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"""
25
# XXX: If we do weaves this way, will a merge still behave the same
26
# way if it's done in a different order? That's a pretty desirable
29
# TODO: Nothing here so far assumes the lines are really \n newlines,
30
# rather than being split up in some other way. We could accomodate
31
# binaries, perhaps by naively splitting on \n or perhaps using
32
# something like a rolling checksum.
34
# TODO: End marker for each version so we can stop reading?
36
# TODO: Check that no insertion occurs inside a deletion that was
37
# active in the version of the insertion.
39
# TODO: In addition to the SHA-1 check, perhaps have some code that
40
# checks structural constraints of the weave: ie that insertions are
41
# properly nested, that there is no text outside of an insertion, that
42
# insertions or deletions are not repeated, etc.
44
# TODO: Parallel-extract that passes back each line along with a
45
# description of which revisions include it. Nice for checking all
46
# shas or calculating stats in parallel.
48
# TODO: Using a single _extract routine and then processing the output
49
# is probably inefficient. It's simple enough that we can afford to
50
# have slight specializations for different ways its used: annotate,
51
# basis for add, get, etc.
53
# TODO: Probably the API should work only in names to hide the integer
54
# indexes from the user.
56
# TODO: Is there any potential performance win by having an add()
57
# variant that is passed a pre-cooked version of the single basis
60
# TODO: Reweave can possibly be made faster by remembering diffs
61
# where the basis and destination are unchanged.
63
# FIXME: Sometimes we will be given a parents list for a revision
64
# that includes some redundant parents (i.e. already a parent of
65
# something in the list.) We should eliminate them. This can
66
# be done fairly efficiently because the sequence numbers constrain
67
# the possible relationships.
72
from difflib import SequenceMatcher
74
from bzrlib.trace import mutter
75
from bzrlib.errors import (WeaveError, WeaveFormatError, WeaveParentMismatch,
76
WeaveRevisionNotPresent, WeaveRevisionAlreadyPresent)
77
import bzrlib.errors as errors
78
from bzrlib.tsort import topo_sort
82
"""weave - versioned text file storage.
84
A Weave manages versions of line-based text files, keeping track
85
of the originating version for each line.
87
To clients the "lines" of the file are represented as a list of strings.
88
These strings will typically have terminal newline characters, but
89
this is not required. In particular files commonly do not have a newline
90
at the end of the file.
92
Texts can be identified in either of two ways:
94
* a nonnegative index number.
96
* a version-id string.
98
Typically the index number will be valid only inside this weave and
99
the version-id is used to reference it in the larger world.
101
The weave is represented as a list mixing edit instructions and
102
literal text. Each entry in _weave can be either a string (or
103
unicode), or a tuple. If a string, it means that the given line
104
should be output in the currently active revisions.
106
If a tuple, it gives a processing instruction saying in which
107
revisions the enclosed lines are active. The tuple has the form
108
(instruction, version).
110
The instruction can be '{' or '}' for an insertion block, and '['
111
and ']' for a deletion block respectively. The version is the
112
integer version index. There is no replace operator, only deletes
113
and inserts. For '}', the end of an insertion, there is no
114
version parameter because it always closes the most recently
119
* A later version can delete lines that were introduced by any
120
number of ancestor versions; this implies that deletion
121
instructions can span insertion blocks without regard to the
122
insertion block's nesting.
124
* Similarly, deletions need not be properly nested with regard to
125
each other, because they might have been generated by
126
independent revisions.
128
* Insertions are always made by inserting a new bracketed block
129
into a single point in the previous weave. This implies they
130
can nest but not overlap, and the nesting must always have later
131
insertions on the inside.
133
* It doesn't seem very useful to have an active insertion
134
inside an inactive insertion, but it might happen.
136
* Therefore, all instructions are always"considered"; that
137
is passed onto and off the stack. An outer inactive block
138
doesn't disable an inner block.
140
* Lines are enabled if the most recent enclosing insertion is
141
active and none of the enclosing deletions are active.
143
* There is no point having a deletion directly inside its own
144
insertion; you might as well just not write it. And there
145
should be no way to get an earlier version deleting a later
149
Text of the weave; list of control instruction tuples and strings.
