« back to all changes in this revision
Viewing changes to bzrlib/weave.py
- browse files at revision 1393.1.39
- compare with another revision
- download diff
- download tarball
- view history from revision 1393.1.39
- Committer: Martin Pool
- Date: 2005-10-04 11:13:33 UTC
- mto: (1185.13.3)
- mto: This revision was merged to the branch mainline in revision 1403.
- Revision ID: mbp@sourcefrog.net-20051004111332-f7b8a6bd41b9fe22
- tweak capture_tree formatting
- files added:
- HACKING
- bzrlib/plugins
- bzrlib/selftest
- bzrlib/store
- bzrlib/transport
- bzrlib/util
- bzrlib/util/effbot
- bzrlib/util/effbot/org
- contrib/emacs
- patches
- tools
- files removed:
- bzrlib/remotebranch.py
- files renamed:
- bzrlib/store.py => bzrlib/store/__init__.py
- elementtree/ => bzrlib/util/elementtree/
- urlgrabber/ => bzrlib/util/urlgrabber/
- contrib/bash/bzr => contrib/bash/bzr.simple
- bzrlib/newinventory.py => contrib/newinventory.py
- files modified:
- .bzrignore
added
removed
bzrlib/weave.py
1
#! /usr/bin/python
2
3
# Copyright (C) 2005 Canonical Ltd
4
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
9
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
14
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
19
# Author: Martin Pool <mbp@canonical.com>
20
21
22
"""Weave - storage of related text file versions"""
23
24
25
# TODO: Perhaps have copy method for Weave instances?
26
27
# XXX: If we do weaves this way, will a merge still behave the same
28
# way if it's done in a different order? That's a pretty desirable
29
# property.
30
31
# TODO: Nothing here so far assumes the lines are really \n newlines,
32
# rather than being split up in some other way. We could accomodate
33
# binaries, perhaps by naively splitting on \n or perhaps using
34
# something like a rolling checksum.
35
36
# TODO: End marker for each version so we can stop reading?
37
38
# TODO: Check that no insertion occurs inside a deletion that was
39
# active in the version of the insertion.
40
41
# TODO: In addition to the SHA-1 check, perhaps have some code that
42
# checks structural constraints of the weave: ie that insertions are
43
# properly nested, that there is no text outside of an insertion, that
44
# insertions or deletions are not repeated, etc.
45
46
# TODO: Parallel-extract that passes back each line along with a
47
# description of which revisions include it. Nice for checking all
48
# shas or calculating stats in parallel.
49
50
# TODO: Using a single _extract routine and then processing the output
51
# is probably inefficient. It's simple enough that we can afford to
52
# have slight specializations for different ways its used: annotate,
53
# basis for add, get, etc.
54
55
# TODO: Perhaps the API should work only in names to hide the integer
56
# indexes from the user?
57
58
# TODO: Is there any potential performance win by having an add()
59
# variant that is passed a pre-cooked version of the single basis
60
# version?
61
62
63
64
import sha
65
from difflib import SequenceMatcher
66
67
68
69
70
class WeaveError(Exception):
71
"""Exception in processing weave"""
72
73
74
class WeaveFormatError(WeaveError):
75
"""Weave invariant violated"""
76
77
78
class Weave(object):
79
"""weave - versioned text file storage.
80
81
A Weave manages versions of line-based text files, keeping track
82
of the originating version for each line.
83
84
To clients the "lines" of the file are represented as a list of strings.
85
These strings will typically have terminal newline characters, but
86
this is not required. In particular files commonly do not have a newline
87
at the end of the file.
88
89
Texts can be identified in either of two ways:
90
91
* a nonnegative index number.
92
93
* a version-id string.
94
95
Typically the index number will be valid only inside this weave and
96
the version-id is used to reference it in the larger world.
97
98
The weave is represented as a list mixing edit instructions and
99
literal text. Each entry in _weave can be either a string (or
100
unicode), or a tuple. If a string, it means that the given line
101
should be output in the currently active revisions.
102
103
If a tuple, it gives a processing instruction saying in which
104
revisions the enclosed lines are active. The tuple has the form
105
(instruction, version).
