 | revn = len(rev) |
|
271 | 271 | if revn > 3 and (revn % 2) == 0 and rev[-2] == 0: |
|
272 | 272 | rev = rev[:-2] + rev[-1:] |
|
273 | 273 | rev = tuple(rev) |
|
274 | 274 | |
|
275 | 275 | if rev not in tags: |
|
276 | 276 | tags[rev] = [] |
|
277 | 277 | tags[rev].append(match.group(1)) |
|
278 | 278 | branchmap[match.group(1)] = match.group(2) |
|
279 | 279 | |
|
280 | 280 | elif re_31.match(line): |
|
281 | 281 | state = 5 |
|
282 | 282 | elif re_32.match(line): |
|
283 | 283 | state = 0 |
|
284 | 284 | |
|
285 | 285 | elif state == 4: |
|
286 | 286 | # expecting '------' separator before first revision |
|
287 | 287 | if re_31.match(line): |
|
288 | 288 | state = 5 |
|
289 | 289 | else: |
|
290 | 290 | assert not re_32.match(line), _('must have at least ' |
|
291 | 291 | 'some revisions') |
|
292 | 292 | |
|
293 | 293 | elif state == 5: |
|
294 | 294 | # expecting revision number and possibly (ignored) lock indication |
|
295 | 295 | # we create the logentry here from values stored in states 0 to 4, |
|
296 | 296 | # as this state is re-entered for subsequent revisions of a file. |
|
297 | 297 | match = re_50.match(line) |
|
298 | 298 | assert match, _('expected revision number') |
|
299 | 299 | e = logentry(rcs=scache(rcs), file=scache(filename), |
|
300 | 300 | revision=tuple([int(x) for x in match.group(1).split('.')]), |
|
301 | 301 | branches=[], parent=None) |
|
302 | 302 | state = 6 |
|
303 | 303 | |
|
304 | 304 | elif state == 6: |
|
305 | 305 | # expecting date, author, state, lines changed |
|
306 | 306 | match = re_60.match(line) |
|
307 | 307 | assert match, _('revision must be followed by date line') |
|
308 | 308 | d = match.group(1) |
|
309 | 309 | if d[2] == '/': |
|
310 | 310 | # Y2K |
|
311 | 311 | d = '19' + d |
|
312 | 312 | |
|
313 | 313 | if len(d.split()) != 3: |
|
314 | 314 | # cvs log dates always in GMT |
|
315 | 315 | d = d + ' UTC' |
|
316 | 316 | e.date = util.parsedate(d, ['%y/%m/%d %H:%M:%S', |
|
317 | 317 | '%Y/%m/%d %H:%M:%S', |
|
318 | 318 | '%Y-%m-%d %H:%M:%S']) |
|
319 | 319 | e.author = scache(match.group(2)) |
|
320 | 320 | e.dead = match.group(3).lower() == 'dead' |
|
321 | 321 | |
|
322 | 322 | if match.group(5): |
|
323 | 323 | if match.group(6): |
|
324 | 324 | e.lines = (int(match.group(5)), int(match.group(6))) |
|
325 | 325 | else: |
|
326 | 326 | e.lines = (int(match.group(5)), 0) |
|
327 | 327 | elif match.group(6): |
|
328 | 328 | e.lines = (0, int(match.group(6))) |
|
329 | 329 | else: |
|
330 | 330 | e.lines = None |
|
331 | 331 | |
|
332 | 332 | if match.group(7): # cvsnt mergepoint |
|
333 | 333 | myrev = match.group(8).split('.') |
|
334 | 334 | if len(myrev) == 2: # head |
|
335 | 335 | e.mergepoint = 'HEAD' |
|
336 | 336 | else: |
|
337 | 337 | myrev = '.'.join(myrev[:-2] + ['0', myrev[-2]]) |
|
338 | 338 | branches = [b for b in branchmap if branchmap[b] == myrev] |
|
339 | 339 | assert len(branches) == 1, 'unknown branch: %s' % e.mergepoint |
|
340 | 340 | e.mergepoint = branches[0] |