152
List of parents, indexed by version number.
153
It is only necessary to store the minimal set of parents for
154
each version; the parent's parents are implied.
157
List of hex SHA-1 of each version.
160
List of symbolic names for each version. Each should be unique.
163
For each name, the version number.
166
Descriptive name of this weave; typically the filename if known.
170
__slots__ = ['_weave', '_parents', '_sha1s', '_names', '_name_map',
173
def __init__(self, weave_name=None):
179
self._weave_name = weave_name
182
return "Weave(%r)" % self._weave_name
186
"""Return a deep copy of self.
188
The copy can be modified without affecting the original weave."""
190
other._weave = self._weave[:]
191
other._parents = self._parents[:]
192
other._sha1s = self._sha1s[:]
193
other._names = self._names[:]
194
other._name_map = self._name_map.copy()
195
other._weave_name = self._weave_name
198
def __eq__(self, other):
199
if not isinstance(other, Weave):
201
return self._parents == other._parents \
202
and self._weave == other._weave \
203
and self._sha1s == other._sha1s
206
def __ne__(self, other):
207
return not self.__eq__(other)
209
def __contains__(self, name):
210
return self._name_map.has_key(name)
212
def maybe_lookup(self, name_or_index):
213
"""Convert possible symbolic name to index, or pass through indexes."""
214
if isinstance(name_or_index, (int, long)):
217
return self.lookup(name_or_index)
220
def lookup(self, name):
221
"""Convert symbolic version name to index."""
223
return self._name_map[name]
225
raise WeaveRevisionNotPresent(name, self)
228
return self._names[:]
230
def iter_names(self):
231
"""Yield a list of all names in this weave."""
232
return iter(self._names)
234
def idx_to_name(self, version):
235
return self._names[version]
237
def _check_repeated_add(self, name, parents, text, sha1):
238
"""Check that a duplicated add is OK.
240
If it is, return the (old) index; otherwise raise an exception.
242
idx = self.lookup(name)
243
if sorted(self._parents[idx]) != sorted(parents) \
244
or sha1 != self._sha1s[idx]:
245
raise WeaveRevisionAlreadyPresent(name, self)
248
def add(self, name, parents, text, sha1=None):
249
"""Add a single text on top of the weave.
251
Returns the index number of the newly added version.
254
Symbolic name for this version.
255
(Typically the revision-id of the revision that added it.)
258
List or set of direct parent version numbers.
261
Sequence of lines to be added in the new version.
263
sha -- SHA-1 of the file, if known. This is trusted to be
266
from bzrlib.osutils import sha_strings
268
assert isinstance(name, basestring)
270
sha1 = sha_strings(text)
271
if name in self._name_map:
272
return self._check_repeated_add(name, parents, text, sha1)
274
parents = map(self.maybe_lookup, parents)
275
self._check_versions(parents)
276
## self._check_lines(text)
277
new_version = len(self._parents)
280
# if we abort after here the (in-memory) weave will be corrupt because only
281
# some fields are updated
282
self._parents.append(parents[:])
283
self._sha1s.append(sha1)
284
self._names.append(name)
285
self._name_map[name] = new_version
289
# special case; adding with no parents revision; can do
290
# this more quickly by just appending unconditionally.
291
# even more specially, if we're adding an empty text we
292
# need do nothing at all.
294
self._weave.append(('{', new_version))
295
self._weave.extend(text)
296
self._weave.append(('}', None))
300
if len(parents) == 1:
301
pv = list(parents)[0]
302
if sha1 == self._sha1s[pv]:
303
# special case: same as the single parent
307
ancestors = self.inclusions(parents)
311
# basis a list of (origin, lineno, line)
314
for origin, lineno, line in self._extract(ancestors):
315
basis_lineno.append(lineno)
316
basis_lines.append(line)
318
# another small special case: a merge, producing the same text
320
if text == basis_lines:
323
# add a sentinal, because we can also match against the final line
324
basis_lineno.append(len(self._weave))
