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
# TODO: Perhaps have copy method for Weave instances?
26
# XXX: If we do weaves this way, will a merge still behave the same
27
# way if it's done in a different order? That's a pretty desirable
30
# TODO: How to write these to disk? One option is cPickle, which
31
# would be fast but less friendly to C, and perhaps not portable. Another is
33
# TODO: Nothing here so far assumes the lines are really \n newlines,
34
# rather than being split up in some other way. We could accomodate
35
# binaries, perhaps by naively splitting on \n or perhaps using
36
# something like a rolling checksum.
38
# TODO: Perhaps track SHA-1 in the header for protection? This would
39
# be redundant with it being stored in the inventory, but perhaps
42
# TODO: Track version names as well as indexes.
44
# TODO: Probably do transitive expansion when specifying parents?
46
# TODO: Separate out some code to read and write weaves.
48
# TODO: End marker for each version so we can stop reading?
50
# TODO: Check that no insertion occurs inside a deletion that was
51
# active in the version of the insertion.
53
# TODO: Perhaps a special slower check() method that verifies more
54
# nesting constraints and the MD5 of each version?
62
from sets import Set, ImmutableSet
64
frozenset = ImmutableSet
68
class WeaveError(Exception):
69
"""Exception in processing weave"""
72
class WeaveFormatError(WeaveError):
73
"""Weave invariant violated"""
77
"""weave - versioned text file storage.
79
A Weave manages versions of line-based text files, keeping track
80
of the originating version for each line.
82
To clients the "lines" of the file are represented as a list of strings.
83
These strings will typically have terminal newline characters, but
84
this is not required. In particular files commonly do not have a newline
85
at the end of the file.
87
Texts can be identified in either of two ways:
89
* a nonnegative index number.
91
* a version-id string.
93
Typically the index number will be valid only inside this weave and
94
the version-id is used to reference it in the larger world.
96
The weave is represented as a list mixing edit instructions and
97
literal text. Each entry in _l can be either a string (or
98
unicode), or a tuple. If a string, it means that the given line
99
should be output in the currently active revisions.
101
If a tuple, it gives a processing instruction saying in which
102
revisions the enclosed lines are active. The tuple has the form
103
(instruction, version).
105
The instruction can be '{' or '}' for an insertion block, and '['
106
and ']' for a deletion block respectively. The version is the
107
integer version index. There is no replace operator, only deletes
112
* A later version can delete lines that were introduced by any
113
number of ancestor versions; this implies that deletion
114
instructions can span insertion blocks without regard to the
115
insertion block's nesting.
117
* Similarly, deletions need not be properly nested with regard to
118
each other, because they might have been generated by
119
independent revisions.
121
* Insertions are always made by inserting a new bracketed block
122
into a single point in the previous weave. This implies they
123
can nest but not overlap, and the nesting must always have later
124
insertions on the inside.
126
* It doesn't seem very useful to have an active insertion
127
inside an inactive insertion, but it might happen.
129
* Therefore, all instructions are always"considered"; that
130
is passed onto and off the stack. An outer inactive block
131
doesn't disable an inner block.
133
* Lines are enabled if the most recent enclosing insertion is
134
active and none of the enclosing deletions are active.
136
* There is no point having a deletion directly inside its own
137
insertion; you might as well just not write it. And there
138
should be no way to get an earlier version deleting a later
145
List of parents, indexed by version number.
146
It is only necessary to store the minimal set of parents for
147
each version; the parent's parents are implied.
150
Full set of inclusions for each revision.
153
List of hex SHA-1 of each version, or None if not recorded.
162
def __eq__(self, other):
163
if not isinstance(other, Weave):
165
return self._v == other._v \
166
and self._l == other._l
169
def __ne__(self, other):
170
return not self.__eq__(other)
173
def add(self, parents, text):
174
"""Add a single text on top of the weave.
176
Returns the index number of the newly added version.
179
List or set of direct parent version numbers.
182
Sequence of lines to be added in the new version."""
183
## self._check_versions(parents)
184
## self._check_lines(text)
195
ancestors = self.inclusions(parents)
196
delta = self._delta(ancestors, text)
198
# offset gives the number of lines that have been inserted
199
# into the weave up to the current point; if the original edit instruction
200
# says to change line A then we actually change (A+offset)
203
for i1, i2, newlines in delta:
206
assert i2 <= len(self._l)
208
# the deletion and insertion are handled separately.
209
# first delete the region.
211
self._l.insert(i1+offset, ('[', idx))
212
self._l.insert(i2+offset+1, (']', idx))
217
# there may have been a deletion spanning up to
218
# i2; we want to insert after this region to make sure
219
# we don't destroy ourselves
221
self._l[i:i] = [('{', idx)] \
224
offset += 2 + len(newlines)
226
self._addversion(parents)
228
# special case; adding with no parents revision; can do this
229
# more quickly by just appending unconditionally
230
self._l.append(('{', idx))
232
self._l.append(('}', idx))
234
self._addversion(None)
236
self._sha1s.append(sha1)
241
def _expand(self, version):
242
if version in self._i:
243
return self._i[version]
246
for pv in self._v[version]:
247
i.update(self._expand(pv))
252
def inclusions(self, versions):
253
"""Expand out everything included by versions."""