|
341 | 341 | else: |
|
342 | 342 | e.mergepoint = None |
|
343 | 343 | e.comment = [] |
|
344 | 344 | state = 7 |
|
345 | 345 | |
|
346 | 346 | elif state == 7: |
|
347 | 347 | # read the revision numbers of branches that start at this revision |
|
348 | 348 | # or store the commit log message otherwise |
|
349 | 349 | m = re_70.match(line) |
|
350 | 350 | if m: |
|
351 | 351 | e.branches = [tuple([int(y) for y in x.strip().split('.')]) |
|
352 | 352 | for x in m.group(1).split(';')] |
|
353 | 353 | state = 8 |
|
354 | 354 | elif re_31.match(line) and re_50.match(peek): |
|
355 | 355 | state = 5 |
|
356 | 356 | store = True |
|
357 | 357 | elif re_32.match(line): |
|
358 | 358 | state = 0 |
|
359 | 359 | store = True |
|
360 | 360 | else: |
|
361 | 361 | e.comment.append(line) |
|
362 | 362 | |
|
363 | 363 | elif state == 8: |
|
364 | 364 | # store commit log message |
|
365 | 365 | if re_31.match(line): |
|
366 | cpeek = peek | |
|
367 | if cpeek.endswith('\n'): | |
|
368 | cpeek = cpeek[:-1] | |
|
369 | if re_50.match(cpeek): | |
|
366 | 370 | state = 5 |
|
367 | 371 | store = True |
|
372 | else: | |
|
373 | e.comment.append(line) | |
|
368 | 374 | elif re_32.match(line): |
|
369 | 375 | state = 0 |
|
370 | 376 | store = True |
|
371 | 377 | else: |
|
372 | 378 | e.comment.append(line) |
|
373 | 379 | |
|
374 | 380 | # When a file is added on a branch B1, CVS creates a synthetic |
|
375 | 381 | # dead trunk revision 1.1 so that the branch has a root. |
|
376 | 382 | # Likewise, if you merge such a file to a later branch B2 (one |
|
377 | 383 | # that already existed when the file was added on B1), CVS |
|
378 | 384 | # creates a synthetic dead revision 1.1.x.1 on B2. Don't drop |
|
379 | 385 | # these revisions now, but mark them synthetic so |
|
380 | 386 | # createchangeset() can take care of them. |
|
381 | 387 | if (store and |
|
382 | 388 | e.dead and |
|
383 | 389 | e.revision[-1] == 1 and # 1.1 or 1.1.x.1 |
|
384 | 390 | len(e.comment) == 1 and |
|
385 | 391 | file_added_re.match(e.comment[0])): |
|
386 | 392 | ui.debug('found synthetic revision in %s: %r\n' |
|
387 | 393 | % (e.rcs, e.comment[0])) |
|
388 | 394 | e.synthetic = True |
|
389 | 395 | |
|
390 | 396 | if store: |
|
391 | 397 | # clean up the results and save in the log. |
|
392 | 398 | store = False |
|
393 | 399 | e.tags = sorted([scache(x) for x in tags.get(e.revision, [])]) |
|
394 | 400 | e.comment = scache('\n'.join(e.comment)) |
|
395 | 401 | |
|
396 | 402 | revn = len(e.revision) |
|
397 | 403 | if revn > 3 and (revn % 2) == 0: |
|
398 | 404 | e.branch = tags.get(e.revision[:-1], [None])[0] |
|
399 | 405 | else: |
|
400 | 406 | e.branch = None |
|
401 | 407 | |
|
402 | 408 | # find the branches starting from this revision |
|
403 | 409 | branchpoints = set() |
|
404 | 410 | for branch, revision in branchmap.iteritems(): |
|
405 | 411 | revparts = tuple([int(i) for i in revision.split('.')]) |
|
406 | 412 | if len(revparts) < 2: # bad tags |
|