326
# XXX: which line of the weave should we really consider
327
# matches the end of the file? the current code says it's the
328
# last line of the weave?
330
#print 'basis_lines:', basis_lines
331
#print 'new_lines: ', lines
333
s = SequenceMatcher(None, basis_lines, text)
335
# offset gives the number of lines that have been inserted
336
# into the weave up to the current point; if the original edit instruction
337
# says to change line A then we actually change (A+offset)
340
for tag, i1, i2, j1, j2 in s.get_opcodes():
341
# i1,i2 are given in offsets within basis_lines; we need to map them
342
# back to offsets within the entire weave
343
#print 'raw match', tag, i1, i2, j1, j2
347
i1 = basis_lineno[i1]
348
i2 = basis_lineno[i2]
350
assert 0 <= j1 <= j2 <= len(text)
352
#print tag, i1, i2, j1, j2
354
# the deletion and insertion are handled separately.
355
# first delete the region.
357
self._weave.insert(i1+offset, ('[', new_version))
358
self._weave.insert(i2+offset+1, (']', new_version))
362
# there may have been a deletion spanning up to
363
# i2; we want to insert after this region to make sure
364
# we don't destroy ourselves
366
self._weave[i:i] = ([('{', new_version)]
369
offset += 2 + (j2 - j1)
373
def add_identical(self, old_rev_id, new_rev_id, parents):
374
"""Add an identical text to old_rev_id as new_rev_id."""
375
old_lines = self.get(self.lookup(old_rev_id))
376
self.add(new_rev_id, parents, old_lines)
378
def inclusions(self, versions):
379
"""Return set of all ancestors of given version(s)."""
381
for v in xrange(max(versions), 0, -1):
383
# include all its parents
384
i.update(self._parents[v])
386
## except IndexError:
387
## raise ValueError("version %d not present in weave" % v)
390
def parents(self, version):
391
return self._parents[version]
394
def parent_names(self, version):
395
"""Return version names for parents of a version."""
396
return map(self.idx_to_name, self._parents[self.lookup(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):
458
(lineno, insert, deletes, text)
459
for each literal line.
465
lineno = 0 # line of weave, 0-based
467
for l in self._weave:
468
if isinstance(l, tuple):
481
raise WeaveFormatError('unexpected instruction %r' % v)
483
assert isinstance(l, basestring)
485
yield lineno, istack[-1], dset, l
489
raise WeaveFormatError("unclosed insertion blocks "
490
"at end of weave: %s" % istack)
492
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
495
def _extract(self, versions):
496
"""Yield annotation of lines in included set.
498
Yields a sequence of tuples (origin, lineno, text), where
499
origin is the origin version, lineno the index in the weave,
500
and text the text of the line.
502
The set typically but not necessarily corresponds to a version.
505
if not isinstance(i, int):
508
included = self.inclusions(versions)
513
lineno = 0 # line of weave, 0-based
519
WFE = WeaveFormatError
521
for l in self._weave:
522
if isinstance(l, tuple):
526
assert v not in istack
540
assert isinstance(l, basestring)
542
isactive = (not dset) and istack and (istack[-1] in included)
544
result.append((istack[-1], lineno, l))
547
raise WeaveFormatError("unclosed insertion blocks "
548
"at end of weave: %s" % istack)
550
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
555
def get_iter(self, name_or_index):
556
"""Yield lines for the specified version."""
557
incls = [self.maybe_lookup(name_or_index)]
562
# We don't have sha1 sums for multiple entries
564
for origin, lineno, line in self._extract(incls):
569
expected_sha1 = self._sha1s[index]
570
measured_sha1 = cur_sha.hexdigest()
571
if measured_sha1 != expected_sha1:
572
raise errors.WeaveInvalidChecksum(
573
'file %s, revision %s, expected: %s, measured %s'
574
% (self._weave_name, self._names[index],
575
expected_sha1, measured_sha1))
578
def get_text(self, name_or_index):
579
return ''.join(self.get_iter(name_or_index))
580
assert isinstance(version, int)
583
def get_lines(self, name_or_index):
584
return list(self.get_iter(name_or_index))
590
def get_sha1(self, name):
591
"""Get the stored sha1 sum for the given revision.
