3
# Copyright (C) 2005 Canonical Ltd
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.
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.
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
19
# Author: Martin Pool <mbp@canonical.com>
22
"""Weave - storage of related text file versions"""
24
# before intset (r923) 2000 versions in 41.5s
25
# with intset (r926) 2000 versions in 93s !!!
26
# better to just use plain sets.
28
# making _extract build and return a list, rather than being a generator
31
# with python -O, r923 does 2000 versions in 36.87s
33
# with optimizations to avoid mutating lists - 35.75! I guess copying
34
# all the elements every time costs more than the small manipulations.
35
# a surprisingly small change.
37
# r931, which avoids using a generator for extract, does 36.98s
39
# with memoized inclusions, takes 41.49s; not very good
41
# with slots, takes 37.35s; without takes 39.16, a bit surprising
43
# with the delta calculation mixed in with the add method, rather than
44
# separated, takes 36.78s
46
# with delta folded in and mutation of the list, 36.13s
48
# with all this and simplification of add code, 33s
54
# TODO: Perhaps have copy method for Weave instances?
56
# XXX: If we do weaves this way, will a merge still behave the same
57
# way if it's done in a different order? That's a pretty desirable
60
# TODO: Nothing here so far assumes the lines are really \n newlines,
61
# rather than being split up in some other way. We could accomodate
62
# binaries, perhaps by naively splitting on \n or perhaps using
63
# something like a rolling checksum.
65
# TODO: End marker for each version so we can stop reading?
67
# TODO: Check that no insertion occurs inside a deletion that was
68
# active in the version of the insertion.
70
# TODO: In addition to the SHA-1 check, perhaps have some code that
71
# checks structural constraints of the weave: ie that insertions are
72
# properly nested, that there is no text outside of an insertion, that
73
# insertions or deletions are not repeated, etc.
75
# TODO: Parallel-extract that passes back each line along with a
76
# description of which revisions include it. Nice for checking all
79
# TODO: Using a single _extract routine and then processing the output
80
# is probably inefficient. It's simple enough that we can afford to
81
# have slight specializations for different ways its used: annotate,
82
# basis for add, get, etc.
84
# TODO: Perhaps the API should work only in names to hide the integer
85
# indexes from the user?
90
from difflib import SequenceMatcher
92
from cStringIO import StringIO
94
from bzrlib.osutils import sha_strings
97
class WeaveError(Exception):
98
"""Exception in processing weave"""
101
class WeaveFormatError(WeaveError):
102
"""Weave invariant violated"""
106
"""weave - versioned text file storage.
108
A Weave manages versions of line-based text files, keeping track
109
of the originating version for each line.
111
To clients the "lines" of the file are represented as a list of strings.
112
These strings will typically have terminal newline characters, but
113
this is not required. In particular files commonly do not have a newline
114
at the end of the file.
116
Texts can be identified in either of two ways:
118
* a nonnegative index number.
120
* a version-id string.
122
Typically the index number will be valid only inside this weave and
123
the version-id is used to reference it in the larger world.
125
The weave is represented as a list mixing edit instructions and
126
literal text. Each entry in _weave can be either a string (or
127
unicode), or a tuple. If a string, it means that the given line
128
should be output in the currently active revisions.
130
If a tuple, it gives a processing instruction saying in which
131
revisions the enclosed lines are active. The tuple has the form
132
(instruction, version).
134
The instruction can be '{' or '}' for an insertion block, and '['
135
and ']' for a deletion block respectively. The version is the
136
integer version index. There is no replace operator, only deletes
137
and inserts. For '}', the end of an insertion, there is no
138
version parameter because it always closes the most recently
143
* A later version can delete lines that were introduced by any
144
number of ancestor versions; this implies that deletion
145
instructions can span insertion blocks without regard to the
146
insertion block's nesting.
148
* Similarly, deletions need not be properly nested with regard to
149
each other, because they might have been generated by
150
independent revisions.
152
* Insertions are always made by inserting a new bracketed block
153
into a single point in the previous weave. This implies they
154
can nest but not overlap, and the nesting must always have later
155
insertions on the inside.
157
* It doesn't seem very useful to have an active insertion
158
inside an inactive insertion, but it might happen.