407 | 413 | continue |
|
408 | 414 | if revparts[-2] == 0 and revparts[-1] % 2 == 0: |
|
409 | 415 | # normal branch |
|
410 | 416 | if revparts[:-2] == e.revision: |
|
411 | 417 | branchpoints.add(branch) |
|
412 | 418 | elif revparts == (1, 1, 1): # vendor branch |
|
413 | 419 | if revparts in e.branches: |
|
414 | 420 | branchpoints.add(branch) |
|
415 | 421 | e.branchpoints = branchpoints |
|
416 | 422 | |
|
417 | 423 | log.append(e) |
|
418 | 424 | |
|
419 | 425 | if len(log) % 100 == 0: |
|
420 | 426 | ui.status(util.ellipsis('%d %s' % (len(log), e.file), 80)+'\n') |
|
421 | 427 | |
|
422 | 428 | log.sort(key=lambda x: (x.rcs, x.revision)) |
|
423 | 429 | |
|
424 | 430 | # find parent revisions of individual files |
|
425 | 431 | versions = {} |
|
426 | 432 | for e in log: |
|
427 | 433 | branch = e.revision[:-1] |
|
428 | 434 | p = versions.get((e.rcs, branch), None) |
|
429 | 435 | if p is None: |
|
430 | 436 | p = e.revision[:-2] |
|
431 | 437 | e.parent = p |
|
432 | 438 | versions[(e.rcs, branch)] = e.revision |
|
433 | 439 | |
|
434 | 440 | # update the log cache |
|
435 | 441 | if cache: |
|
436 | 442 | if log: |
|
437 | 443 | # join up the old and new logs |
|
438 | 444 | log.sort(key=lambda x: x.date) |
|
439 | 445 | |
|
440 | 446 | if oldlog and oldlog[-1].date >= log[0].date: |
|
441 | 447 | raise logerror(_('log cache overlaps with new log entries,' |
|
442 | 448 | ' re-run without cache.')) |
|
443 | 449 | |
|
444 | 450 | log = oldlog + log |
|
445 | 451 | |
|
446 | 452 | # write the new cachefile |
|
447 | 453 | ui.note(_('writing cvs log cache %s\n') % cachefile) |
|
448 | 454 | pickle.dump(log, open(cachefile, 'w')) |
|
449 | 455 | else: |
|
450 | 456 | log = oldlog |
|
451 | 457 | |
|
452 | 458 | ui.status(_('%d log entries\n') % len(log)) |
|
453 | 459 | |
|
454 | 460 | hook.hook(ui, None, "cvslog", True, log=log) |
|
455 | 461 | |
|
456 | 462 | return log |
|
457 | 463 | |
|
458 | 464 | |
|
459 | 465 | class changeset(object): |
|
460 | 466 | '''Class changeset has the following attributes: |
|
461 | 467 | .id - integer identifying this changeset (list index) |
|
462 | 468 | .author - author name as CVS knows it |
|
463 | 469 | .branch - name of branch this changeset is on, or None |
|
464 | 470 | .comment - commit message |
|
465 | 471 | .date - the commit date as a (time,tz) tuple |
|
466 | 472 | .entries - list of logentry objects in this changeset |
|
467 | 473 | .parents - list of one or two parent changesets |
|
468 | 474 | .tags - list of tags on this changeset |
|
469 | 475 | .synthetic - from synthetic revision "file ... added on branch ..." |
|
470 | 476 | .mergepoint- the branch that has been merged from |
|
471 | 477 | (if present in rlog output) |
|
472 | 478 | .branchpoints- the branches that start at the current entry |
|
473 | 479 | ''' |
|
474 | 480 | def __init__(self, **entries): |
|
475 | 481 | self.synthetic = False |
|
476 | 482 | self.__dict__.update(entries) |
|
477 | 483 | |
|