593
:param name: The name of the version to lookup
595
return self._sha1s[self.lookup(name)]
597
def mash_iter(self, included):
598
"""Return composed version of multiple included versions."""
599
included = map(self.maybe_lookup, included)
600
for origin, lineno, text in self._extract(included):
604
def dump(self, to_file):
605
from pprint import pprint
606
print >>to_file, "Weave._weave = ",
607
pprint(self._weave, to_file)
608
print >>to_file, "Weave._parents = ",
609
pprint(self._parents, to_file)
613
def numversions(self):
614
l = len(self._parents)
615
assert l == len(self._sha1s)
620
return self.numversions()
622
def check(self, progress_bar=None):
623
# check no circular inclusions
624
for version in range(self.numversions()):
625
inclusions = list(self._parents[version])
628
if inclusions[-1] >= version:
629
raise WeaveFormatError("invalid included version %d for index %d"
630
% (inclusions[-1], version))
632
# try extracting all versions; parallel extraction is used
633
nv = self.numversions()
634
sha1s = [sha.new() for i in range(nv)]
635
texts = [[] for i in range(nv)]
638
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
639
# The problem is that set membership is much more expensive
641
for p in self._parents[i]:
642
new_inc.update(inclusions[p])
644
#assert set(new_inc) == self.inclusions([i]), 'failed %s != %s' % (new_inc, self.inclusions([i]))
645
inclusions.append(new_inc)
647
nlines = len(self._weave)
649
update_text = 'checking weave'
651
short_name = os.path.basename(self._weave_name)
652
update_text = 'checking %s' % (short_name,)
653
update_text = update_text[:25]
655
for lineno, insert, deleteset, line in self._walk():
657
progress_bar.update(update_text, lineno, nlines)
659
for j, j_inc in enumerate(inclusions):
660
# The active inclusion must be an ancestor,
661
# and no ancestors must have deleted this line,
662
# because we don't support resurrection.
663
if (insert in j_inc) and not (deleteset & j_inc):
664
sha1s[j].update(line)
666
for version in range(nv):
667
hd = sha1s[version].hexdigest()
668
expected = self._sha1s[version]
670
raise errors.WeaveInvalidChecksum(
671
"mismatched sha1 for version %s: "
672
"got %s, expected %s"
673
% (self._names[version], hd, expected))
675
# TODO: check insertions are properly nested, that there are
676
# no lines outside of insertion blocks, that deletions are
677
# properly paired, etc.
679
def _delta(self, included, lines):
680
"""Return changes from basis to new revision.
682
The old text for comparison is the union of included revisions.
684
This is used in inserting a new text.
686
Delta is returned as a sequence of
687
(weave1, weave2, newlines).
689
This indicates that weave1:weave2 of the old weave should be
690
replaced by the sequence of lines in newlines. Note that
691
these line numbers are positions in the total weave and don't
692
correspond to the lines in any extracted version, or even the
693
extracted union of included versions.
695
If line1=line2, this is a pure insert; if newlines=[] this is a
696
pure delete. (Similar to difflib.)
698
raise NotImplementedError()
701
def plan_merge(self, ver_a, ver_b):
702
"""Return pseudo-annotation indicating how the two versions merge.
704
This is computed between versions a and b and their common
707
Weave lines present in none of them are skipped entirely.
709
inc_a = self.inclusions([ver_a])
710
inc_b = self.inclusions([ver_b])
711
inc_c = inc_a & inc_b
713
for lineno, insert, deleteset, line in self._walk():
714
if deleteset & inc_c:
715
# killed in parent; can't be in either a or b
716
# not relevant to our work
717
yield 'killed-base', line
718
elif insert in inc_c:
719
# was inserted in base
720
killed_a = bool(deleteset & inc_a)
721
killed_b = bool(deleteset & inc_b)
722
if killed_a and killed_b:
723
yield 'killed-both', line
725
yield 'killed-a', line
727
yield 'killed-b', line
729
yield 'unchanged', line
730
elif insert in inc_a:
731
if deleteset & inc_a:
732
yield 'ghost-a', line
736
elif insert in inc_b:
737
if deleteset & inc_b:
738
yield 'ghost-b', line
742
# not in either revision
743
yield 'irrelevant', line
745
yield 'unchanged', '' # terminator
749
def weave_merge(self, plan):