160
* Therefore, all instructions are always"considered"; that
161
is passed onto and off the stack. An outer inactive block
162
doesn't disable an inner block.
164
* Lines are enabled if the most recent enclosing insertion is
165
active and none of the enclosing deletions are active.
167
* There is no point having a deletion directly inside its own
168
insertion; you might as well just not write it. And there
169
should be no way to get an earlier version deleting a later
173
Text of the weave; list of control instruction tuples and strings.
176
List of parents, indexed by version number.
177
It is only necessary to store the minimal set of parents for
178
each version; the parent's parents are implied.
181
List of hex SHA-1 of each version.
184
List of symbolic names for each version. Each should be unique.
187
For each name, the version number.
190
Descriptive name of this weave; typically the filename if known.
194
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
197
def __init__(self, weave_name=None):
203
self._weave_name = weave_name
206
def __eq__(self, other):
207
if not isinstance(other, Weave):
209
return self._parents == other._parents \
210
and self._weave == other._weave \
211
and self._sha1s == other._sha1s
214
def __ne__(self, other):
215
return not self.__eq__(other)
218
def maybe_lookup(self, name_or_index):
219
"""Convert possible symbolic name to index, or pass through indexes."""
220
if isinstance(name_or_index, (int, long)):
223
return self.lookup(name_or_index)
226
def lookup(self, name):
227
"""Convert symbolic version name to index."""
229
return self._name_map[name]
231
raise WeaveError("name %r not present in weave %r" %
232
(name, self._weave_name))
235
def idx_to_name(self, version):
236
return self._names[version]
239
def _check_repeated_add(self, name, parents, text, sha1):
240
"""Check that a duplicated add is OK.
242
If it is, return the (old) index; otherwise raise an exception.
244
idx = self.lookup(name)
245
if sorted(self._parents[idx]) != sorted(parents):
246
raise WeaveError("name \"%s\" already present in weave "
247
"with different parents" % name)
248
if sha1 != self._sha1s[idx]:
249
raise WeaveError("name \"%s\" already present in weave "
250
"with different text" % name)
255
def add(self, name, parents, text, sha1=None):
256
"""Add a single text on top of the weave.
258
Returns the index number of the newly added version.
261
Symbolic name for this version.
262
(Typically the revision-id of the revision that added it.)
265
List or set of direct parent version numbers.
268
Sequence of lines to be added in the new version.
270
sha -- SHA-1 of the file, if known. This is trusted to be
274
assert isinstance(name, basestring)
276
sha1 = sha_strings(text)
277
if name in self._name_map:
278
return self._check_repeated_add(name, parents, text, sha1)
280
parents = map(self.maybe_lookup, parents)
281
self._check_versions(parents)
282
## self._check_lines(text)
283
new_version = len(self._parents)
286
# if we abort after here the (in-memory) weave will be corrupt because only
287
# some fields are updated
288
self._parents.append(parents[:])
289
self._sha1s.append(sha1)
290
self._names.append(name)
291
self._name_map[name] = new_version
295
# special case; adding with no parents revision; can do
296
# this more quickly by just appending unconditionally.
297
# even more specially, if we're adding an empty text we
298
# need do nothing at all.
300
self._weave.append(('{', new_version))
301
self._weave.extend(text)
302
self._weave.append(('}', None))
306
if len(parents) == 1:
307
pv = list(parents)[0]
308
if sha1 == self._sha1s[pv]:
309
# special case: same as the single parent
313
ancestors = self.inclusions(parents)
317
# basis a list of (origin, lineno, line)
320
for origin, lineno, line in self._extract(ancestors):
321
basis_lineno.append(lineno)
322
basis_lines.append(line)
324
# another small special case: a merge, producing the same text
326
if text == basis_lines:
329
# add a sentinal, because we can also match against the final line
330
basis_lineno.append(len(self._weave))
