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
# XXX: FIXME implement a succeed-or-fail of the rest of this routine.
283
# - Robert Collins 20060226
284
self._parents.append(parents[:])
285
self._sha1s.append(sha1)
286
self._names.append(name)
287
self._name_map[name] = new_version
291
# special case; adding with no parents revision; can do
292
# this more quickly by just appending unconditionally.
293
# even more specially, if we're adding an empty text we
294
# need do nothing at all.
296
self._weave.append(('{', new_version))
297
self._weave.extend(text)
298
self._weave.append(('}', None))
302
if len(parents) == 1:
303
pv = list(parents)[0]
304
if sha1 == self._sha1s[pv]:
305
# special case: same as the single parent
309
ancestors = self.inclusions(parents)
313
# basis a list of (origin, lineno, line)
316
for origin, lineno, line in self._extract(ancestors):
317
basis_lineno.append(lineno)
318
basis_lines.append(line)
320
# another small special case: a merge, producing the same text
322
if text == basis_lines:
325
# add a sentinal, because we can also match against the final line
326
basis_lineno.append(len(self._weave))
328
# XXX: which line of the weave should we really consider
329
# matches the end of the file? the current code says it's the
330
# last line of the weave?
332
#print 'basis_lines:', basis_lines
333
#print 'new_lines: ', lines
335
s = SequenceMatcher(None, basis_lines, text)
337
# offset gives the number of lines that have been inserted
338
# into the weave up to the current point; if the original edit instruction
339
# says to change line A then we actually change (A+offset)
342
for tag, i1, i2, j1, j2 in s.get_opcodes():
343
# i1,i2 are given in offsets within basis_lines; we need to map them
344
# back to offsets within the entire weave
345
#print 'raw match', tag, i1, i2, j1, j2
349
i1 = basis_lineno[i1]
350
i2 = basis_lineno[i2]
352
assert 0 <= j1 <= j2 <= len(text)
354
#print tag, i1, i2, j1, j2
356
# the deletion and insertion are handled separately.
357
# first delete the region.
359
self._weave.insert(i1+offset, ('[', new_version))
360
self._weave.insert(i2+offset+1, (']', new_version))
364
# there may have been a deletion spanning up to
365
# i2; we want to insert after this region to make sure
366
# we don't destroy ourselves
368
self._weave[i:i] = ([('{', new_version)]
371
offset += 2 + (j2 - j1)
375
def add_identical(self, old_rev_id, new_rev_id, parents):
376
"""Add an identical text to old_rev_id as new_rev_id."""
377
old_lines = self.get(self.lookup(old_rev_id))
378
self.add(new_rev_id, parents, old_lines)
380
def inclusions(self, versions):
381
"""Return set of all ancestors of given version(s)."""
383
for v in xrange(max(versions), 0, -1):
385
# include all its parents
386
i.update(self._parents[v])
388
## except IndexError:
389
## raise ValueError("version %d not present in weave" % v)
392
def parents(self, version):
393
return self._parents[version]
396
def parent_names(self, version):
397
"""Return version names for parents of a version."""
398
return map(self.idx_to_name, self._parents[self.lookup(version)])
401
def minimal_parents(self, version):
402
"""Find the minimal set of parents for the version."""
403
included = self._parents[version]
408
li.sort(reverse=True)
416
gotit.update(self.inclusions(pv))
418
assert mininc[0] >= 0
419
assert mininc[-1] < version
424
def _check_lines(self, text):
425
if not isinstance(text, list):
426
raise ValueError("text should be a list, not %s" % type(text))
429
if not isinstance(l, basestring):
430
raise ValueError("text line should be a string or unicode, not %s"
435
def _check_versions(self, indexes):
436
"""Check everything in the sequence of indexes is valid"""
441
raise IndexError("invalid version number %r" % i)
444
def annotate(self, name_or_index):
445
return list(self.annotate_iter(name_or_index))
448
def annotate_iter(self, name_or_index):
449
"""Yield list of (index-id, line) pairs for the specified version.