753
# TODO: Return a structured form of the conflicts (e.g. 2-tuples for
754
# conflicted regions), rather than just inserting the markers.
756
# TODO: Show some version information (e.g. author, date) on
757
# conflicted regions.
758
for state, line in plan:
759
if state == 'unchanged' or state == 'killed-both':
760
# resync and flush queued conflicts changes if any
761
if not lines_a and not lines_b:
763
elif ch_a and not ch_b:
765
for l in lines_a: yield l
766
elif ch_b and not ch_a:
767
for l in lines_b: yield l
768
elif lines_a == lines_b:
769
for l in lines_a: yield l
772
for l in lines_a: yield l
774
for l in lines_b: yield l
781
if state == 'unchanged':
784
elif state == 'killed-a':
787
elif state == 'killed-b':
790
elif state == 'new-a':
793
elif state == 'new-b':
797
assert state in ('irrelevant', 'ghost-a', 'ghost-b', 'killed-base',
802
def join(self, other, pb=None, msg=None):
804
"""Integrate versions from other into this weave.
806
The resulting weave contains all the history of both weaves;
807
any version you could retrieve from either self or other can be
808
retrieved from self after this call.
810
It is illegal for the two weaves to contain different values
811
or different parents for any version. See also reweave().
813
:param other: The other weave to pull into this one
814
:param pb: An optional progress bar
815
:param msg: An optional message to display for progress
817
if other.numversions() == 0:
818
return # nothing to update, easy
819
# two loops so that we do not change ourselves before verifying it
821
# work through in index order to make sure we get all dependencies
824
for other_idx, name in enumerate(other._names):
825
self._check_version_consistent(other, other_idx, name)
826
sha1 = other._sha1s[other_idx]
830
if name in self._name_map:
831
idx = self.lookup(name)
832
n1 = set(map(other.idx_to_name, other._parents[other_idx]))
833
n2 = set(map(self.idx_to_name, self._parents[idx]))
834
if sha1 == self._sha1s[idx] and n1 == n2:
837
names_to_join.append((other_idx, name))
844
for other_idx, name in names_to_join:
845
# TODO: If all the parents of the other version are already
846
# present then we can avoid some work by just taking the delta
847
# and adjusting the offsets.
848
new_parents = self._imported_parents(other, other_idx)
849
sha1 = other._sha1s[other_idx]
854
pb.update(msg, merged, len(names_to_join))
856
lines = other.get_lines(other_idx)
857
self.add(name, new_parents, lines, sha1)
859
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
860
merged, processed, self._weave_name, time.time( )-time0))
862
def _imported_parents(self, other, other_idx):
863
"""Return list of parents in self corresponding to indexes in other."""
865
for parent_idx in other._parents[other_idx]:
866
parent_name = other._names[parent_idx]
867
if parent_name not in self._names:
868
# should not be possible
869
raise WeaveError("missing parent {%s} of {%s} in %r"
870
% (parent_name, other._name_map[other_idx], self))
871
new_parents.append(self._name_map[parent_name])
874
def _check_version_consistent(self, other, other_idx, name):
875
"""Check if a version in consistent in this and other.
877
To be consistent it must have:
880
* the same direct parents (by name, not index, and disregarding
883
If present & correct return True;
884
if not present in self return False;
885
if inconsistent raise error."""
886
this_idx = self._name_map.get(name, -1)
888
if self._sha1s[this_idx] != other._sha1s[other_idx]:
889
raise WeaveError("inconsistent texts for version {%s} "
890
"when joining weaves"
892
self_parents = self._parents[this_idx]
893
other_parents = other._parents[other_idx]
894
n1 = set([self._names[i] for i in self_parents])
895
n2 = set([other._names[i] for i in other_parents])
897
raise WeaveParentMismatch("inconsistent parents "
898
"for version {%s}: %s vs %s" % (name, n1, n2))
904
def reweave(self, other, pb=None, msg=None):
905
"""Reweave self with other.