332
# XXX: which line of the weave should we really consider
333
# matches the end of the file? the current code says it's the
334
# last line of the weave?
336
#print 'basis_lines:', basis_lines
337
#print 'new_lines: ', lines
339
s = SequenceMatcher(None, basis_lines, text)
341
# offset gives the number of lines that have been inserted
342
# into the weave up to the current point; if the original edit instruction
343
# says to change line A then we actually change (A+offset)
346
for tag, i1, i2, j1, j2 in s.get_opcodes():
347
# i1,i2 are given in offsets within basis_lines; we need to map them
348
# back to offsets within the entire weave
349
#print 'raw match', tag, i1, i2, j1, j2
353
i1 = basis_lineno[i1]
354
i2 = basis_lineno[i2]
356
assert 0 <= j1 <= j2 <= len(text)
358
#print tag, i1, i2, j1, j2
360
# the deletion and insertion are handled separately.
361
# first delete the region.
363
self._weave.insert(i1+offset, ('[', new_version))
364
self._weave.insert(i2+offset+1, (']', new_version))
368
# there may have been a deletion spanning up to
369
# i2; we want to insert after this region to make sure
370
# we don't destroy ourselves
372
self._weave[i:i] = ([('{', new_version)]
375
offset += 2 + (j2 - j1)
380
def inclusions(self, versions):
381
"""Return set of all ancestors of given version(s)."""
387
# include all its parents
388
i.update(self._parents[v])
392
raise ValueError("version %d not present in weave" % v)
395
def parents(self, version):
396
return self._parents[version]
399
def minimal_parents(self, version):
400
"""Find the minimal set of parents for the version."""
401
included = self._parents[version]
406
li.sort(reverse=True)
414
gotit.update(self.inclusions(pv))
416
assert mininc[0] >= 0
417
assert mininc[-1] < version
422
def _check_lines(self, text):
423
if not isinstance(text, list):
424
raise ValueError("text should be a list, not %s" % type(text))
427
if not isinstance(l, basestring):
428
raise ValueError("text line should be a string or unicode, not %s"
433
def _check_versions(self, indexes):
434
"""Check everything in the sequence of indexes is valid"""
439
raise IndexError("invalid version number %r" % i)
442
def annotate(self, name_or_index):
443
return list(self.annotate_iter(name_or_index))
446
def annotate_iter(self, name_or_index):
447
"""Yield list of (index-id, line) pairs for the specified version.
449
The index indicates when the line originated in the weave."""
450
incls = [self.maybe_lookup(name_or_index)]
451
for origin, lineno, text in self._extract(incls):
459
(lineno, insert, deletes, text)
460
for each literal line.
466
lineno = 0 # line of weave, 0-based
468
for l in self._weave:
469
if isinstance(l, tuple):
482
raise WeaveFormatError('unexpected instruction %r'
485
assert isinstance(l, basestring)
487
yield lineno, istack[-1], dset, l
492
def _extract(self, versions):
493
"""Yield annotation of lines in included set.
495
Yields a sequence of tuples (origin, lineno, text), where
496
origin is the origin version, lineno the index in the weave,
497
and text the text of the line.
499
The set typically but not necessarily corresponds to a version.
502
if not isinstance(i, int):
505
included = self.inclusions(versions)
510
lineno = 0 # line of weave, 0-based
516
WFE = WeaveFormatError
518
for l in self._weave:
519
if isinstance(l, tuple):
523
assert v not in istack
537
assert isinstance(l, basestring)
539
isactive = (not dset) and istack and (istack[-1] in included)
541
result.append((istack[-1], lineno, l))
545
raise WFE("unclosed insertion blocks at end of weave",
548
raise WFE("unclosed deletion blocks at end of weave",
555
def get_iter(self, name_or_index):
556
"""Yield lines for the specified version."""
557
incls = [self.maybe_lookup(name_or_index)]
558
for origin, lineno, line in self._extract(incls):
562
def get_text(self, version):
563
assert isinstance(version, int)
565
s.writelines(self.get_iter(version))
569
def get(self, name_or_index):
570
return list(self.get_iter(name_or_index))
573
def mash_iter(self, included):
574
"""Return composed version of multiple included versions."""
