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
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):
156
max_origin_len = max(len(origin) for origin, text in annotations)
157
for origin, text in annotations:
158
if full or last_rev_id != origin:
162
to_file.write('%*s | %s' % (max_origin_len, this, text))
167
def _annotations(repo, file_id, rev_id):
168
"""Return the list of (origin_revision_id, line_text) for a revision of a file in a repository."""
169
annotations = repo.texts.annotate((file_id, rev_id))
171
return [(key[-1], line) for (key, line) in annotations]
174
def _expand_annotations(annotations, branch, current_rev=None):
175
"""Expand a a files annotations into command line UI ready tuples.
177
Each tuple includes detailed information, such as the author name, and date
178
string for the commit, rather than just the revision id.
180
:param annotations: The annotations to expand.
181
:param revision_id_to_revno: A map from id to revision numbers.
182
:param branch: A locked branch to query for revision details.
184
repository = branch.repository
185
if current_rev is not None:
186
# This can probably become a function on MutableTree, get_revno_map there,
188
last_revision = current_rev.revision_id
189
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
190
graph = repository.get_graph()
191
revision_graph = dict(((key, value) for key, value in
192
graph.iter_ancestry(current_rev.parent_ids) if value is not None))
193
revision_graph = _strip_NULL_ghosts(revision_graph)
194
revision_graph[last_revision] = current_rev.parent_ids
195
merge_sorted_revisions = tsort.merge_sort(
200
revision_id_to_revno = dict((rev_id, revno)
201
for seq_num, rev_id, depth, revno, end_of_merge in
202
merge_sorted_revisions)
204
revision_id_to_revno = branch.get_revision_id_to_revno_map()
206
revision_ids = set(o for o, t in annotations)
208
if CURRENT_REVISION in revision_ids:
209
revision_id_to_revno[CURRENT_REVISION] = (
210
"%d?" % (branch.revno() + 1),)
211
revisions[CURRENT_REVISION] = current_rev
212
revision_ids = [o for o in revision_ids if
213
repository.has_revision(o)]
214
revisions.update((r.revision_id, r) for r in
215
repository.get_revisions(revision_ids))
216
for origin, text in annotations:
217
text = text.rstrip('\r\n')
218
if origin == last_origin:
219
(revno_str, author, date_str) = ('','','')
222
if origin not in revisions:
223
(revno_str, author, date_str) = ('?','?','?')
225
revno_str = '.'.join(str(i) for i in
226
revision_id_to_revno[origin])
227
rev = revisions[origin]
228
tz = rev.timezone or 0
229
date_str = time.strftime('%Y%m%d',
230
time.gmtime(rev.timestamp + tz))
231
# a lazy way to get something like the email address
232
# TODO: Get real email address
233
author = rev.get_apparent_author()
235
author = extract_email_address(author)
236
except errors.NoEmailInUsername:
237
pass # use the whole name
238
yield (revno_str, author, date_str, origin, text)
241
def reannotate(parents_lines, new_lines, new_revision_id,
242
_left_matching_blocks=None,
243
heads_provider=None):
244
"""Create a new annotated version from new lines and parent annotations.
246
:param parents_lines: List of annotated lines for all parents
247
:param new_lines: The un-annotated new lines
248
:param new_revision_id: The revision-id to associate with new lines
249
(will often be CURRENT_REVISION)
250
:param left_matching_blocks: a hint about which areas are common
251
between the text and its left-hand-parent. The format is
252
the SequenceMatcher.get_matching_blocks format
253
(start_left, start_right, length_of_match).
254
:param heads_provider: An object which provids a .heads() call to resolve
255
if any revision ids are children of others.
256
If None, then any ancestry disputes will be resolved with
259
if len(parents_lines) == 0:
260
lines = [(new_revision_id, line) for line in new_lines]
261
elif len(parents_lines) == 1:
262
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
263
_left_matching_blocks)
264
elif len(parents_lines) == 2:
265
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
266
_left_matching_blocks)
267
lines = _reannotate_annotated(parents_lines[1], new_lines,
268
new_revision_id, left,
271
reannotations = [_reannotate(parents_lines[0], new_lines,
272
new_revision_id, _left_matching_blocks)]
273
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
274
for p in parents_lines[1:])
276
for annos in zip(*reannotations):
277
origins = set(a for a, l in annos)
278
if len(origins) == 1:
279
# All the parents agree, so just return the first one
280
lines.append(annos[0])
283
if len(origins) == 2 and new_revision_id in origins:
284
origins.remove(new_revision_id)
285
if len(origins) == 1:
286
lines.append((origins.pop(), line))
288
lines.append((new_revision_id, line))
292
def _reannotate(parent_lines, new_lines, new_revision_id,
293
matching_blocks=None):
295
if matching_blocks is None:
296
plain_parent_lines = [l for r, l in parent_lines]
297
matcher = patiencediff.PatienceSequenceMatcher(None,
298
plain_parent_lines, new_lines)
299
matching_blocks = matcher.get_matching_blocks()
301
for i, j, n in matching_blocks:
302
for line in new_lines[new_cur:j]:
303
lines.append((new_revision_id, line))
304
lines.extend(parent_lines[i:i+n])
309
def _get_matching_blocks(old, new):
310
matcher = patiencediff.PatienceSequenceMatcher(None,
312
return matcher.get_matching_blocks()
315
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
316
child_lines, start_child, end_child,
317
right_lines, start_right, end_right,
318
heads_provider, revision_id):
319
"""Find lines in plain_right_lines that match the existing lines.