257
i.update(self._expand(v))
259
raise ValueError("version %d not present in weave" % v)
263
def minimal_parents(self, version):
264
"""Find the minimal set of parents for the version."""
265
included = self._v[version]
279
gotit.update(self._v[pv])
281
assert mininc[0] >= 0
282
assert mininc[-1] < version
286
def _addversion(self, parents):
288
self._v.append(parents)
290
self._v.append(frozenset())
293
def _check_lines(self, text):
294
if not isinstance(text, list):
295
raise ValueError("text should be a list, not %s" % type(text))
298
if not isinstance(l, basestring):
299
raise ValueError("text line should be a string or unicode, not %s"
304
def _check_versions(self, indexes):
305
"""Check everything in the sequence of indexes is valid"""
310
raise IndexError("invalid version number %r" % i)
313
def annotate(self, index):
314
return list(self.annotate_iter(index))
317
def annotate_iter(self, version):
318
"""Yield list of (index-id, line) pairs for the specified version.
320
The index indicates when the line originated in the weave."""
321
included = self.inclusions([version])
322
for origin, lineno, text in self._extract(self.inclusions(included)):
326
def _extract(self, included):
327
"""Yield annotation of lines in included set.
329
Yields a sequence of tuples (origin, lineno, text), where
330
origin is the origin version, lineno the index in the weave,
331
and text the text of the line.
333
The set typically but not necessarily corresponds to a version.
339
lineno = 0 # line of weave, 0-based
343
WFE = WeaveFormatError
346
if isinstance(l, tuple):
349
assert v not in istack
352
isactive = (v in included)
356
isactive = (not dset) and (istack and istack[-1] in included)
367
isactive = (not dset) and (istack and istack[-1] in included)
369
assert isinstance(l, basestring)
371
yield istack[-1], lineno, l
375
raise WFE("unclosed insertion blocks at end of weave",
378
raise WFE("unclosed deletion blocks at end of weave",
382
def get_iter(self, version):
383
"""Yield lines for the specified version."""
384
for origin, lineno, line in self._extract(self.inclusions([version])):
388
def get(self, index):
389
return list(self.get_iter(index))
392
def mash_iter(self, included):
393
"""Return composed version of multiple included versions."""
394
included = frozenset(included)
395
for origin, lineno, text in self._extract(self.inclusions(included)):
399
def dump(self, to_file):
400
from pprint import pprint
401
print >>to_file, "Weave._l = ",
402
pprint(self._l, to_file)
403
print >>to_file, "Weave._v = ",
404
pprint(self._v, to_file)
408
def numversions(self):
410
assert l == len(self._sha1s)
415
# check no circular inclusions
416
for version in range(self.numversions()):
417
inclusions = list(self._v[version])
420
if inclusions[-1] >= version:
421
raise WeaveFormatError("invalid included version %d for index %d"
422
% (inclusions[-1], version))
424
# try extracting all versions; this is a bit slow and parallel
425
# extraction could be used
427
for version in range(self.numversions()):
429
for l in self.get_iter(version):
432
expected = self._sha1s[version]
434
raise WeaveError("mismatched sha1 for version %d; "
435
"got %s, expected %s"
436
% (version, hd, expected))
438
# TODO: check insertions are properly nested, that there are
439
# no lines outside of insertion blocks, that deletions are
440
# properly paired, etc.
444
def merge(self, merge_versions):
445
"""Automerge and mark conflicts between versions.
447
This returns a sequence, each entry describing alternatives
448
for a chunk of the file. Each of the alternatives is given as
451
If there is a chunk of the file where there's no diagreement,
452
only one alternative is given.
455
# approach: find the included versions common to all the
457
raise NotImplementedError()
461
def _delta(self, included, lines):
462
"""Return changes from basis to new revision.
464
The old text for comparison is the union of included revisions.
466
This is used in inserting a new text.
468
Delta is returned as a sequence of
469
(weave1, weave2, newlines).
471
This indicates that weave1:weave2 of the old weave should be
472
replaced by the sequence of lines in newlines. Note that
473
these line numbers are positions in the total weave and don't
474
correspond to the lines in any extracted version, or even the
475
extracted union of included versions.
477
If line1=line2, this is a pure insert; if newlines=[] this is a
478
pure delete. (Similar to difflib.)