575
for origin, lineno, text in self._extract(included):
579
def dump(self, to_file):
580
from pprint import pprint
581
print >>to_file, "Weave._weave = ",
582
pprint(self._weave, to_file)
583
print >>to_file, "Weave._parents = ",
584
pprint(self._parents, to_file)
588
def numversions(self):
589
l = len(self._parents)
590
assert l == len(self._sha1s)
595
return self.numversions()
598
def check(self, progress_bar=None):
599
# check no circular inclusions
600
for version in range(self.numversions()):
601
inclusions = list(self._parents[version])
604
if inclusions[-1] >= version:
605
raise WeaveFormatError("invalid included version %d for index %d"
606
% (inclusions[-1], version))
608
# try extracting all versions; this is a bit slow and parallel
609
# extraction could be used
610
nv = self.numversions()
611
for version in range(nv):
613
progress_bar.update('checking text', version, nv)
615
for l in self.get_iter(version):
618
expected = self._sha1s[version]
620
raise WeaveError("mismatched sha1 for version %d; "
621
"got %s, expected %s"
622
% (version, hd, expected))
624
# TODO: check insertions are properly nested, that there are
625
# no lines outside of insertion blocks, that deletions are
626
# properly paired, etc.
630
def merge(self, merge_versions):
631
"""Automerge and mark conflicts between versions.
633
This returns a sequence, each entry describing alternatives
634
for a chunk of the file. Each of the alternatives is given as
637
If there is a chunk of the file where there's no diagreement,
638
only one alternative is given.
641
# approach: find the included versions common to all the
643
raise NotImplementedError()
647
def _delta(self, included, lines):
648
"""Return changes from basis to new revision.
650
The old text for comparison is the union of included revisions.
652
This is used in inserting a new text.
654
Delta is returned as a sequence of
655
(weave1, weave2, newlines).
657
This indicates that weave1:weave2 of the old weave should be
658
replaced by the sequence of lines in newlines. Note that
659
these line numbers are positions in the total weave and don't
660
correspond to the lines in any extracted version, or even the
661
extracted union of included versions.
663
If line1=line2, this is a pure insert; if newlines=[] this is a
664
pure delete. (Similar to difflib.)
669
def plan_merge(self, ver_a, ver_b):
670
"""Return pseudo-annotation indicating how the two versions merge.
672
This is computed between versions a and b and their common
675
Weave lines present in none of them are skipped entirely.
677
inc_a = self.inclusions([ver_a])
678
inc_b = self.inclusions([ver_b])
679
inc_c = inc_a & inc_b
681
for lineno, insert, deleteset, line in self._walk():
682
if deleteset & inc_c:
683
# killed in parent; can't be in either a or b
684
# not relevant to our work
685
yield 'killed-base', line
686
elif insert in inc_c:
687
# was inserted in base
688
killed_a = bool(deleteset & inc_a)
689
killed_b = bool(deleteset & inc_b)
690
if killed_a and killed_b:
691
yield 'killed-both', line
693
yield 'killed-a', line
695
yield 'killed-b', line
697
yield 'unchanged', line
698
elif insert in inc_a:
699
if deleteset & inc_a:
700
yield 'ghost-a', line
704
elif insert in inc_b:
705
if deleteset & inc_b:
706
yield 'ghost-b', line
710
# not in either revision
711
yield 'irrelevant', line
713
yield 'unchanged', '' # terminator
717
def weave_merge(self, plan):
722
for state, line in plan:
723
if state == 'unchanged' or state == 'killed-both':
724
# resync and flush queued conflicts changes if any
725
if not lines_a and not lines_b:
727
elif ch_a and not ch_b:
729
for l in lines_a: yield l
730
elif ch_b and not ch_a:
731
for l in lines_b: yield l
732
elif lines_a == lines_b:
733
for l in lines_a: yield l
736
for l in lines_a: yield l
738
for l in lines_b: yield l
745
if state == 'unchanged':
748
elif state == 'killed-a':
751
elif state == 'killed-b':
754
elif state == 'new-a':
757
elif state == 'new-b':
761
assert state in ('irrelevant', 'ghost-a', 'ghost-b', 'killed-base',
772
"""Show the weave's table-of-contents"""
773
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
774
for i in (6, 50, 10, 10):
777
for i in range(w.numversions()):
780
parent_str = ' '.join(map(str, w._parents[i]))
781
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
785
def weave_stats(weave_file):
786
from bzrlib.progress import ProgressBar
787
from bzrlib.weavefile import read_weave
791
wf = file(weave_file, 'rb')
793
# FIXME: doesn't work on pipes
794
weave_size = wf.tell()
798
for i in range(vers):
799
pb.update('checking sizes', i, vers)
800
for line in w.get_iter(i):
805
print 'versions %9d' % vers
806
print 'weave file %9d bytes' % weave_size
807
print 'total contents %9d bytes' % total
808
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
811
print 'average size %9d bytes' % avg
812
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
816
print """bzr weave tool
818
Experimental tool for weave algorithm.