478 | 484 | def __repr__(self): |
|
479 | 485 | return "<%s at 0x%x: %s>" % (self.__class__.__name__, |
|
480 | 486 | id(self), |
|
481 | 487 | getattr(self, 'id', "(no id)")) |
|
482 | 488 | |
|
483 | 489 | def createchangeset(ui, log, fuzz=60, mergefrom=None, mergeto=None): |
|
484 | 490 | '''Convert log into changesets.''' |
|
485 | 491 | |
|
486 | 492 | ui.status(_('creating changesets\n')) |
|
487 | 493 | |
|
488 | 494 | # Merge changesets |
|
489 | 495 | |
|
490 | 496 | log.sort(key=lambda x: (x.comment, x.author, x.branch, x.date)) |
|
491 | 497 | |
|
492 | 498 | changesets = [] |
|
493 | 499 | files = set() |
|
494 | 500 | c = None |
|
495 | 501 | for i, e in enumerate(log): |
|
496 | 502 | |
|
497 | 503 | # Check if log entry belongs to the current changeset or not. |
|
498 | 504 | |
|
499 | 505 | # Since CVS is file centric, two different file revisions with |
|
500 | 506 | # different branchpoints should be treated as belonging to two |
|
501 | 507 | # different changesets (and the ordering is important and not |
|
502 | 508 | # honoured by cvsps at this point). |
|
503 | 509 | # |
|
504 | 510 | # Consider the following case: |
|
505 | 511 | # foo 1.1 branchpoints: [MYBRANCH] |
|
506 | 512 | # bar 1.1 branchpoints: [MYBRANCH, MYBRANCH2] |
|
507 | 513 | # |
|
508 | 514 | # Here foo is part only of MYBRANCH, but not MYBRANCH2, e.g. a |
|
509 | 515 | # later version of foo may be in MYBRANCH2, so foo should be the |
|
510 | 516 | # first changeset and bar the next and MYBRANCH and MYBRANCH2 |
|
511 | 517 | # should both start off of the bar changeset. No provisions are |
|
512 | 518 | # made to ensure that this is, in fact, what happens. |
|
513 | 519 | if not (c and |
|
514 | 520 | e.comment == c.comment and |
|
515 | 521 | e.author == c.author and |
|
516 | 522 | e.branch == c.branch and |
|
517 | 523 | (not util.safehasattr(e, 'branchpoints') or |
|
518 | 524 | not util.safehasattr (c, 'branchpoints') or |
|
519 | 525 | e.branchpoints == c.branchpoints) and |
|
520 | 526 | ((c.date[0] + c.date[1]) <= |
|
521 | 527 | (e.date[0] + e.date[1]) <= |
|
522 | 528 | (c.date[0] + c.date[1]) + fuzz) and |
|
523 | 529 | e.file not in files): |
|
524 | 530 | c = changeset(comment=e.comment, author=e.author, |
|
525 | 531 | branch=e.branch, date=e.date, entries=[], |
|
526 | 532 | mergepoint=getattr(e, 'mergepoint', None), |
|
527 | 533 | branchpoints=getattr(e, 'branchpoints', set())) |
|
528 | 534 | changesets.append(c) |
|
529 | 535 | files = set() |
|
530 | 536 | if len(changesets) % 100 == 0: |
|
531 | 537 | t = '%d %s' % (len(changesets), repr(e.comment)[1:-1]) |
|
532 | 538 | ui.status(util.ellipsis(t, 80) + '\n') |
|
533 | 539 | |
|
534 | 540 | c.entries.append(e) |
|
535 | 541 | files.add(e.file) |
|
536 | 542 | c.date = e.date # changeset date is date of latest commit in it |
|
537 | 543 | |
|
538 | 544 | # Mark synthetic changesets |
|
539 | 545 | |
|
540 | 546 | for c in changesets: |
|
541 | 547 | # Synthetic revisions always get their own changeset, because |
