1
# Copyright (C) 2004, 2005, 2006, 2007 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
"""File annotate based on weave storage"""
19
# TODO: Choice of more or less verbose formats:
21
# interposed: show more details between blocks of modified lines
23
# TODO: Show which revision caused a line to merge into the parent
25
# TODO: perhaps abbreviate timescales depending on how recent they are
26
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
37
from bzrlib.config import extract_email_address
38
from bzrlib.repository import _strip_NULL_ghosts
39
from bzrlib.revision import CURRENT_REVISION, Revision
42
def annotate_file(branch, rev_id, file_id, verbose=False, full=False,
43
to_file=None, show_ids=False):
44
"""Annotate file_id at revision rev_id in branch.
46
The branch should already be read_locked() when annotate_file is called.
48
:param branch: The branch to look for revision numbers and history from.
49
:param rev_id: The revision id to annotate.
50
:param file_id: The file_id to annotate.
51
:param verbose: Show all details rather than truncating to ensure
52
reasonable text width.
53
:param full: XXXX Not sure what this does.
54
:param to_file: The file to output the annotation to; if None stdout is
56
:param show_ids: Show revision ids in the annotation output.
61
# Handle the show_ids case
62
annotations = _annotations(branch.repository, file_id, rev_id)
64
return _show_id_annotations(annotations, to_file, full)
66
# Calculate the lengths of the various columns
67
annotation = list(_expand_annotations(annotations, branch))
68
_print_annotations(annotation, verbose, to_file, full)
71
def annotate_file_tree(tree, file_id, to_file, verbose=False, full=False,
73
"""Annotate file_id in a tree.
75
The tree should already be read_locked() when annotate_file_tree is called.
77
:param tree: The tree to look for revision numbers and history from.
78
:param file_id: The file_id to annotate.
79
:param to_file: The file to output the annotation to.
80
:param verbose: Show all details rather than truncating to ensure
81
reasonable text width.
82
:param full: XXXX Not sure what this does.
83
:param show_ids: Show revision ids in the annotation output.
85
rev_id = tree.last_revision()
88
# Handle the show_ids case
89
annotations = list(tree.annotate_iter(file_id))
91
return _show_id_annotations(annotations, to_file, full)
93
# Create a virtual revision to represent the current tree state.
94
# Should get some more pending commit attributes, like pending tags,
96
current_rev = Revision(CURRENT_REVISION)
97
current_rev.parent_ids = tree.get_parent_ids()
98
current_rev.committer = tree.branch.get_config().username()
99
current_rev.message = "?"
100
current_rev.timestamp = round(time.time(), 3)
101
current_rev.timezone = osutils.local_time_offset()
102
annotation = list(_expand_annotations(annotations, tree.branch,
104
_print_annotations(annotation, verbose, to_file, full)
107
def _print_annotations(annotation, verbose, to_file, full):
108
"""Print annotations to to_file.
110
:param to_file: The file to output the annotation to.
111
:param verbose: Show all details rather than truncating to ensure
112
reasonable text width.
113
:param full: XXXX Not sure what this does.
115
if len(annotation) == 0:
116
max_origin_len = max_revno_len = max_revid_len = 0
118
max_origin_len = max(len(x[1]) for x in annotation)
119
max_revno_len = max(len(x[0]) for x in annotation)
120
max_revid_len = max(len(x[3]) for x in annotation)
122
max_revno_len = min(max_revno_len, 12)
123
max_revno_len = max(max_revno_len, 3)
125
# Output the annotations
127
encoding = getattr(to_file, 'encoding', None) or \
128
osutils.get_terminal_encoding()
129
for (revno_str, author, date_str, line_rev_id, text) in annotation:
131
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
132
max_origin_len, author, date_str)
134
if len(revno_str) > max_revno_len:
135
revno_str = revno_str[:max_revno_len-1] + '>'
136
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
137
if anno.lstrip() == "" and full:
141
except UnicodeEncodeError:
142
# cmd_annotate should be passing in an 'exact' object, which means
143
# we have a direct handle to sys.stdout or equivalent. It may not
144
# be able to handle the exact Unicode characters, but 'annotate' is
145
# a user function (non-scripting), so shouldn't die because of
146
# unrepresentable annotation characters. So encode using 'replace',
147
# and write them again.
148
to_file.write(anno.encode(encoding, 'replace'))
149
to_file.write('| %s\n' % (text,))
153
def _show_id_annotations(annotations, to_file, full):
157
max_origin_len = max(len(origin) for origin, text in annotations)
158
for origin, text in annotations:
159
if full or last_rev_id != origin:
163
to_file.write('%*s | %s' % (max_origin_len, this, text))
168
def _annotations(repo, file_id, rev_id):
169
"""Return the list of (origin_revision_id, line_text) for a revision of a file in a repository."""