106
107
The instruction can be '{' or '}' for an insertion block, and '['
108
and ']' for a deletion block respectively. The version is the
109
integer version index. There is no replace operator, only deletes
110
and inserts. For '}', the end of an insertion, there is no
111
version parameter because it always closes the most recently
112
opened insertion.
113
114
Constraints/notes:
115
116
* A later version can delete lines that were introduced by any
117
number of ancestor versions; this implies that deletion
118
instructions can span insertion blocks without regard to the
119
insertion block's nesting.
120
121
* Similarly, deletions need not be properly nested with regard to
122
each other, because they might have been generated by
123
independent revisions.
124
125
* Insertions are always made by inserting a new bracketed block
126
into a single point in the previous weave. This implies they
127
can nest but not overlap, and the nesting must always have later
128
insertions on the inside.
129
130
* It doesn't seem very useful to have an active insertion
131
inside an inactive insertion, but it might happen.
132
133
* Therefore, all instructions are always"considered"; that
134
is passed onto and off the stack. An outer inactive block
135
doesn't disable an inner block.
136
137
* Lines are enabled if the most recent enclosing insertion is
138
active and none of the enclosing deletions are active.
139
140
* There is no point having a deletion directly inside its own
141
insertion; you might as well just not write it. And there
142
should be no way to get an earlier version deleting a later
143
version.
144
145
_weave
146
Text of the weave; list of control instruction tuples and strings.
147
148
_parents
149
List of parents, indexed by version number.
150
It is only necessary to store the minimal set of parents for
151
each version; the parent's parents are implied.
152
153
_sha1s
154
List of hex SHA-1 of each version.
155
156
_names
157
List of symbolic names for each version. Each should be unique.
158
159
_name_map
160
For each name, the version number.
161
162
_weave_name
163
Descriptive name of this weave; typically the filename if known.
164
Set by read_weave.
165
"""
166
167
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
168
'_weave_name']
169
170
def __init__(self, weave_name=None):
171
self._weave = []
172
self._parents = []
173
self._sha1s = []
174
self._names = []
175
self._name_map = {}
176
self._weave_name = weave_name
177
178
179
def __eq__(self, other):
180
if not isinstance(other, Weave):
181
return False
182
return self._parents == other._parents \
183
and self._weave == other._weave \
184
and self._sha1s == other._sha1s
185
186
187
def __ne__(self, other):
188
return not self.__eq__(other)
189
190
191
def maybe_lookup(self, name_or_index):
192
"""Convert possible symbolic name to index, or pass through indexes."""
193
if isinstance(name_or_index, (int, long)):
194
return name_or_index
195
else:
196
return self.lookup(name_or_index)
197
198
199
def lookup(self, name):
200
"""Convert symbolic version name to index."""
201
try:
202
return self._name_map[name]
203
except KeyError:
204
raise WeaveError("name %r not present in weave %r" %
205
(name, self._weave_name))
206
207
208
def idx_to_name(self, version):
209
return self._names[version]
210
211
212
def _check_repeated_add(self, name, parents, text, sha1):
213
"""Check that a duplicated add is OK.
214
215
If it is, return the (old) index; otherwise raise an exception.
216
"""
217
idx = self.lookup(name)
218
if sorted(self._parents[idx]) != sorted(parents):
219
raise WeaveError("name \"%s\" already present in weave "
220
"with different parents" % name)
221
if sha1 != self._sha1s[idx]:
222
raise WeaveError("name \"%s\" already present in weave "
223
"with different text" % name)
224
return idx
225
226
227
228
def add(self, name, parents, text, sha1=None):
229
"""Add a single text on top of the weave.
230
231
Returns the index number of the newly added version.
232
233
name
234
Symbolic name for this version.
235
(Typically the revision-id of the revision that added it.)
236
237
parents
238
List or set of direct parent version numbers.
239
240
text
241
Sequence of lines to be added in the new version.
242
243
sha -- SHA-1 of the file, if known. This is trusted to be
244
correct if supplied.