321
:param output_lines: Append final annotated lines to this list
322
:param plain_child_lines: The unannotated new lines for the child text
323
:param child_lines: Lines for the child text which have been annotated
325
:param start_child: Position in plain_child_lines and child_lines to start the
327
:param end_child: Last position in plain_child_lines and child_lines to search
329
:param right_lines: The annotated lines for the whole text for the right
331
:param start_right: Position in right_lines to start the match
332
:param end_right: Last position in right_lines to search for a match
333
:param heads_provider: When parents disagree on the lineage of a line, we
334
need to check if one side supersedes the other
335
:param revision_id: The label to give if a line should be labeled 'tip'
337
output_extend = output_lines.extend
338
output_append = output_lines.append
339
# We need to see if any of the unannotated lines match
340
plain_right_subset = [l for a,l in right_lines[start_right:end_right]]
341
plain_child_subset = plain_child_lines[start_child:end_child]
342
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
346
for right_idx, child_idx, match_len in match_blocks:
347
# All the lines that don't match are just passed along
348
if child_idx > last_child_idx:
349
output_extend(child_lines[start_child + last_child_idx
350
:start_child + child_idx])
351
for offset in xrange(match_len):
352
left = child_lines[start_child+child_idx+offset]
353
right = right_lines[start_right+right_idx+offset]
354
if left[0] == right[0]:
355
# The annotations match, just return the left one
357
elif left[0] == revision_id:
358
# The left parent marked this as unmatched, so let the
359
# right parent claim it
362
# Left and Right both claim this line
363
if heads_provider is None:
364
output_append((revision_id, left[1]))
366
heads = heads_provider.heads((left[0], right[0]))
368
output_append((iter(heads).next(), left[1]))
370
# Both claim different origins, sort lexicographically
371
# so that we always get a stable result.
372
output_append(sorted([left, right])[0])
373
last_child_idx = child_idx + match_len
376
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
377
annotated_lines, heads_provider):
378
"""Update the annotations for a node based on another parent.
380
:param right_parent_lines: A list of annotated lines for the right-hand
382
:param new_lines: The unannotated new lines.
383
:param new_revision_id: The revision_id to attribute to lines which are not
384
present in either parent.
385
:param annotated_lines: A list of annotated lines. This should be the
386
annotation of new_lines based on parents seen so far.
387
:param heads_provider: When parents disagree on the lineage of a line, we
388
need to check if one side supersedes the other.
390
if len(new_lines) != len(annotated_lines):
391
raise AssertionError("mismatched new_lines and annotated_lines")
392
# First compare the newly annotated lines with the right annotated lines.
393
# Lines which were not changed in left or right should match. This tends to
394
# be the bulk of the lines, and they will need no further processing.
396
lines_extend = lines.extend
397
last_right_idx = 0 # The line just after the last match from the right side
399
matching_left_and_right = _get_matching_blocks(right_parent_lines,
401
for right_idx, left_idx, match_len in matching_left_and_right:
402
# annotated lines from last_left_idx to left_idx did not match the lines from
404
# to right_idx, the raw lines should be compared to determine what annotations
406
if last_right_idx == right_idx or last_left_idx == left_idx:
407
# One of the sides is empty, so this is a pure insertion
408
lines_extend(annotated_lines[last_left_idx:left_idx])
410
# We need to see if any of the unannotated lines match
411
_find_matching_unannotated_lines(lines,
412
new_lines, annotated_lines,
413
last_left_idx, left_idx,
415
last_right_idx, right_idx,
418
last_right_idx = right_idx + match_len
419
last_left_idx = left_idx + match_len
420
# If left and right agree on a range, just push that into the output
421
lines_extend(annotated_lines[left_idx:left_idx + match_len])