822
Create an empty weave file
823
weave get WEAVEFILE VERSION
824
Write out specified version.
825
weave check WEAVEFILE
826
Check consistency of all versions.
828
Display table of contents.
829
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
830
Add NEWTEXT, with specified parent versions.
831
weave annotate WEAVEFILE VERSION
832
Display origin of each line.
833
weave mash WEAVEFILE VERSION...
834
Display composite of all selected versions.
835
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
836
Auto-merge two versions and display conflicts.
840
% weave init foo.weave
842
% weave add foo.weave ver0 < foo.txt
845
(create updated version)
847
% weave get foo.weave 0 | diff -u - foo.txt
848
% weave add foo.weave ver1 0 < foo.txt
851
% weave get foo.weave 0 > foo.txt (create forked version)
853
% weave add foo.weave ver2 0 < foo.txt
856
% weave merge foo.weave 1 2 > foo.txt (merge them)
857
% vi foo.txt (resolve conflicts)
858
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
867
from weavefile import write_weave, read_weave
868
from bzrlib.progress import ProgressBar
883
return read_weave(file(argv[2], 'rb'))
889
# at the moment, based on everything in the file
891
parents = map(int, argv[4:])
892
lines = sys.stdin.readlines()
893
ver = w.add(name, parents, lines)
894
write_weave(w, file(argv[2], 'wb'))
895
print 'added version %r %d' % (name, ver)
898
if os.path.exists(fn):
899
raise IOError("file exists")
901
write_weave(w, file(fn, 'wb'))
902
elif cmd == 'get': # get one version
904
sys.stdout.writelines(w.get_iter(int(argv[3])))
906
elif cmd == 'mash': # get composite
908
sys.stdout.writelines(w.mash_iter(map(int, argv[3:])))
910
elif cmd == 'annotate':
912
# newline is added to all lines regardless; too hard to get
913
# reasonable formatting otherwise
915
for origin, text in w.annotate(int(argv[3])):
916
text = text.rstrip('\r\n')
918
print ' | %s' % (text)
920
print '%5d | %s' % (origin, text)
934
print '%d versions ok' % w.numversions()
936
elif cmd == 'inclusions':
938
print ' '.join(map(str, w.inclusions([int(argv[3])])))
940
elif cmd == 'parents':
942
print ' '.join(map(str, w._parents[int(argv[3])]))
944
elif cmd == 'plan-merge':
946
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
948
print '%14s | %s' % (state, line),
952
p = w.plan_merge(int(argv[3]), int(argv[4]))
953
sys.stdout.writelines(w.weave_merge(p))
955
elif cmd == 'mash-merge':
961
v1, v2 = map(int, argv[3:5])
963
basis = w.inclusions([v1]).intersection(w.inclusions([v2]))
965
base_lines = list(w.mash_iter(basis))
966
a_lines = list(w.get(v1))
967
b_lines = list(w.get(v2))
969
from bzrlib.merge3 import Merge3
970
m3 = Merge3(base_lines, a_lines, b_lines)
972
name_a = 'version %d' % v1
973
name_b = 'version %d' % v2
974
sys.stdout.writelines(m3.merge_lines(name_a=name_a, name_b=name_b))
976
raise ValueError('unknown command %r' % cmd)
980
def profile_main(argv):
981
import tempfile, hotshot, hotshot.stats
983
prof_f = tempfile.NamedTemporaryFile()
985
prof = hotshot.Profile(prof_f.name)
987
ret = prof.runcall(main, argv)
990
stats = hotshot.stats.load(prof_f.name)
992
stats.sort_stats('cumulative')
993
## XXX: Might like to write to stderr or the trace file instead but
994
## print_stats seems hardcoded to stdout
995
stats.print_stats(20)
1000
if __name__ == '__main__':
1002
if '--profile' in sys.argv:
1004
args.remove('--profile')
1005
sys.exit(profile_main(args))
1007
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