480
# basis a list of (origin, lineno, line)
483
for origin, lineno, line in self._extract(self.inclusions(included)):
484
basis_lineno.append(lineno)
485
basis_lines.append(line)
487
# add a sentinal, because we can also match against the final line
488
basis_lineno.append(len(self._l))
490
# XXX: which line of the weave should we really consider
491
# matches the end of the file? the current code says it's the
492
# last line of the weave?
494
from difflib import SequenceMatcher
495
s = SequenceMatcher(None, basis_lines, lines)
497
# TODO: Perhaps return line numbers from composed weave as well?
499
for tag, i1, i2, j1, j2 in s.get_opcodes():
500
##print tag, i1, i2, j1, j2
505
# i1,i2 are given in offsets within basis_lines; we need to map them
506
# back to offsets within the entire weave
507
real_i1 = basis_lineno[i1]
508
real_i2 = basis_lineno[i2]
512
assert j2 <= len(lines)
514
yield real_i1, real_i2, lines[j1:j2]
518
def weave_info(filename, out):
519
"""Show some text information about the weave."""
520
from weavefile import read_weave
521
wf = file(filename, 'rb')
523
# FIXME: doesn't work on pipes
524
weave_size = wf.tell()
525
print >>out, "weave file size %d bytes" % weave_size
526
print >>out, "weave contains %d versions" % len(w._v)
529
print '%6s %6s %8s %40s %20s' % ('ver', 'lines', 'bytes', 'sha1', 'parents')
530
for i in (6, 6, 8, 40, 20):
533
for i in range(len(w._v)):
536
bytes = sum((len(a) for a in text))
538
print '%6d %6d %8d %40s' % (i, lines, bytes, sha1),
544
print >>out, "versions total %d bytes" % total
545
print >>out, "compression ratio %.3f" % (float(total)/float(weave_size))
549
print """bzr weave tool
551
Experimental tool for weave algorithm.
555
Create an empty weave file
556
weave get WEAVEFILE VERSION
557
Write out specified version.
558
weave check WEAVEFILE
559
Check consistency of all versions.
561
Display table of contents.
562
weave add WEAVEFILE [BASE...] < NEWTEXT
563
Add NEWTEXT, with specified parent versions.
564
weave annotate WEAVEFILE VERSION
565
Display origin of each line.
566
weave mash WEAVEFILE VERSION...
567
Display composite of all selected versions.
568
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
569
Auto-merge two versions and display conflicts.
573
% weave init foo.weave
575
% weave add foo.weave < foo.txt
578
(create updated version)
580
% weave get foo.weave 0 | diff -u - foo.txt
581
% weave add foo.weave 0 < foo.txt
584
% weave get foo.weave 0 > foo.txt (create forked version)
586
% weave add foo.weave 0 < foo.txt
589
% weave merge foo.weave 1 2 > foo.txt (merge them)
590
% vi foo.txt (resolve conflicts)
591
% weave add foo.weave 1 2 < foo.txt (commit merged version)
600
from weavefile import write_weave, read_weave
604
return read_weave(file(argv[2], 'rb'))
610
# at the moment, based on everything in the file
611
parents = map(int, argv[3:])
612
lines = sys.stdin.readlines()
613
ver = w.add(parents, lines)
614
write_weave(w, file(argv[2], 'wb'))
615
print 'added version %d' % ver
618
if os.path.exists(fn):
619
raise IOError("file exists")
621
write_weave(w, file(fn, 'wb'))
622
elif cmd == 'get': # get one version
624
sys.stdout.writelines(w.get_iter(int(argv[3])))
626
elif cmd == 'mash': # get composite
628
sys.stdout.writelines(w.mash_iter(map(int, argv[3:])))
630
elif cmd == 'annotate':
632
# newline is added to all lines regardless; too hard to get
633
# reasonable formatting otherwise
635
for origin, text in w.annotate(int(argv[3])):
636
text = text.rstrip('\r\n')
638
print ' | %s' % (text)
640
print '%5d | %s' % (origin, text)
644
weave_info(argv[2], sys.stdout)
650
elif cmd == 'inclusions':
652
print ' '.join(map(str, w.inclusions([int(argv[3])])))
654
elif cmd == 'parents':
656
print ' '.join(map(str, w._v[int(argv[3])]))
664
v1, v2 = map(int, argv[3:5])
666
basis = w.inclusions([v1]).intersection(w.inclusions([v2]))
668
base_lines = list(w.mash_iter(basis))
669
a_lines = list(w.get(v1))
670
b_lines = list(w.get(v2))
672
from bzrlib.merge3 import Merge3
673
m3 = Merge3(base_lines, a_lines, b_lines)
675
name_a = 'version %d' % v1
676
name_b = 'version %d' % v2
677
sys.stdout.writelines(m3.merge_lines(name_a=name_a, name_b=name_b))
679
raise ValueError('unknown command %r' % cmd)
682
if __name__ == '__main__':
684
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