|
542 | 548 | # the log message includes the filename. E.g. if you add file3 |
|
543 | 549 | # and file4 on a branch, you get four log entries and three |
|
544 | 550 | # changesets: |
|
545 | 551 | # "File file3 was added on branch ..." (synthetic, 1 entry) |
|
546 | 552 | # "File file4 was added on branch ..." (synthetic, 1 entry) |
|
547 | 553 | # "Add file3 and file4 to fix ..." (real, 2 entries) |
|
548 | 554 | # Hence the check for 1 entry here. |
|
549 | 555 | c.synthetic = len(c.entries) == 1 and c.entries[0].synthetic |
|
550 | 556 | |
|
551 | 557 | # Sort files in each changeset |
|
552 | 558 | |
|
553 | 559 | for c in changesets: |
|
554 | 560 | def pathcompare(l, r): |
|
555 | 561 | 'Mimic cvsps sorting order' |
|
556 | 562 | l = l.split('/') |
|
557 | 563 | r = r.split('/') |
|
558 | 564 | nl = len(l) |
|
559 | 565 | nr = len(r) |
|
560 | 566 | n = min(nl, nr) |
|
561 | 567 | for i in range(n): |
|
562 | 568 | if i + 1 == nl and nl < nr: |
|
563 | 569 | return -1 |
|
564 | 570 | elif i + 1 == nr and nl > nr: |
|
565 | 571 | return +1 |
|
566 | 572 | elif l[i] < r[i]: |
|
567 | 573 | return -1 |
|
568 | 574 | elif l[i] > r[i]: |
|
569 | 575 | return +1 |
|
570 | 576 | return 0 |
|
571 | 577 | def entitycompare(l, r): |
|
572 | 578 | return pathcompare(l.file, r.file) |
|
573 | 579 | |
|
574 | 580 | c.entries.sort(entitycompare) |
|
575 | 581 | |
|
576 | 582 | # Sort changesets by date |
|
577 | 583 | |
|
578 | 584 | def cscmp(l, r): |
|
579 | 585 | d = sum(l.date) - sum(r.date) |
|
580 | 586 | if d: |
|
581 | 587 | return d |
|
582 | 588 | |
|
583 | 589 | # detect vendor branches and initial commits on a branch |
|
584 | 590 | le = {} |
|
585 | 591 | for e in l.entries: |
|
586 | 592 | le[e.rcs] = e.revision |
|
587 | 593 | re = {} |
|
588 | 594 | for e in r.entries: |
|
589 | 595 | re[e.rcs] = e.revision |
|
590 | 596 | |
|
591 | 597 | d = 0 |
|
592 | 598 | for e in l.entries: |
|
593 | 599 | if re.get(e.rcs, None) == e.parent: |
|
594 | 600 | assert not d |
|
595 | 601 | d = 1 |
|
596 | 602 | break |
|
597 | 603 | |
|
598 | 604 | for e in r.entries: |
|
599 | 605 | if le.get(e.rcs, None) == e.parent: |
|
600 | 606 | assert not d |
|
601 | 607 | d = -1 |
|
602 | 608 | break |
|
603 | 609 | |
|
604 | 610 | return d |
|
605 | 611 | |
|
606 | 612 | changesets.sort(cscmp) |
|
607 | 613 | |
|
608 | 614 | # Collect tags |
|
609 | 615 | |
|
610 | 616 | globaltags = {} |
|
611 | 617 | for c in changesets: |
|
612 | 618 | for e in c.entries: |
|
613 | 619 | for tag in e.tags: |
|
614 | 620 | # remember which is the latest changeset to have this tag |
|
615 | 621 | globaltags[tag] = c |
|
616 | 622 | |
|
617 | 623 | for c in changesets: |
|
618 | 624 | tags = set() |
|
619 | 625 | for e in c.entries: |
|
620 | 626 | tags.update(e.tags) |
|
621 | 627 | # remember tags only if this is the latest changeset to have it |
|
622 | 628 | c.tags = sorted(tag for tag in tags if globaltags[tag] is c) |
|
623 | 629 | |
|
