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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 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
40
def annotate_file(branch, rev_id, file_id, verbose=False, full=False,
41
to_file=None, show_ids=False):
42
"""Annotate file_id at revision rev_id in branch.
44
The branch should already be read_locked() when annotate_file is called.
46
:param branch: The branch to look for revision numbers and history from.
47
:param rev_id: The revision id to annotate.
48
:param file_id: The file_id to annotate.
49
:param verbose: Show all details rather than truncating to ensure
50
reasonable text width.
51
:param full: XXXX Not sure what this does.
52
:param to_file: The file to output the annotation to; if None stdout is
54
:param show_ids: Show revision ids in the annotation output.
59
# Handle the show_ids case
62
annotations = _annotations(branch.repository, file_id, rev_id)
63
max_origin_len = max(len(origin) for origin, text in annotations)
64
for origin, text in annotations:
65
if full or last_rev_id != origin:
69
to_file.write('%*s | %s' % (max_origin_len, this, text))
73
# Calculate the lengths of the various columns
74
annotation = list(_annotate_file(branch, rev_id, file_id))
75
if len(annotation) == 0:
76
max_origin_len = max_revno_len = max_revid_len = 0
78
max_origin_len = max(len(x[1]) for x in annotation)
79
max_revno_len = max(len(x[0]) for x in annotation)
80
max_revid_len = max(len(x[3]) for x in annotation)
82
max_revno_len = min(max_revno_len, 12)
83
max_revno_len = max(max_revno_len, 3)
85
# Output the annotations
87
encoding = getattr(to_file, 'encoding', None) or \
88
osutils.get_terminal_encoding()
89
for (revno_str, author, date_str, line_rev_id, text) in annotation:
91
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
92
max_origin_len, author, date_str)
94
if len(revno_str) > max_revno_len:
95
revno_str = revno_str[:max_revno_len-1] + '>'
96
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
97
if anno.lstrip() == "" and full:
101
except UnicodeEncodeError:
102
# cmd_annotate should be passing in an 'exact' object, which means
103
# we have a direct handle to sys.stdout or equivalent. It may not
104
# be able to handle the exact Unicode characters, but 'annotate' is
105
# a user function (non-scripting), so shouldn't die because of
106
# unrepresentable annotation characters. So encode using 'replace',
107
# and write them again.
108
to_file.write(anno.encode(encoding, 'replace'))
109
to_file.write('| %s\n' % (text,))
113
def _annotations(repo, file_id, rev_id):
114
"""Return the list of (origin,text) for a revision of a file in a repository."""
115
w = repo.weave_store.get_weave(file_id, repo.get_transaction())
116
return w.annotate(rev_id)
119
def _annotate_file(branch, rev_id, file_id):
120
"""Yield the origins for each line of a file.
122
This includes detailed information, such as the author name, and
123
date string for the commit, rather than just the revision id.
125
revision_id_to_revno = branch.get_revision_id_to_revno_map()
126
annotations = _annotations(branch.repository, file_id, rev_id)
128
revision_ids = set(o for o, t in annotations)
129
revision_ids = [o for o in revision_ids if
130
branch.repository.has_revision(o)]
131
revisions = dict((r.revision_id, r) for r in
132
branch.repository.get_revisions(revision_ids))
133
for origin, text in annotations:
134
text = text.rstrip('\r\n')
135
if origin == last_origin:
136
(revno_str, author, date_str) = ('','','')
139
if origin not in revisions:
140
(revno_str, author, date_str) = ('?','?','?')
142
revno_str = '.'.join(str(i) for i in
143
revision_id_to_revno[origin])
144
rev = revisions[origin]
145
tz = rev.timezone or 0
146
date_str = time.strftime('%Y%m%d',
147
time.gmtime(rev.timestamp + tz))
148
# a lazy way to get something like the email address
149
# TODO: Get real email address
150
author = rev.get_apparent_author()
152
author = extract_email_address(author)
153
except errors.NoEmailInUsername:
154
pass # use the whole name
155
yield (revno_str, author, date_str, origin, text)
158
def reannotate(parents_lines, new_lines, new_revision_id,
159
_left_matching_blocks=None,
160
heads_provider=None):
161
"""Create a new annotated version from new lines and parent annotations.
163
:param parents_lines: List of annotated lines for all parents
164
:param new_lines: The un-annotated new lines
165
:param new_revision_id: The revision-id to associate with new lines
166
(will often be CURRENT_REVISION)
167
:param left_matching_blocks: a hint about which areas are common
168
between the text and its left-hand-parent. The format is
169
the SequenceMatcher.get_matching_blocks format
170
(start_left, start_right, length_of_match).
171
:param heads_provider: An object which provids a .heads() call to resolve
172
if any revision ids are children of others.