245
"""
246
from bzrlib.osutils import sha_strings
247
248
assert isinstance(name, basestring)
249
if sha1 is None:
250
sha1 = sha_strings(text)
251
if name in self._name_map:
252
return self._check_repeated_add(name, parents, text, sha1)
253
254
parents = map(self.maybe_lookup, parents)
255
self._check_versions(parents)
256
## self._check_lines(text)
257
new_version = len(self._parents)
258
259
260
# if we abort after here the (in-memory) weave will be corrupt because only
261
# some fields are updated
262
self._parents.append(parents[:])
263
self._sha1s.append(sha1)
264
self._names.append(name)
265
self._name_map[name] = new_version
266
267
268
if not parents:
269
# special case; adding with no parents revision; can do
270
# this more quickly by just appending unconditionally.
271
# even more specially, if we're adding an empty text we
272
# need do nothing at all.
273
if text:
274
self._weave.append(('{', new_version))
275
self._weave.extend(text)
276
self._weave.append(('}', None))
277
278
return new_version
279
280
if len(parents) == 1:
281
pv = list(parents)[0]
282
if sha1 == self._sha1s[pv]:
283
# special case: same as the single parent
284
return new_version
285
286
287
ancestors = self.inclusions(parents)
288
289
l = self._weave
290
291
# basis a list of (origin, lineno, line)
292
basis_lineno = []
293
basis_lines = []
294
for origin, lineno, line in self._extract(ancestors):
295
basis_lineno.append(lineno)
296
basis_lines.append(line)
297
298
# another small special case: a merge, producing the same text
299
# as auto-merge
300
if text == basis_lines:
301
return new_version
302
303
# add a sentinal, because we can also match against the final line
304
basis_lineno.append(len(self._weave))
305
306
# XXX: which line of the weave should we really consider
307
# matches the end of the file? the current code says it's the
308
# last line of the weave?
309
310
#print 'basis_lines:', basis_lines
311
#print 'new_lines: ', lines
312
313
s = SequenceMatcher(None, basis_lines, text)
314
315
# offset gives the number of lines that have been inserted
316
# into the weave up to the current point; if the original edit instruction
317
# says to change line A then we actually change (A+offset)
318
offset = 0
319
320
for tag, i1, i2, j1, j2 in s.get_opcodes():
321
# i1,i2 are given in offsets within basis_lines; we need to map them
322
# back to offsets within the entire weave
323
#print 'raw match', tag, i1, i2, j1, j2
324
if tag == 'equal':
325
continue
326
327
i1 = basis_lineno[i1]
328
i2 = basis_lineno[i2]
329
330
assert 0 <= j1 <= j2 <= len(text)
331
332
#print tag, i1, i2, j1, j2
333
334
# the deletion and insertion are handled separately.
335
# first delete the region.
336
if i1 != i2:
337
self._weave.insert(i1+offset, ('[', new_version))
338
self._weave.insert(i2+offset+1, (']', new_version))
339
offset += 2
340
341
if j1 != j2:
342
# there may have been a deletion spanning up to
343
# i2; we want to insert after this region to make sure
344
# we don't destroy ourselves
345
i = i2 + offset
346
self._weave[i:i] = ([('{', new_version)]
347
+ text[j1:j2]
348
+ [('}', None)])
349
offset += 2 + (j2 - j1)
350
351
return new_version
352
353
def add_identical(self, old_rev_id, new_rev_id, parents):
354
"""Add an identical text to old_rev_id as new_rev_id."""
355
old_lines = self.get(self.lookup(old_rev_id))
356
self.add(new_rev_id, parents, old_lines)
357
358
def inclusions(self, versions):
359
"""Return set of all ancestors of given version(s)."""
360
i = set(versions)
361
for v in xrange(max(versions), 0, -1):
362
if v in i:
363
# include all its parents
364
i.update(self._parents[v])
365
return i
366
## except IndexError:
367
## raise ValueError("version %d not present in weave" % v)
368
369
370
def parents(self, version):
371
return self._parents[version]
372
373
374
def minimal_parents(self, version):
375
"""Find the minimal set of parents for the version."""