451
The index indicates when the line originated in the weave."""
452
incls = [self.maybe_lookup(name_or_index)]
453
for origin, lineno, text in self._extract(incls):
460
(lineno, insert, deletes, text)
461
for each literal line.
467
lineno = 0 # line of weave, 0-based
469
for l in self._weave:
470
if isinstance(l, tuple):
483
raise WeaveFormatError('unexpected instruction %r' % v)
485
assert isinstance(l, basestring)
487
yield lineno, istack[-1], dset, l
491
raise WeaveFormatError("unclosed insertion blocks "
492
"at end of weave: %s" % istack)
494
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
497
def _extract(self, versions):
498
"""Yield annotation of lines in included set.
500
Yields a sequence of tuples (origin, lineno, text), where
501
origin is the origin version, lineno the index in the weave,
502
and text the text of the line.
504
The set typically but not necessarily corresponds to a version.
507
if not isinstance(i, int):
510
included = self.inclusions(versions)
515
lineno = 0 # line of weave, 0-based
521
WFE = WeaveFormatError
523
for l in self._weave:
524
if isinstance(l, tuple):
528
assert v not in istack
542
assert isinstance(l, basestring)
544
isactive = (not dset) and istack and (istack[-1] in included)
546
result.append((istack[-1], lineno, l))
549
raise WeaveFormatError("unclosed insertion blocks "
550
"at end of weave: %s" % istack)
552
raise WeaveFormatError("unclosed deletion blocks at end of weave: %s"
557
def get_iter(self, name_or_index):
558
"""Yield lines for the specified version."""
559
incls = [self.maybe_lookup(name_or_index)]
564
# We don't have sha1 sums for multiple entries
566
for origin, lineno, line in self._extract(incls):
571
expected_sha1 = self._sha1s[index]
572
measured_sha1 = cur_sha.hexdigest()
573
if measured_sha1 != expected_sha1:
574
raise errors.WeaveInvalidChecksum(
575
'file %s, revision %s, expected: %s, measured %s'
576
% (self._weave_name, self._names[index],
577
expected_sha1, measured_sha1))
580
def get_text(self, name_or_index):
581
return ''.join(self.get_iter(name_or_index))
582
assert isinstance(version, int)
585
def get_lines(self, name_or_index):
586
return list(self.get_iter(name_or_index))
592
def get_sha1(self, name):
593
"""Get the stored sha1 sum for the given revision.
595
:param name: The name of the version to lookup
597
return self._sha1s[self.lookup(name)]
599
def mash_iter(self, included):
600
"""Return composed version of multiple included versions."""
601
included = map(self.maybe_lookup, included)
602
for origin, lineno, text in self._extract(included):
606
def dump(self, to_file):
607
from pprint import pprint
608
print >>to_file, "Weave._weave = ",
609
pprint(self._weave, to_file)
610
print >>to_file, "Weave._parents = ",
611
pprint(self._parents, to_file)
615
def numversions(self):
616
l = len(self._parents)
617
assert l == len(self._sha1s)
622
return self.numversions()
624
def check(self, progress_bar=None):
625
# check no circular inclusions
626
for version in range(self.numversions()):
627
inclusions = list(self._parents[version])
630
if inclusions[-1] >= version:
631
raise WeaveFormatError("invalid included version %d for index %d"
632
% (inclusions[-1], version))
634
# try extracting all versions; parallel extraction is used
635
nv = self.numversions()
636
sha1s = [sha.new() for i in range(nv)]
637
texts = [[] for i in range(nv)]
640
# For creating the ancestry, IntSet is much faster (3.7s vs 0.17s)
641
# The problem is that set membership is much more expensive
643
for p in self._parents[i]:
644
new_inc.update(inclusions[p])
646
#assert set(new_inc) == self.inclusions([i]), 'failed %s != %s' % (new_inc, self.inclusions([i]))
647
inclusions.append(new_inc)
649
nlines = len(self._weave)
651
update_text = 'checking weave'
653
short_name = os.path.basename(self._weave_name)
654
update_text = 'checking %s' % (short_name,)
655
update_text = update_text[:25]
657
for lineno, insert, deleteset, line in self._walk():
659
progress_bar.update(update_text, lineno, nlines)
661
for j, j_inc in enumerate(inclusions):