170
annotations = repo.texts.annotate((file_id, rev_id))
172
return [(key[-1], line) for (key, line) in annotations]
175
def _expand_annotations(annotations, branch, current_rev=None):
176
"""Expand a file's annotations into command line UI ready tuples.
178
Each tuple includes detailed information, such as the author name, and date
179
string for the commit, rather than just the revision id.
181
:param annotations: The annotations to expand.
182
:param revision_id_to_revno: A map from id to revision numbers.
183
:param branch: A locked branch to query for revision details.
185
repository = branch.repository
186
if current_rev is not None:
187
# This can probably become a function on MutableTree, get_revno_map there,
189
last_revision = current_rev.revision_id
190
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
191
graph = repository.get_graph()
192
revision_graph = dict(((key, value) for key, value in
193
graph.iter_ancestry(current_rev.parent_ids) if value is not None))
194
revision_graph = _strip_NULL_ghosts(revision_graph)
195
revision_graph[last_revision] = current_rev.parent_ids
196
merge_sorted_revisions = tsort.merge_sort(
201
revision_id_to_revno = dict((rev_id, revno)
202
for seq_num, rev_id, depth, revno, end_of_merge in
203
merge_sorted_revisions)
205
revision_id_to_revno = branch.get_revision_id_to_revno_map()
207
revision_ids = set(o for o, t in annotations)
209
if CURRENT_REVISION in revision_ids:
210
revision_id_to_revno[CURRENT_REVISION] = (
211
"%d?" % (branch.revno() + 1),)
212
revisions[CURRENT_REVISION] = current_rev
213
revision_ids = [o for o in revision_ids if
214
repository.has_revision(o)]
215
revisions.update((r.revision_id, r) for r in
216
repository.get_revisions(revision_ids))
217
for origin, text in annotations:
218
text = text.rstrip('\r\n')
219
if origin == last_origin:
220
(revno_str, author, date_str) = ('','','')
223
if origin not in revisions:
224
(revno_str, author, date_str) = ('?','?','?')
226
revno_str = '.'.join(str(i) for i in
227
revision_id_to_revno[origin])
228
rev = revisions[origin]
229
tz = rev.timezone or 0
230
date_str = time.strftime('%Y%m%d',
231
time.gmtime(rev.timestamp + tz))
232
# a lazy way to get something like the email address
233
# TODO: Get real email address
234
author = rev.get_apparent_authors()[0]
236
author = extract_email_address(author)
237
except errors.NoEmailInUsername:
238
pass # use the whole name
239
yield (revno_str, author, date_str, origin, text)
242
def reannotate(parents_lines, new_lines, new_revision_id,
243
_left_matching_blocks=None,
244
heads_provider=None):
245
"""Create a new annotated version from new lines and parent annotations.
247
:param parents_lines: List of annotated lines for all parents
248
:param new_lines: The un-annotated new lines
249
:param new_revision_id: The revision-id to associate with new lines
250
(will often be CURRENT_REVISION)
251
:param left_matching_blocks: a hint about which areas are common
252
between the text and its left-hand-parent. The format is
253
the SequenceMatcher.get_matching_blocks format
254
(start_left, start_right, length_of_match).
255
:param heads_provider: An object which provides a .heads() call to resolve
256
if any revision ids are children of others.
257
If None, then any ancestry disputes will be resolved with
260
if len(parents_lines) == 0:
261
lines = [(new_revision_id, line) for line in new_lines]
262
elif len(parents_lines) == 1:
263
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
264
_left_matching_blocks)
265
elif len(parents_lines) == 2:
266
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
267
_left_matching_blocks)
268
lines = _reannotate_annotated(parents_lines[1], new_lines,
269
new_revision_id, left,
272
reannotations = [_reannotate(parents_lines[0], new_lines,
273
new_revision_id, _left_matching_blocks)]
274
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
275
for p in parents_lines[1:])
277
for annos in zip(*reannotations):
278
origins = set(a for a, l in annos)
279
if len(origins) == 1:
280
# All the parents agree, so just return the first one
281
lines.append(annos[0])
284
if len(origins) == 2 and new_revision_id in origins:
285
origins.remove(new_revision_id)
286
if len(origins) == 1:
287
lines.append((origins.pop(), line))
289
lines.append((new_revision_id, line))
293
def _reannotate(parent_lines, new_lines, new_revision_id,
294
matching_blocks=None):
296
if matching_blocks is None:
297
plain_parent_lines = [l for r, l in parent_lines]
298
matcher = patiencediff.PatienceSequenceMatcher(None,
299
plain_parent_lines, new_lines)
300
matching_blocks = matcher.get_matching_blocks()
302
for i, j, n in matching_blocks:
303
for line in new_lines[new_cur:j]:
304
lines.append((new_revision_id, line))
305
lines.extend(parent_lines[i:i+n])
310
def _get_matching_blocks(old, new):
311
matcher = patiencediff.PatienceSequenceMatcher(None,
313
return matcher.get_matching_blocks()
316
def _break_annotation_tie(annotated_lines):
317
"""Chose an attribution between several possible ones.