624 | 630 | # Find parent changesets, handle {{mergetobranch BRANCHNAME}} |
|
625 | 631 | # by inserting dummy changesets with two parents, and handle |
|
626 | 632 | # {{mergefrombranch BRANCHNAME}} by setting two parents. |
|
627 | 633 | |
|
628 | 634 | if mergeto is None: |
|
629 | 635 | mergeto = r'{{mergetobranch ([-\w]+)}}' |
|
630 | 636 | if mergeto: |
|
631 | 637 | mergeto = re.compile(mergeto) |
|
632 | 638 | |
|
633 | 639 | if mergefrom is None: |
|
634 | 640 | mergefrom = r'{{mergefrombranch ([-\w]+)}}' |
|
635 | 641 | if mergefrom: |
|
636 | 642 | mergefrom = re.compile(mergefrom) |
|
637 | 643 | |
|
638 | 644 | versions = {} # changeset index where we saw any particular file version |
|
639 | 645 | branches = {} # changeset index where we saw a branch |
|
640 | 646 | n = len(changesets) |
|
641 | 647 | i = 0 |
|
642 | 648 | while i < n: |
|
643 | 649 | c = changesets[i] |
|
644 | 650 | |
|
645 | 651 | for f in c.entries: |
|
646 | 652 | versions[(f.rcs, f.revision)] = i |
|
647 | 653 | |
|
648 | 654 | p = None |
|
649 | 655 | if c.branch in branches: |
|
650 | 656 | p = branches[c.branch] |
|
651 | 657 | else: |
|
652 | 658 | # first changeset on a new branch |
|
653 | 659 | # the parent is a changeset with the branch in its |
|
654 | 660 | # branchpoints such that it is the latest possible |
|
655 | 661 | # commit without any intervening, unrelated commits. |
|
656 | 662 | |
|
657 | 663 | for candidate in xrange(i): |
|
658 | 664 | if c.branch not in changesets[candidate].branchpoints: |
|
659 | 665 | if p is not None: |
|
660 | 666 | break |
|
661 | 667 | continue |
|
662 | 668 | p = candidate |
|
663 | 669 | |
|
664 | 670 | c.parents = [] |
|
665 | 671 | if p is not None: |
|
666 | 672 | p = changesets[p] |
|
667 | 673 | |
|
668 | 674 | # Ensure no changeset has a synthetic changeset as a parent. |
|
669 | 675 | while p.synthetic: |
|
670 | 676 | assert len(p.parents) <= 1, \ |
|
671 | 677 | _('synthetic changeset cannot have multiple parents') |
|
672 | 678 | if p.parents: |
|
673 | 679 | p = p.parents[0] |
|
674 | 680 | else: |
|
675 | 681 | p = None |
|
676 | 682 | break |
|
677 | 683 | |
|
678 | 684 | if p is not None: |
|
679 | 685 | c.parents.append(p) |
|
680 | 686 | |
|
681 | 687 | if c.mergepoint: |
|
682 | 688 | if c.mergepoint == 'HEAD': |
|
683 | 689 | c.mergepoint = None |
|
684 | 690 | c.parents.append(changesets[branches[c.mergepoint]]) |
|
685 | 691 | |
|
686 | 692 | if mergefrom: |
|
687 | 693 | m = mergefrom.search(c.comment) |
|
688 | 694 | if m: |
|
689 | 695 | m = m.group(1) |
|
690 | 696 | if m == 'HEAD': |
|
691 | 697 | m = None |
|
692 | 698 | try: |
|
693 | 699 | candidate = changesets[branches[m]] |
|
694 | 700 | except KeyError: |
|
695 | 701 | ui.warn(_("warning: CVS commit message references " |
|
696 | 702 | "non-existent branch %r:\n%s\n") |
|
697 | 703 | % (m, c.comment)) |
|
698 | 704 | if m in branches and c.branch != m and not candidate.synthetic: |
|
699 | 705 | c.parents.append(candidate) |