376
included = self._parents[version]
377
if not included:
378
return []
379
380
li = list(included)
381
li.sort(reverse=True)
382
383
mininc = []
384
gotit = set()
385
386
for pv in li:
387
if pv not in gotit:
388
mininc.append(pv)
389
gotit.update(self.inclusions(pv))
390
391
assert mininc[0] >= 0
392
assert mininc[-1] < version
393
return mininc
394
395
396
397
def _check_lines(self, text):
398
if not isinstance(text, list):
399
raise ValueError("text should be a list, not %s" % type(text))
400
401
for l in text:
402
if not isinstance(l, basestring):
403
raise ValueError("text line should be a string or unicode, not %s"
404
% type(l))
405
406
407
408
def _check_versions(self, indexes):
409
"""Check everything in the sequence of indexes is valid"""
410
for i in indexes:
411
try:
412
self._parents[i]
413
except IndexError:
414
raise IndexError("invalid version number %r" % i)
415
416
417
def annotate(self, name_or_index):
418
return list(self.annotate_iter(name_or_index))
419
420
421
def annotate_iter(self, name_or_index):
422
"""Yield list of (index-id, line) pairs for the specified version.
423
424
The index indicates when the line originated in the weave."""
425
incls = [self.maybe_lookup(name_or_index)]
426
for origin, lineno, text in self._extract(incls):
427
yield origin, text
428
429
430
def _walk(self):
431
"""Walk the weave.
432
433
Yields sequence of
434
(lineno, insert, deletes, text)
435
for each literal line.
436
"""
437
438
istack = []
439
dset = set()
440
441
lineno = 0 # line of weave, 0-based
442
443
for l in self._weave:
444
if isinstance(l, tuple):
445
c, v = l
446
isactive = None
447
if c == '{':
448
istack.append(v)
449
elif c == '}':
450
istack.pop()
451
elif c == '[':
452
assert v not in dset
453
dset.add(v)
454
elif c == ']':
455
dset.remove(v)
456
else:
457
raise WeaveFormatError('unexpected instruction %r'
458
% v)
459
else:
460
assert isinstance(l, basestring)
461
assert istack
462
yield lineno, istack[-1], dset, l
463
lineno += 1
464
465
466
467
def _extract(self, versions):
468
"""Yield annotation of lines in included set.
469
470
Yields a sequence of tuples (origin, lineno, text), where
471
origin is the origin version, lineno the index in the weave,
472
and text the text of the line.
473
474
The set typically but not necessarily corresponds to a version.
475
"""
476
for i in versions:
477
if not isinstance(i, int):
478
raise ValueError(i)
479
480
included = self.inclusions(versions)
481
482
istack = []
483
dset = set()
484
485
lineno = 0 # line of weave, 0-based
486
487
isactive = None
488
489
result = []
490
491
WFE = WeaveFormatError
492
493
for l in self._weave:
494
if isinstance(l, tuple):
495
c, v = l
496
isactive = None
497
if c == '{':
498
assert v not in istack
499
istack.append(v)
500
elif c == '}':
501
istack.pop()
502
elif c == '[':
503
if v in included:
504
assert v not in dset
505
dset.add(v)
506
else:
507
assert c == ']'
508
if v in included:
509
assert v in dset
510
dset.remove(v)
511
else:
512
assert isinstance(l, basestring)
513
if isactive is None:
514
isactive = (not dset) and istack and (istack[-1] in included)
515
if isactive:
516
result.append((istack[-1], lineno, l))
517
lineno += 1
518
519
if istack:
520
raise WFE("unclosed insertion blocks at end of weave",
521
istack)
522
if dset:
523
raise WFE("unclosed deletion blocks at end of weave",
524
dset)
525
526
return result
527
528
529
530
def get_iter(self, name_or_index):
531
"""Yield lines for the specified version."""
532
incls = [self.maybe_lookup(name_or_index)]
533
for origin, lineno, line in self._extract(incls):
534
yield line
535
536
537
def get_text(self, name_or_index):
538
return ''.join(self.get_iter(name_or_index))
539
assert isinstance(version, int)
540
541
542
def get_lines(self, name_or_index):
543
return list(self.get_iter(name_or_index))
544
545
546
get = get_lines
547
548
549
def mash_iter(self, included):
550
"""Return composed version of multiple included versions."""