662
# The active inclusion must be an ancestor,
663
# and no ancestors must have deleted this line,
664
# because we don't support resurrection.
665
if (insert in j_inc) and not (deleteset & j_inc):
666
sha1s[j].update(line)
668
for version in range(nv):
669
hd = sha1s[version].hexdigest()
670
expected = self._sha1s[version]
672
raise errors.WeaveInvalidChecksum(
673
"mismatched sha1 for version %s: "
674
"got %s, expected %s"
675
% (self._names[version], hd, expected))
677
# TODO: check insertions are properly nested, that there are
678
# no lines outside of insertion blocks, that deletions are
679
# properly paired, etc.
681
def _delta(self, included, lines):
682
"""Return changes from basis to new revision.
684
The old text for comparison is the union of included revisions.
686
This is used in inserting a new text.
688
Delta is returned as a sequence of
689
(weave1, weave2, newlines).
691
This indicates that weave1:weave2 of the old weave should be
692
replaced by the sequence of lines in newlines. Note that
693
these line numbers are positions in the total weave and don't
694
correspond to the lines in any extracted version, or even the
695
extracted union of included versions.
697
If line1=line2, this is a pure insert; if newlines=[] this is a
698
pure delete. (Similar to difflib.)
700
raise NotImplementedError()
703
def plan_merge(self, ver_a, ver_b):
704
"""Return pseudo-annotation indicating how the two versions merge.
706
This is computed between versions a and b and their common
709
Weave lines present in none of them are skipped entirely.
711
inc_a = self.inclusions([ver_a])
712
inc_b = self.inclusions([ver_b])
713
inc_c = inc_a & inc_b
715
for lineno, insert, deleteset, line in self._walk():
716
if deleteset & inc_c:
717
# killed in parent; can't be in either a or b
718
# not relevant to our work
719
yield 'killed-base', line
720
elif insert in inc_c:
721
# was inserted in base
722
killed_a = bool(deleteset & inc_a)
723
killed_b = bool(deleteset & inc_b)
724
if killed_a and killed_b:
725
yield 'killed-both', line
727
yield 'killed-a', line
729
yield 'killed-b', line
731
yield 'unchanged', line
732
elif insert in inc_a:
733
if deleteset & inc_a:
734
yield 'ghost-a', line
738
elif insert in inc_b:
739
if deleteset & inc_b:
740
yield 'ghost-b', line
744
# not in either revision
745
yield 'irrelevant', line
747
yield 'unchanged', '' # terminator
751
def weave_merge(self, plan, a_marker='<<<<<<< \n', b_marker='>>>>>>> \n'):
755
# TODO: Return a structured form of the conflicts (e.g. 2-tuples for
756
# conflicted regions), rather than just inserting the markers.
758
# TODO: Show some version information (e.g. author, date) on
759
# conflicted regions.
760
for state, line in plan:
761
if state == 'unchanged' or state == 'killed-both':
762
# resync and flush queued conflicts changes if any
763
if not lines_a and not lines_b:
765
elif ch_a and not ch_b:
767
for l in lines_a: yield l
768
elif ch_b and not ch_a:
769
for l in lines_b: yield l
770
elif lines_a == lines_b:
771
for l in lines_a: yield l
774
for l in lines_a: yield l
776
for l in lines_b: yield l
783
if state == 'unchanged':
786
elif state == 'killed-a':
789
elif state == 'killed-b':
792
elif state == 'new-a':
795
elif state == 'new-b':
799
assert state in ('irrelevant', 'ghost-a', 'ghost-b', 'killed-base',
804
def join(self, other, pb=None, msg=None):
806
"""Integrate versions from other into this weave.