|
700 | 706 | |
|
701 | 707 | if mergeto: |
|
702 | 708 | m = mergeto.search(c.comment) |
|
703 | 709 | if m: |
|
704 | 710 | try: |
|
705 | 711 | m = m.group(1) |
|
706 | 712 | if m == 'HEAD': |
|
707 | 713 | m = None |
|
708 | 714 | except: |
|
709 | 715 | m = None # if no group found then merge to HEAD |
|
710 | 716 | if m in branches and c.branch != m: |
|
711 | 717 | # insert empty changeset for merge |
|
712 | 718 | cc = changeset( |
|
713 | 719 | author=c.author, branch=m, date=c.date, |
|
714 | 720 | comment='convert-repo: CVS merge from branch %s' |
|
715 | 721 | % c.branch, |
|
716 | 722 | entries=[], tags=[], |
|
717 | 723 | parents=[changesets[branches[m]], c]) |
|
718 | 724 | changesets.insert(i + 1, cc) |
|
719 | 725 | branches[m] = i + 1 |
|
720 | 726 | |
|
721 | 727 | # adjust our loop counters now we have inserted a new entry |
|
722 | 728 | n += 1 |
|
723 | 729 | i += 2 |
|
724 | 730 | continue |
|
725 | 731 | |
|
726 | 732 | branches[c.branch] = i |
|
727 | 733 | i += 1 |
|
728 | 734 | |
|
729 | 735 | # Drop synthetic changesets (safe now that we have ensured no other |
|
730 | 736 | # changesets can have them as parents). |
|
731 | 737 | i = 0 |
|
732 | 738 | while i < len(changesets): |
|
733 | 739 | if changesets[i].synthetic: |
|
734 | 740 | del changesets[i] |
|
735 | 741 | else: |
|
736 | 742 | i += 1 |
|
737 | 743 | |
|
738 | 744 | # Number changesets |
|
739 | 745 | |
|
740 | 746 | for i, c in enumerate(changesets): |
|
741 | 747 | c.id = i + 1 |
|
742 | 748 | |
|
743 | 749 | ui.status(_('%d changeset entries\n') % len(changesets)) |
|
744 | 750 | |
|
745 | 751 | hook.hook(ui, None, "cvschangesets", True, changesets=changesets) |
|
746 | 752 | |
|
747 | 753 | return changesets |
|
748 | 754 | |
|
749 | 755 | |
|
750 | 756 | def debugcvsps(ui, *args, **opts): |
|
751 | 757 | '''Read CVS rlog for current directory or named path in |
|
752 | 758 | repository, and convert the log to changesets based on matching |
|
753 | 759 | commit log entries and dates. |
|
754 | 760 | ''' |
|
755 | 761 | if opts["new_cache"]: |
|
756 | 762 | cache = "write" |
|
757 | 763 | elif opts["update_cache"]: |
|
758 | 764 | cache = "update" |
|
759 | 765 | else: |
|
760 | 766 | cache = None |
|
761 | 767 | |
|
762 | 768 | revisions = opts["revisions"] |
|
763 | 769 | |
|
764 | 770 | try: |
|
765 | 771 | if args: |
|
766 | 772 | log = [] |
|
767 | 773 | for d in args: |
|
768 | 774 | log += createlog(ui, d, root=opts["root"], cache=cache) |
|
769 | 775 | else: |
|
770 | 776 | log = createlog(ui, root=opts["root"], cache=cache) |
|
771 | 777 | except logerror, e: |
|
772 | 778 | ui.write("%r\n"%e) |
|
773 | 779 | return |
|
774 | 780 | |
|
775 | 781 | changesets = createchangeset(ui, log, opts["fuzz"]) |
|
776 | 782 | del log |
|
777 | 783 | |
|
778 | 784 | # Print changesets (optionally filtered) |
|
779 | 785 | |
|
780 | 786 | off = len(revisions) |
|
781 | 787 | branches = {} # latest version number in each branch |
|