907
:param other: The other weave to merge
908
:param pb: An optional progress bar, indicating how far done we are
909
:param msg: An optional message for the progress
911
new_weave = reweave(self, other, pb=pb, msg=msg)
912
for attr in self.__slots__:
913
setattr(self, attr, getattr(new_weave, attr))
916
def reweave(wa, wb, pb=None, msg=None):
917
"""Combine two weaves and return the result.
919
This works even if a revision R has different parents in
920
wa and wb. In the resulting weave all the parents are given.
922
This is done by just building up a new weave, maintaining ordering
923
of the versions in the two inputs. More efficient approaches
924
might be possible but it should only be necessary to do
925
this operation rarely, when a new previously ghost version is
928
:param pb: An optional progress bar, indicating how far done we are
929
:param msg: An optional message for the progress
933
queue_a = range(wa.numversions())
934
queue_b = range(wb.numversions())
935
# first determine combined parents of all versions
936
# map from version name -> all parent names
937
combined_parents = _reweave_parent_graphs(wa, wb)
938
mutter("combined parents: %r", combined_parents)
939
order = topo_sort(combined_parents.iteritems())
940
mutter("order to reweave: %r", order)
945
for idx, name in enumerate(order):
947
pb.update(msg, idx, len(order))
948
if name in wa._name_map:
949
lines = wa.get_lines(name)
950
if name in wb._name_map:
951
lines_b = wb.get_lines(name)
953
mutter('Weaves differ on content. rev_id {%s}', name)
954
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
956
lines = list(difflib.unified_diff(lines, lines_b,
957
wa._weave_name, wb._weave_name))
958
mutter('lines:\n%s', ''.join(lines))
959
raise errors.WeaveTextDiffers(name, wa, wb)
961
lines = wb.get_lines(name)
962
wr.add(name, combined_parents[name], lines)
966
def _reweave_parent_graphs(wa, wb):
967
"""Return combined parent ancestry for two weaves.
969
Returned as a list of (version_name, set(parent_names))"""
971
for weave in [wa, wb]:
972
for idx, name in enumerate(weave._names):
973
p = combined.setdefault(name, set())
974
p.update(map(weave.idx_to_name, weave._parents[idx]))
979
"""Show the weave's table-of-contents"""
980
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
981
for i in (6, 50, 10, 10):
984
for i in range(w.numversions()):
987
parent_str = ' '.join(map(str, w._parents[i]))
988
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
992
def weave_stats(weave_file, pb):
993
from bzrlib.weavefile import read_weave
995
wf = file(weave_file, 'rb')
997
# FIXME: doesn't work on pipes
998
weave_size = wf.tell()
1002
for i in range(vers):
1003
pb.update('checking sizes', i, vers)
1004
for origin, lineno, line in w._extract([i]):
1009
print 'versions %9d' % vers
1010
print 'weave file %9d bytes' % weave_size
1011
print 'total contents %9d bytes' % total
1012
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1015
print 'average size %9d bytes' % avg
1016
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1020
print """bzr weave tool
1022
Experimental tool for weave algorithm.
1025
weave init WEAVEFILE
1026
Create an empty weave file
1027
weave get WEAVEFILE VERSION
1028
Write out specified version.
1029
weave check WEAVEFILE
1030
Check consistency of all versions.
1032
Display table of contents.
1033
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1034
Add NEWTEXT, with specified parent versions.
1035
weave annotate WEAVEFILE VERSION
1036
Display origin of each line.
1037
weave mash WEAVEFILE VERSION...
1038
Display composite of all selected versions.
1039
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1040
Auto-merge two versions and display conflicts.
1041
weave diff WEAVEFILE VERSION1 VERSION2
1042
Show differences between two versions.