551
included = map(self.maybe_lookup, included)
552
for origin, lineno, text in self._extract(included):
553
yield text
554
555
556
def dump(self, to_file):
557
from pprint import pprint
558
print >>to_file, "Weave._weave = ",
559
pprint(self._weave, to_file)
560
print >>to_file, "Weave._parents = ",
561
pprint(self._parents, to_file)
562
563
564
565
def numversions(self):
566
l = len(self._parents)
567
assert l == len(self._sha1s)
568
return l
569
570
571
def __len__(self):
572
return self.numversions()
573
574
575
def check(self, progress_bar=None):
576
# check no circular inclusions
577
for version in range(self.numversions()):
578
inclusions = list(self._parents[version])
579
if inclusions:
580
inclusions.sort()
581
if inclusions[-1] >= version:
582
raise WeaveFormatError("invalid included version %d for index %d"
583
% (inclusions[-1], version))
584
585
# try extracting all versions; this is a bit slow and parallel
586
# extraction could be used
587
nv = self.numversions()
588
for version in range(nv):
589
if progress_bar:
590
progress_bar.update('checking text', version, nv)
591
s = sha.new()
592
for l in self.get_iter(version):
593
s.update(l)
594
hd = s.hexdigest()
595
expected = self._sha1s[version]
596
if hd != expected:
597
raise WeaveError("mismatched sha1 for version %d; "
598
"got %s, expected %s"
599
% (version, hd, expected))
600
601
# TODO: check insertions are properly nested, that there are
602
# no lines outside of insertion blocks, that deletions are
603
# properly paired, etc.
604
605
606
607
def merge(self, merge_versions):
608
"""Automerge and mark conflicts between versions.
609
610
This returns a sequence, each entry describing alternatives
611
for a chunk of the file. Each of the alternatives is given as
612
a list of lines.
613
614
If there is a chunk of the file where there's no diagreement,
615
only one alternative is given.
616
"""
617
618
# approach: find the included versions common to all the
619
# merged versions
620
raise NotImplementedError()
621
622
623
624
def _delta(self, included, lines):
625
"""Return changes from basis to new revision.
626
627
The old text for comparison is the union of included revisions.
628
629
This is used in inserting a new text.
630
631
Delta is returned as a sequence of
632
(weave1, weave2, newlines).
633
634
This indicates that weave1:weave2 of the old weave should be
635
replaced by the sequence of lines in newlines. Note that
636
these line numbers are positions in the total weave and don't
637
correspond to the lines in any extracted version, or even the
638
extracted union of included versions.
639
640
If line1=line2, this is a pure insert; if newlines=[] this is a
641
pure delete. (Similar to difflib.)
642
"""
643
644
645
646
def plan_merge(self, ver_a, ver_b):
647
"""Return pseudo-annotation indicating how the two versions merge.
648
649
This is computed between versions a and b and their common
650
base.
651
652
Weave lines present in none of them are skipped entirely.
653
"""
654
inc_a = self.inclusions([ver_a])
655
inc_b = self.inclusions([ver_b])
656
inc_c = inc_a & inc_b
657
658
for lineno, insert, deleteset, line in self._walk():
659
if deleteset & inc_c:
660
# killed in parent; can't be in either a or b
661
# not relevant to our work
662
yield 'killed-base', line
663
elif insert in inc_c:
664
# was inserted in base
665
killed_a = bool(deleteset & inc_a)
666
killed_b = bool(deleteset & inc_b)
667
if killed_a and killed_b:
668
yield 'killed-both', line
669
elif killed_a:
670
yield 'killed-a', line
671
elif killed_b:
672
yield 'killed-b', line
673
else:
674
yield 'unchanged', line
675
elif insert in inc_a:
676
if deleteset & inc_a:
677
yield 'ghost-a', line
678
else:
679
# new in A; not in B
680
yield 'new-a', line
681
elif insert in inc_b:
682
if deleteset & inc_b:
683
yield 'ghost-b', line
684
else:
685
yield 'new-b', line
686
else:
687
# not in either revision
688