808
The resulting weave contains all the history of both weaves;
809
any version you could retrieve from either self or other can be
810
retrieved from self after this call.
812
It is illegal for the two weaves to contain different values
813
or different parents for any version. See also reweave().
815
:param other: The other weave to pull into this one
816
:param pb: An optional progress bar
817
:param msg: An optional message to display for progress
819
if other.numversions() == 0:
820
return # nothing to update, easy
821
# two loops so that we do not change ourselves before verifying it
823
# work through in index order to make sure we get all dependencies
826
for other_idx, name in enumerate(other._names):
827
self._check_version_consistent(other, other_idx, name)
828
sha1 = other._sha1s[other_idx]
832
if name in self._name_map:
833
idx = self.lookup(name)
834
n1 = set(map(other.idx_to_name, other._parents[other_idx]))
835
n2 = set(map(self.idx_to_name, self._parents[idx]))
836
if sha1 == self._sha1s[idx] and n1 == n2:
839
names_to_join.append((other_idx, name))
846
for other_idx, name in names_to_join:
847
# TODO: If all the parents of the other version are already
848
# present then we can avoid some work by just taking the delta
849
# and adjusting the offsets.
850
new_parents = self._imported_parents(other, other_idx)
851
sha1 = other._sha1s[other_idx]
856
pb.update(msg, merged, len(names_to_join))
858
lines = other.get_lines(other_idx)
859
self.add(name, new_parents, lines, sha1)
861
mutter("merged = %d, processed = %d, file_id=%s; deltat=%d"%(
862
merged, processed, self._weave_name, time.time( )-time0))
864
def _imported_parents(self, other, other_idx):
865
"""Return list of parents in self corresponding to indexes in other."""
867
for parent_idx in other._parents[other_idx]:
868
parent_name = other._names[parent_idx]
869
if parent_name not in self._names:
870
# should not be possible
871
raise WeaveError("missing parent {%s} of {%s} in %r"
872
% (parent_name, other._name_map[other_idx], self))
873
new_parents.append(self._name_map[parent_name])
876
def _check_version_consistent(self, other, other_idx, name):
877
"""Check if a version in consistent in this and other.
879
To be consistent it must have:
882
* the same direct parents (by name, not index, and disregarding
885
If present & correct return True;
886
if not present in self return False;
887
if inconsistent raise error."""
888
this_idx = self._name_map.get(name, -1)
890
if self._sha1s[this_idx] != other._sha1s[other_idx]:
891
raise WeaveError("inconsistent texts for version {%s} "
892
"when joining weaves"
894
self_parents = self._parents[this_idx]
895
other_parents = other._parents[other_idx]
896
n1 = set([self._names[i] for i in self_parents])
897
n2 = set([other._names[i] for i in other_parents])
899
raise WeaveParentMismatch("inconsistent parents "
900
"for version {%s}: %s vs %s" % (name, n1, n2))
906
def reweave(self, other, pb=None, msg=None):
907
"""Reweave self with other.
909
:param other: The other weave to merge
910
:param pb: An optional progress bar, indicating how far done we are
911
:param msg: An optional message for the progress
913
new_weave = reweave(self, other, pb=pb, msg=msg)
914
for attr in self.__slots__:
915
setattr(self, attr, getattr(new_weave, attr))
918
def reweave(wa, wb, pb=None, msg=None):
919
"""Combine two weaves and return the result.
921
This works even if a revision R has different parents in
922
wa and wb. In the resulting weave all the parents are given.