319
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
320
the lines are identical but the revids different while no parent
321
relation exist between them
323
:return : The "winning" line. This must be one with a revid that
324
guarantees that further criss-cross merges will converge. Failing to
325
do so have performance implications.
327
# sort lexicographically so that we always get a stable result.
329
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
330
# with the current implementation, chosing the oldest revision is known to
331
# provide better results (as in matching user expectations). The most
332
# common use case being manual cherry-pick from an already existing
334
return sorted(annotated_lines)[0]
337
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
338
child_lines, start_child, end_child,
339
right_lines, start_right, end_right,
340
heads_provider, revision_id):
341
"""Find lines in plain_right_lines that match the existing lines.
343
:param output_lines: Append final annotated lines to this list
344
:param plain_child_lines: The unannotated new lines for the child text
345
:param child_lines: Lines for the child text which have been annotated
348
:param start_child: Position in plain_child_lines and child_lines to start
350
:param end_child: Last position in plain_child_lines and child_lines to
352
:param right_lines: The annotated lines for the whole text for the right
354
:param start_right: Position in right_lines to start the match
355
:param end_right: Last position in right_lines to search for a match
356
:param heads_provider: When parents disagree on the lineage of a line, we
357
need to check if one side supersedes the other
358
:param revision_id: The label to give if a line should be labeled 'tip'
360
output_extend = output_lines.extend
361
output_append = output_lines.append
362
# We need to see if any of the unannotated lines match
363
plain_right_subset = [l for a,l in right_lines[start_right:end_right]]
364
plain_child_subset = plain_child_lines[start_child:end_child]
365
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
369
for right_idx, child_idx, match_len in match_blocks:
370
# All the lines that don't match are just passed along
371
if child_idx > last_child_idx:
372
output_extend(child_lines[start_child + last_child_idx
373
:start_child + child_idx])
374
for offset in xrange(match_len):
375
left = child_lines[start_child+child_idx+offset]
376
right = right_lines[start_right+right_idx+offset]
377
if left[0] == right[0]:
378
# The annotations match, just return the left one
380
elif left[0] == revision_id:
381
# The left parent marked this as unmatched, so let the
382
# right parent claim it
385
# Left and Right both claim this line
386
if heads_provider is None:
387
output_append((revision_id, left[1]))
389
heads = heads_provider.heads((left[0], right[0]))
391
output_append((iter(heads).next(), left[1]))
393
# Both claim different origins, get a stable result.
394
# If the result is not stable, there is a risk a
395
# performance degradation as criss-cross merges will
396
# flip-flop the attribution.
397
output_append(_break_annotation_tie([left, right]))
398
last_child_idx = child_idx + match_len
401
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
402
annotated_lines, heads_provider):
403
"""Update the annotations for a node based on another parent.
405
:param right_parent_lines: A list of annotated lines for the right-hand
407
:param new_lines: The unannotated new lines.
408
:param new_revision_id: The revision_id to attribute to lines which are not
409
present in either parent.
410
:param annotated_lines: A list of annotated lines. This should be the
411
annotation of new_lines based on parents seen so far.
412
:param heads_provider: When parents disagree on the lineage of a line, we
413
need to check if one side supersedes the other.
415
if len(new_lines) != len(annotated_lines):
416
raise AssertionError("mismatched new_lines and annotated_lines")
417
# First compare the newly annotated lines with the right annotated lines.
418
# Lines which were not changed in left or right should match. This tends to
419
# be the bulk of the lines, and they will need no further processing.
421
lines_extend = lines.extend
422
last_right_idx = 0 # The line just after the last match from the right side
424
matching_left_and_right = _get_matching_blocks(right_parent_lines,
426
for right_idx, left_idx, match_len in matching_left_and_right:
427
# annotated lines from last_left_idx to left_idx did not match the
428
# lines from last_right_idx to right_idx, the raw lines should be
429
# compared to determine what annotations need to be updated
430
if last_right_idx == right_idx or last_left_idx == left_idx:
431
# One of the sides is empty, so this is a pure insertion
432
lines_extend(annotated_lines[last_left_idx:left_idx])
434
# We need to see if any of the unannotated lines match
435
_find_matching_unannotated_lines(lines,
436
new_lines, annotated_lines,
437
last_left_idx, left_idx,
439
last_right_idx, right_idx,
442
last_right_idx = right_idx + match_len
443
last_left_idx = left_idx + match_len
444
# If left and right agree on a range, just push that into the output
445
lines_extend(annotated_lines[left_idx:left_idx + match_len])
added
removed
bzrlib/annotate.py