782 | 788 | ancestors = {} # parent branch |
|
783 | 789 | for cs in changesets: |
|
784 | 790 | |
|
785 | 791 | if opts["ancestors"]: |
|
786 | 792 | if cs.branch not in branches and cs.parents and cs.parents[0].id: |
|
787 | 793 | ancestors[cs.branch] = (changesets[cs.parents[0].id - 1].branch, |
|
788 | 794 | cs.parents[0].id) |
|
789 | 795 | branches[cs.branch] = cs.id |
|
790 | 796 | |
|
791 | 797 | # limit by branches |
|
792 | 798 | if opts["branches"] and (cs.branch or 'HEAD') not in opts["branches"]: |
|
793 | 799 | continue |
|
794 | 800 | |
|
795 | 801 | if not off: |
|
796 | 802 | # Note: trailing spaces on several lines here are needed to have |
|
797 | 803 | # bug-for-bug compatibility with cvsps. |
|
798 | 804 | ui.write('---------------------\n') |
|
799 | 805 | ui.write('PatchSet %d \n' % cs.id) |
|
800 | 806 | ui.write('Date: %s\n' % util.datestr(cs.date, |
|
801 | 807 | '%Y/%m/%d %H:%M:%S %1%2')) |
|
802 | 808 | ui.write('Author: %s\n' % cs.author) |
|
803 | 809 | ui.write('Branch: %s\n' % (cs.branch or 'HEAD')) |
|
804 | 810 | ui.write('Tag%s: %s \n' % (['', 's'][len(cs.tags) > 1], |
|
805 | 811 | ','.join(cs.tags) or '(none)')) |
|
806 | 812 | branchpoints = getattr(cs, 'branchpoints', None) |
|
807 | 813 | if branchpoints: |
|
808 | 814 | ui.write('Branchpoints: %s \n' % ', '.join(branchpoints)) |
|
809 | 815 | if opts["parents"] and cs.parents: |
|
810 | 816 | if len(cs.parents) > 1: |
|
811 | 817 | ui.write('Parents: %s\n' % |
|
812 | 818 | (','.join([str(p.id) for p in cs.parents]))) |
|
813 | 819 | else: |
|
814 | 820 | ui.write('Parent: %d\n' % cs.parents[0].id) |
|
815 | 821 | |
|
816 | 822 | if opts["ancestors"]: |
|
817 | 823 | b = cs.branch |
|
818 | 824 | r = [] |
|
819 | 825 | while b: |
|
820 | 826 | b, c = ancestors[b] |
|
821 | 827 | r.append('%s:%d:%d' % (b or "HEAD", c, branches[b])) |
|
822 | 828 | if r: |
|
823 | 829 | ui.write('Ancestors: %s\n' % (','.join(r))) |
|
824 | 830 | |
|
825 | 831 | ui.write('Log:\n') |
|
826 | 832 | ui.write('%s\n\n' % cs.comment) |
|
827 | 833 | ui.write('Members: \n') |
|
828 | 834 | for f in cs.entries: |
|
829 | 835 | fn = f.file |
|
830 | 836 | if fn.startswith(opts["prefix"]): |
|
831 | 837 | fn = fn[len(opts["prefix"]):] |
|
832 | 838 | ui.write('\t%s:%s->%s%s \n' % ( |
|
833 | 839 | fn, '.'.join([str(x) for x in f.parent]) or 'INITIAL', |
|
834 | 840 | '.'.join([str(x) for x in f.revision]), |
|
835 | 841 | ['', '(DEAD)'][f.dead])) |
|
836 | 842 | ui.write('\n') |
|
837 | 843 | |
|
838 | 844 | # have we seen the start tag? |
|
839 | 845 | if revisions and off: |
|
840 | 846 | if revisions[0] == str(cs.id) or \ |
|
841 | 847 | revisions[0] in cs.tags: |
|
842 | 848 | off = False |
|
843 | 849 | |
|
844 | 850 | # see if we reached the end tag |
|
845 | 851 | if len(revisions) > 1 and not off: |
|
846 | 852 | if revisions[1] == str(cs.id) or \ |
|
847 | 853 | revisions[1] in cs.tags: |
|
848 | 854 | break |
General Comments 0
You need to be logged in to leave comments.
Login now