924
This is done by just building up a new weave, maintaining ordering
925
of the versions in the two inputs. More efficient approaches
926
might be possible but it should only be necessary to do
927
this operation rarely, when a new previously ghost version is
930
:param pb: An optional progress bar, indicating how far done we are
931
:param msg: An optional message for the progress
935
queue_a = range(wa.numversions())
936
queue_b = range(wb.numversions())
937
# first determine combined parents of all versions
938
# map from version name -> all parent names
939
combined_parents = _reweave_parent_graphs(wa, wb)
940
mutter("combined parents: %r", combined_parents)
941
order = topo_sort(combined_parents.iteritems())
942
mutter("order to reweave: %r", order)
947
for idx, name in enumerate(order):
949
pb.update(msg, idx, len(order))
950
if name in wa._name_map:
951
lines = wa.get_lines(name)
952
if name in wb._name_map:
953
lines_b = wb.get_lines(name)
955
mutter('Weaves differ on content. rev_id {%s}', name)
956
mutter('weaves: %s, %s', wa._weave_name, wb._weave_name)
958
lines = list(difflib.unified_diff(lines, lines_b,
959
wa._weave_name, wb._weave_name))
960
mutter('lines:\n%s', ''.join(lines))
961
raise errors.WeaveTextDiffers(name, wa, wb)
963
lines = wb.get_lines(name)
964
wr.add(name, combined_parents[name], lines)
968
def _reweave_parent_graphs(wa, wb):
969
"""Return combined parent ancestry for two weaves.
971
Returned as a list of (version_name, set(parent_names))"""
973
for weave in [wa, wb]:
974
for idx, name in enumerate(weave._names):
975
p = combined.setdefault(name, set())
976
p.update(map(weave.idx_to_name, weave._parents[idx]))
981
"""Show the weave's table-of-contents"""
982
print '%6s %50s %10s %10s' % ('ver', 'name', 'sha1', 'parents')
983
for i in (6, 50, 10, 10):
986
for i in range(w.numversions()):
989
parent_str = ' '.join(map(str, w._parents[i]))
990
print '%6d %-50.50s %10.10s %s' % (i, name, sha1, parent_str)
994
def weave_stats(weave_file, pb):
995
from bzrlib.weavefile import read_weave
997
wf = file(weave_file, 'rb')
999
# FIXME: doesn't work on pipes
1000
weave_size = wf.tell()
1004
for i in range(vers):
1005
pb.update('checking sizes', i, vers)
1006
for origin, lineno, line in w._extract([i]):
1011
print 'versions %9d' % vers
1012
print 'weave file %9d bytes' % weave_size
1013
print 'total contents %9d bytes' % total
1014
print 'compression ratio %9.2fx' % (float(total) / float(weave_size))
1017
print 'average size %9d bytes' % avg
1018
print 'relative size %9.2fx' % (float(weave_size) / float(avg))
1022
print """bzr weave tool
1024
Experimental tool for weave algorithm.
1027
weave init WEAVEFILE
1028
Create an empty weave file
1029
weave get WEAVEFILE VERSION
1030
Write out specified version.
1031
weave check WEAVEFILE
1032
Check consistency of all versions.
1034
Display table of contents.
1035
weave add WEAVEFILE NAME [BASE...] < NEWTEXT
1036
Add NEWTEXT, with specified parent versions.
1037
weave annotate WEAVEFILE VERSION
1038
Display origin of each line.
1039
weave mash WEAVEFILE VERSION...
1040
Display composite of all selected versions.
1041
weave merge WEAVEFILE VERSION1 VERSION2 > OUT
1042
Auto-merge two versions and display conflicts.
1043
weave diff WEAVEFILE VERSION1 VERSION2
1044
Show differences between two versions.
1048
% weave init foo.weave
1050
% weave add foo.weave ver0 < foo.txt
1053
(create updated version)
1055
% weave get foo.weave 0 | diff -u - foo.txt
1056
% weave add foo.weave ver1 0 < foo.txt
1059
% weave get foo.weave 0 > foo.txt (create forked version)
1061
% weave add foo.weave ver2 0 < foo.txt
1064
% weave merge foo.weave 1 2 > foo.txt (merge them)
1065
% vi foo.txt (resolve conflicts)
1066
% weave add foo.weave merged 1 2 < foo.txt (commit merged version)
1078
# in case we're run directly from the subdirectory
1079
sys.path.append('..')