yield 'irrelevant', line
689
690
yield 'unchanged', '' # terminator
691
692
693
694
def weave_merge(self, plan):
695
lines_a = []
696
lines_b = []
697
ch_a = ch_b = False
698
699
for state, line in plan:
700
if state == 'unchanged' or state == 'killed-both':
701
# resync and flush queued conflicts changes if any
702
if not lines_a and not lines_b:
703
pass
704
elif ch_a and not ch_b:
705
# one-sided change:
706
for l in lines_a: yield l
707
elif ch_b and not ch_a:
708
for l in lines_b: yield l
709
elif lines_a == lines_b:
710
for l in lines_a: yield l
711
else:
712
yield '<<<<\n'
713
for l in lines_a: yield l
714
yield '====\n'
715
for l in lines_b: yield l
716
yield '>>>>\n'
717
718
del lines_a[:]
719
del lines_b[:]
720
ch_a = ch_b = False
721
722
if state == 'unchanged':
723
if line:
724
yield line
725
elif state == 'killed-a':
726
ch_a = True
727
lines_b.append(line)
728
elif state == 'killed-b':
729
ch_b = True
730
lines_a.append(line)
731
elif state == 'new-a':
732
ch_a = True
733
lines_a.append(line)
734
elif state == 'new-b':
735
ch_b = True
736
lines_b.append(line)
737
else:
738
assert state in ('irrelevant', 'ghost-a', 'ghost-b', 'killed-base',
739
'killed-both'), \
740
state
741
742
743
744
745
746
747
748
def weave_toc(w):
749
"""Show the weave's table-of-contents"""
750
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
751
for i in (6, 50, 10, 10):
752
print '-' * i,
753
print
754
for i in range(w.numversions()):
755
sha1 = w._sha1s[i]
756
name = w._names[i]
757
parent_str = ' '.join(map(str, w._parents[i]))
758
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
759
760
761
762
def weave_stats(weave_file, pb):
763
from bzrlib.weavefile import read_weave
764
765
wf = file(weave_file, 'rb')
766
w = read_weave(wf)
767
# FIXME: doesn't work on pipes
768
weave_size = wf.tell()
769
770
total = 0
771
vers = len(w)
772
for i in range(vers):
773
pb.update('checking sizes', i, vers)
774
for origin, lineno, line in w._extract([i]):
775
total += len(line)
776
777
pb.clear()
778
779
print 'versions %9d' % vers
780
print 'weave file %9d bytes' % weave_size
781
print 'total contents %9d bytes' % total
782
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
783
if vers:
784
avg = total/vers
785
print 'average size %9d bytes' % avg
786
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
787
788
789
def usage():
790
print """bzr weave tool
791
792
Experimental tool for weave algorithm.
793
794
usage:
795
weave init WEAVEFILE
796
Create an empty weave file
797
weave get WEAVEFILE VERSION
798
Write out specified version.
799
weave check WEAVEFILE
800
Check consistency of all versions.
801
weave toc WEAVEFILE
802
Display table of contents.
803
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
804
Add NEWTEXT, with specified parent versions.
805
weave annotate WEAVEFILE VERSION
806
Display origin of each line.
807
weave mash WEAVEFILE VERSION...
808
Display composite of all selected versions.
809
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
810
Auto-merge two versions and display conflicts.
811
weave diff WEAVEFILE VERSION1 VERSION2
812
Show differences between two versions.
813
814
example:
815
816
% weave init foo.weave
817
% vi foo.txt
818
% weave add foo.weave ver0 < foo.txt
819
added version 0
820
821
(create updated version)
822
% vi foo.txt
823
% weave get foo.weave 0 | diff -u - foo.txt
824
% weave add foo.weave ver1 0 < foo.txt
825
added version 1
826
827
% weave get foo.weave 0 > foo.txt (create forked version)
828
% vi foo.txt
829
% weave add foo.weave ver2 0 < foo.txt
830
added version 2
831
832
% weave merge foo.weave 1 2 > foo.txt (merge them)
833
% vi foo.txt (resolve conflicts)
834
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
835
836
"""
837
838
839
840
def main(argv):
841
import sys
842
import os
843
try:
844
import bzrlib
845
except ImportError:
846
# in case we're run directly from the subdirectory
847
sys.path.append('..')