1046
% weave init foo.weave
1048
% weave add foo.weave ver0 < foo.txt
1051
(create updated version)
1053
% weave get foo.weave 0 | diff -u - foo.txt
1054
% weave add foo.weave ver1 0 < foo.txt
1057
% weave get foo.weave 0 > foo.txt (create forked version)
1059
% weave add foo.weave ver2 0 < foo.txt
1062
% weave merge foo.weave 1 2 > foo.txt (merge them)
1063
% vi foo.txt (resolve conflicts)
1064
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1076
# in case we're run directly from the subdirectory
1077
sys.path.append('..')
1079
from bzrlib.weavefile import write_weave, read_weave
1080
from bzrlib.progress import ProgressBar
1095
return read_weave(file(argv[2], 'rb'))
1101
# at the moment, based on everything in the file
1103
parents = map(int, argv[4:])
1104
lines = sys.stdin.readlines()
1105
ver = w.add(name, parents, lines)
1106
write_weave(w, file(argv[2], 'wb'))
1107
print 'added version %r %d' % (name, ver)
1110
if os.path.exists(fn):
1111
raise IOError("file exists")
1113
write_weave(w, file(fn, 'wb'))
1114
elif cmd == 'get': # get one version
1116
sys.stdout.writelines(w.get_iter(int(argv[3])))
1118
elif cmd == 'mash': # get composite
1120
sys.stdout.writelines(w.mash_iter(map(int, argv[3:])))
1123
from difflib import unified_diff
1126
v1, v2 = map(int, argv[3:5])
1129
diff_gen = unified_diff(lines1, lines2,
1130
'%s version %d' % (fn, v1),
1131
'%s version %d' % (fn, v2))
1132
sys.stdout.writelines(diff_gen)
1134
elif cmd == 'annotate':
1136
# newline is added to all lines regardless; too hard to get
1137
# reasonable formatting otherwise
1139
for origin, text in w.annotate(int(argv[3])):
1140
text = text.rstrip('\r\n')
1142
print ' | %s' % (text)
1144
print '%5d | %s' % (origin, text)
1150
elif cmd == 'stats':
1151
weave_stats(argv[2], ProgressBar())
1153
elif cmd == 'check':
1158
print '%d versions ok' % w.numversions()
1160
elif cmd == 'inclusions':
1162
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1164
elif cmd == 'parents':
1166
print ' '.join(map(str, w._parents[int(argv[3])]))
1168
elif cmd == 'plan-merge':
1170
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1172
print '%14s | %s' % (state, line),
1174
elif cmd == 'merge':
1176
p = w.plan_merge(int(argv[3]), int(argv[4]))
1177
sys.stdout.writelines(w.weave_merge(p))
1179
elif cmd == 'mash-merge':
1185
v1, v2 = map(int, argv[3:5])
1187
basis = w.inclusions([v1]).intersection(w.inclusions([v2]))
1189
base_lines = list(w.mash_iter(basis))
1190
a_lines = list(w.get(v1))
1191
b_lines = list(w.get(v2))
1193
from bzrlib.merge3 import Merge3
1194
m3 = Merge3(base_lines, a_lines, b_lines)
1196
name_a = 'version %d' % v1
1197
name_b = 'version %d' % v2
1198
sys.stdout.writelines(m3.merge_lines(name_a=name_a, name_b=name_b))
1200
raise ValueError('unknown command %r' % cmd)
1204
def profile_main(argv):
1205
import tempfile, hotshot, hotshot.stats
1207
prof_f = tempfile.NamedTemporaryFile()
1209
prof = hotshot.Profile(prof_f.name)
1211
ret = prof.runcall(main, argv)
1214
stats = hotshot.stats.load(prof_f.name)
1216
stats.sort_stats('cumulative')
1217
## XXX: Might like to write to stderr or the trace file instead but
1218
## print_stats seems hardcoded to stdout
1219
stats.print_stats(20)
1224
def lsprofile_main(argv):
1225
from bzrlib.lsprof import profile
1226
ret,stats = profile(main, argv)
1232
if __name__ == '__main__':
1234
if '--profile' in sys.argv:
1236
args.remove('--profile')
1237
sys.exit(profile_main(args))
1238
elif '--lsprof' in sys.argv:
1240
args.remove('--lsprof')
1241
sys.exit(lsprofile_main(args))
1243
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