1081
from bzrlib.weavefile import write_weave, read_weave
1082
from bzrlib.progress import ProgressBar
1097
return read_weave(file(argv[2], 'rb'))
1103
# at the moment, based on everything in the file
1105
parents = map(int, argv[4:])
1106
lines = sys.stdin.readlines()
1107
ver = w.add(name, parents, lines)
1108
write_weave(w, file(argv[2], 'wb'))
1109
print 'added version %r %d' % (name, ver)
1112
if os.path.exists(fn):
1113
raise IOError("file exists")
1115
write_weave(w, file(fn, 'wb'))
1116
elif cmd == 'get': # get one version
1118
sys.stdout.writelines(w.get_iter(int(argv[3])))
1120
elif cmd == 'mash': # get composite
1122
sys.stdout.writelines(w.mash_iter(map(int, argv[3:])))
1125
from difflib import unified_diff
1128
v1, v2 = map(int, argv[3:5])
1131
diff_gen = unified_diff(lines1, lines2,
1132
'%s version %d' % (fn, v1),
1133
'%s version %d' % (fn, v2))
1134
sys.stdout.writelines(diff_gen)
1136
elif cmd == 'annotate':
1138
# newline is added to all lines regardless; too hard to get
1139
# reasonable formatting otherwise
1141
for origin, text in w.annotate(int(argv[3])):
1142
text = text.rstrip('\r\n')
1144
print ' | %s' % (text)
1146
print '%5d | %s' % (origin, text)
1152
elif cmd == 'stats':
1153
weave_stats(argv[2], ProgressBar())
1155
elif cmd == 'check':
1160
print '%d versions ok' % w.numversions()
1162
elif cmd == 'inclusions':
1164
print ' '.join(map(str, w.inclusions([int(argv[3])])))
1166
elif cmd == 'parents':
1168
print ' '.join(map(str, w._parents[int(argv[3])]))
1170
elif cmd == 'plan-merge':
1172
for state, line in w.plan_merge(int(argv[3]), int(argv[4])):
1174
print '%14s | %s' % (state, line),
1176
elif cmd == 'merge':
1178
p = w.plan_merge(int(argv[3]), int(argv[4]))
1179
sys.stdout.writelines(w.weave_merge(p))
1181
elif cmd == 'mash-merge':
1187
v1, v2 = map(int, argv[3:5])
1189
basis = w.inclusions([v1]).intersection(w.inclusions([v2]))
1191
base_lines = list(w.mash_iter(basis))
1192
a_lines = list(w.get(v1))
1193
b_lines = list(w.get(v2))
1195
from bzrlib.merge3 import Merge3
1196
m3 = Merge3(base_lines, a_lines, b_lines)
1198
name_a = 'version %d' % v1
1199
name_b = 'version %d' % v2
1200
sys.stdout.writelines(m3.merge_lines(name_a=name_a, name_b=name_b))
1202
raise ValueError('unknown command %r' % cmd)
1206
def profile_main(argv):
1207
import tempfile, hotshot, hotshot.stats
1209
prof_f = tempfile.NamedTemporaryFile()
1211
prof = hotshot.Profile(prof_f.name)
1213
ret = prof.runcall(main, argv)
1216
stats = hotshot.stats.load(prof_f.name)
1218
stats.sort_stats('cumulative')
1219
## XXX: Might like to write to stderr or the trace file instead but
1220
## print_stats seems hardcoded to stdout
1221
stats.print_stats(20)
1226
def lsprofile_main(argv):
1227
from bzrlib.lsprof import profile
1228
ret,stats = profile(main, argv)
1234
if __name__ == '__main__':
1236
if '--profile' in sys.argv:
1238
args.remove('--profile')
1239
sys.exit(profile_main(args))
1240
elif '--lsprof' in sys.argv:
1242
args.remove('--lsprof')
1243
sys.exit(lsprofile_main(args))
1245
sys.exit(main(sys.argv))
added
removed
bzrlib/weave.py