848
import bzrlib
849
from bzrlib.weavefile import write_weave, read_weave
850
from bzrlib.progress import ProgressBar
851
852
try:
853
import psyco
854
psyco.full()
855
except ImportError:
856
pass
857
858
if len(argv) < 2:
859
usage()
860
return 0
861
862
cmd = argv[1]
863
864
def readit():
865
return read_weave(file(argv[2], 'rb'))
866
867
if cmd == 'help':
868
usage()
869
elif cmd == 'add':
870
w = readit()
871
# at the moment, based on everything in the file
872
name = argv[3]
873
parents = map(int, argv[4:])
874
lines = sys.stdin.readlines()
875
ver = w.add(name, parents, lines)
876
write_weave(w, file(argv[2], 'wb'))
877
print 'added version %r %d' % (name, ver)
878
elif cmd == 'init':
879
fn = argv[2]
880
if os.path.exists(fn):
881
raise IOError("file exists")
882
w = Weave()
883
write_weave(w, file(fn, 'wb'))
884
elif cmd == 'get': # get one version
885
w = readit()
886
sys.stdout.writelines(w.get_iter(int(argv[3])))
887
888
elif cmd == 'mash': # get composite
889
w = readit()
890
sys.stdout.writelines(w.mash_iter(map(int, argv[3:])))
891
892
elif cmd == 'diff':
893
from difflib import unified_diff
894
w = readit()
895
fn = argv[2]
896
v1, v2 = map(int, argv[3:5])
897
lines1 = w.get(v1)
898
lines2 = w.get(v2)
899
diff_gen = unified_diff(lines1, lines2,
900
'%s version %d' % (fn, v1),
901
'%s version %d' % (fn, v2))
902
sys.stdout.writelines(diff_gen)
903
904
elif cmd == 'annotate':
905
w = readit()
906
# newline is added to all lines regardless; too hard to get
907
# reasonable formatting otherwise
908
lasto = None
909
for origin, text in w.annotate(int(argv[3])):
910
text = text.rstrip('\r\n')
911
if origin == lasto:
912
print ' | %s' % (text)
913
else:
914
print '%5d | %s' % (origin, text)
915
lasto = origin
916
917
elif cmd == 'toc':
918
weave_toc(readit())
919
920
elif cmd == 'stats':
921
weave_stats(argv[2], ProgressBar())
922
923
elif cmd == 'check':
924
w = readit()
925
pb = ProgressBar()
926
w.check(pb)
927
pb.clear()
928
print '%d versions ok' % w.numversions()
929
930
elif cmd == 'inclusions':
931
w = readit()
932
print ' '.join(map(str, w.inclusions([int(argv[3])])))
933
934
elif cmd == 'parents':
935
w = readit()
936
print ' '.join(map(str, w._parents[int(argv[3])]))
937
938
elif cmd == 'plan-merge':
939
w = readit()
940
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
941
if line:
942
print '%14s | %s' % (state, line),
943
944
elif cmd == 'merge':
945
w = readit()
946
p = w.plan_merge(int(argv[3]), int(argv[4]))
947
sys.stdout.writelines(w.weave_merge(p))
948
949
elif cmd == 'mash-merge':
950
if len(argv) != 5:
951
usage()
952
return 1
953
954
w = readit()
955
v1, v2 = map(int, argv[3:5])
956
957
basis = w.inclusions([v1]).intersection(w.inclusions([v2]))
958
959
base_lines = list(w.mash_iter(basis))
960
a_lines = list(w.get(v1))
961
b_lines = list(w.get(v2))
962
963
from bzrlib.merge3 import Merge3
964
m3 = Merge3(base_lines, a_lines, b_lines)
965
966
name_a = 'version %d' % v1
967
name_b = 'version %d' % v2
968
sys.stdout.writelines(m3.merge_lines(name_a=name_a, name_b=name_b))
969
else:
970
raise ValueError('unknown command %r' % cmd)
971
972
973
974
def profile_main(argv):
975
import tempfile, hotshot, hotshot.stats
976
977
prof_f = tempfile.NamedTemporaryFile()
978
979
prof = hotshot.Profile(prof_f.name)
980
981
ret = prof.runcall(main, argv)
982
prof.close()
983
984
stats = hotshot.stats.load(prof_f.name)
985
#stats.strip_dirs()
986
stats.sort_stats('cumulative')
987
## XXX: Might like to write to stderr or the trace file instead but
988
## print_stats seems hardcoded to stdout
989
stats.print_stats(20)
990
991
return ret
992
993
994
if __name__ == '__main__':
995
import sys
996
if '--profile' in sys.argv:
997
args = sys.argv[:]
998
args.remove('--profile')
999
sys.exit(profile_main(args))
1000
else:
1001
sys.exit(main(sys.argv))
1002
Loggerhead is a web-based interface for Breezy
Version: 2.0.1