173
If None, then any ancestry disputes will be resolved with
176
if len(parents_lines) == 0:
177
lines = [(new_revision_id, line) for line in new_lines]
178
elif len(parents_lines) == 1:
179
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
180
_left_matching_blocks)
181
elif len(parents_lines) == 2:
182
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
183
_left_matching_blocks)
184
lines = _reannotate_annotated(parents_lines[1], new_lines,
185
new_revision_id, left,
188
reannotations = [_reannotate(parents_lines[0], new_lines,
189
new_revision_id, _left_matching_blocks)]
190
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
191
for p in parents_lines[1:])
193
for annos in zip(*reannotations):
194
origins = set(a for a, l in annos)
195
if len(origins) == 1:
196
# All the parents agree, so just return the first one
197
lines.append(annos[0])
200
if len(origins) == 2 and new_revision_id in origins:
201
origins.remove(new_revision_id)
202
if len(origins) == 1:
203
lines.append((origins.pop(), line))
205
lines.append((new_revision_id, line))
209
def _reannotate(parent_lines, new_lines, new_revision_id,
210
matching_blocks=None):
212
if matching_blocks is None:
213
plain_parent_lines = [l for r, l in parent_lines]
214
matcher = patiencediff.PatienceSequenceMatcher(None,
215
plain_parent_lines, new_lines)
216
matching_blocks = matcher.get_matching_blocks()
218
for i, j, n in matching_blocks:
219
for line in new_lines[new_cur:j]:
220
lines.append((new_revision_id, line))
221
lines.extend(parent_lines[i:i+n])
226
def _get_matching_blocks(old, new):
227
matcher = patiencediff.PatienceSequenceMatcher(None,
229
return matcher.get_matching_blocks()
232
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
233
child_lines, start_child, end_child,
234
right_lines, start_right, end_right,
235
heads_provider, revision_id):
236
"""Find lines in plain_right_lines that match the existing lines.
238
:param output_lines: Append final annotated lines to this list
239
:param plain_child_lines: The unannotated new lines for the child text
240
:param child_lines: Lines for the child text which have been annotated
242
:param start_child: Position in plain_child_lines and child_lines to start the
244
:param end_child: Last position in plain_child_lines and child_lines to search
246
:param right_lines: The annotated lines for the whole text for the right
248
:param start_right: Position in right_lines to start the match
249
:param end_right: Last position in right_lines to search for a match
250
:param heads_provider: When parents disagree on the lineage of a line, we
251
need to check if one side supersedes the other
252
:param revision_id: The label to give if a line should be labeled 'tip'
254
output_extend = output_lines.extend
255
output_append = output_lines.append
256
# We need to see if any of the unannotated lines match
257
plain_right_subset = [l for a,l in right_lines[start_right:end_right]]
258
plain_child_subset = plain_child_lines[start_child:end_child]
259
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
263
for right_idx, child_idx, match_len in match_blocks:
264
# All the lines that don't match are just passed along
265
if child_idx > last_child_idx:
266
output_extend(child_lines[start_child + last_child_idx
267
:start_child + child_idx])
268
for offset in xrange(match_len):
269
left = child_lines[start_child+child_idx+offset]
270
right = right_lines[start_right+right_idx+offset]
271
if left[0] == right[0]:
272
# The annotations match, just return the left one
274
elif left[0] == revision_id:
275
# The left parent marked this as unmatched, so let the
276
# right parent claim it
279
# Left and Right both claim this line
280
if heads_provider is None:
281
output_append((revision_id, left[1]))
283
heads = heads_provider.heads((left[0], right[0]))
285
output_append((iter(heads).next(), left[1]))
287
# Both claim different origins
288
output_append((revision_id, left[1]))
289
# We know that revision_id is the head for
290
# left and right, so cache it
291
heads_provider.cache(
292
(revision_id, left[0]),
294
heads_provider.cache(
295
(revision_id, right[0]),
297
last_child_idx = child_idx + match_len
300
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
301
annotated_lines, heads_provider):
302
"""Update the annotations for a node based on another parent.
304
:param right_parent_lines: A list of annotated lines for the right-hand
306
:param new_lines: The unannotated new lines.
307
:param new_revision_id: The revision_id to attribute to lines which are not
308
present in either parent.
309
:param annotated_lines: A list of annotated lines. This should be the
310
annotation of new_lines based on parents seen so far.
311
:param heads_provider: When parents disagree on the lineage of a line, we
312
need to check if one side supersedes the other.
314
assert len(new_lines) == len(annotated_lines)
315
# First compare the newly annotated lines with the right annotated lines.
316
# Lines which were not changed in left or right should match. This tends to
317
# be the bulk of the lines, and they will need no further processing.
319
lines_extend = lines.extend
320
last_right_idx = 0 # The line just after the last match from the right side
322
matching_left_and_right = _get_matching_blocks(right_parent_lines,
324
for right_idx, left_idx, match_len in matching_left_and_right:
325
# annotated lines from last_left_idx to left_idx did not match the lines from
327
# to right_idx, the raw lines should be compared to determine what annotations
329
if last_right_idx == right_idx or last_left_idx == left_idx:
330
# One of the sides is empty, so this is a pure insertion
331
lines_extend(annotated_lines[last_left_idx:left_idx])
333
# We need to see if any of the unannotated lines match
334
_find_matching_unannotated_lines(lines,
335
new_lines, annotated_lines,
336
last_left_idx, left_idx,
338
last_right_idx, right_idx,
341
last_right_idx = right_idx + match_len
342
last_left_idx = left_idx + match_len
343
# If left and right agree on a range, just push that into the output
344
assert len(lines) == left_idx
345
lines_extend(annotated_lines[left_idx:left_idx + match_len])
added
removed
bzrlib/annotate.py
