1
# Copyright (C) 2004, 2005 by Canonical Ltd
1
# Copyright (C) 2005-2010 Canonical Ltd
3
3
# This program is free software; you can redistribute it and/or modify
4
4
# it under the terms of the GNU General Public License as published by
5
5
# the Free Software Foundation; either version 2 of the License, or
6
6
# (at your option) any later version.
8
8
# This program is distributed in the hope that it will be useful,
9
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
11
# GNU General Public License for more details.
13
13
# You should have received a copy of the GNU General Public License
14
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
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
17
"""File annotate based on weave storage"""
19
from __future__ import absolute_import
19
21
# TODO: Choice of more or less verbose formats:
21
23
# interposed: show more details between blocks of modified lines
23
25
# TODO: Show which revision caused a line to merge into the parent
25
# TODO: With --long, show entire email address, not just the first bit
27
27
# TODO: perhaps abbreviate timescales depending on how recent they are
28
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
28
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
33
from bzrlib.lazy_import import lazy_import
34
lazy_import(globals(), """
35
44
from bzrlib.config import extract_email_address
38
def annotate_file(branch, rev_id, file_id, verbose=False, full=False,
45
from bzrlib.repository import _strip_NULL_ghosts
46
from bzrlib.revision import (
52
def annotate_file_tree(tree, file_id, to_file, verbose=False, full=False,
53
show_ids=False, branch=None):
54
"""Annotate file_id in a tree.
56
The tree should already be read_locked() when annotate_file_tree is called.
58
:param tree: The tree to look for revision numbers and history from.
59
:param file_id: The file_id to annotate.
60
:param to_file: The file to output the annotation to.
61
:param verbose: Show all details rather than truncating to ensure
62
reasonable text width.
63
:param full: XXXX Not sure what this does.
64
:param show_ids: Show revision ids in the annotation output.
65
:param branch: Branch to use for revision revno lookups
40
69
if to_file is None:
41
70
to_file = sys.stdout
44
for (revno_str, author, date_str, line_rev_id, text ) in \
45
_annotate_file(branch, rev_id, file_id ):
72
# Handle the show_ids case
73
annotations = list(tree.annotate_iter(file_id))
75
return _show_id_annotations(annotations, to_file, full)
77
if not getattr(tree, "get_revision_id", False):
78
# Create a virtual revision to represent the current tree state.
79
# Should get some more pending commit attributes, like pending tags,
81
current_rev = Revision(CURRENT_REVISION)
82
current_rev.parent_ids = tree.get_parent_ids()
84
current_rev.committer = branch.get_config_stack().get('email')
85
except errors.NoWhoami:
86
current_rev.committer = 'local user'
87
current_rev.message = "?"
88
current_rev.timestamp = round(time.time(), 3)
89
current_rev.timezone = osutils.local_time_offset()
92
annotation = list(_expand_annotations(annotations, branch,
94
_print_annotations(annotation, verbose, to_file, full)
97
def _print_annotations(annotation, verbose, to_file, full):
98
"""Print annotations to to_file.
100
:param to_file: The file to output the annotation to.
101
:param verbose: Show all details rather than truncating to ensure
102
reasonable text width.
103
:param full: XXXX Not sure what this does.
105
if len(annotation) == 0:
106
max_origin_len = max_revno_len = max_revid_len = 0
108
max_origin_len = max(len(x[1]) for x in annotation)
109
max_revno_len = max(len(x[0]) for x in annotation)
110
max_revid_len = max(len(x[3]) for x in annotation)
112
max_revno_len = min(max_revno_len, 12)
113
max_revno_len = max(max_revno_len, 3)
115
# Output the annotations
117
encoding = getattr(to_file, 'encoding', None) or \
118
osutils.get_terminal_encoding()
119
for (revno_str, author, date_str, line_rev_id, text) in annotation:
48
anno = '%5s %-12s %8s ' % (revno_str, author[:12], date_str)
50
anno = "%5s %-7s " % ( revno_str, author[:7] )
52
if anno.lstrip() == "" and full: anno = prevanno
53
print >>to_file, '%s| %s' % (anno, text)
56
def _annotate_file(branch, rev_id, file_id ):
58
rh = branch.revision_history()
59
w = branch.weave_store.get_weave(file_id, branch.get_transaction())
121
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
122
max_origin_len, author, date_str)
124
if len(revno_str) > max_revno_len:
125
revno_str = revno_str[:max_revno_len-1] + '>'
126
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
127
if anno.lstrip() == "" and full:
131
except UnicodeEncodeError:
132
# cmd_annotate should be passing in an 'exact' object, which means
133
# we have a direct handle to sys.stdout or equivalent. It may not
134
# be able to handle the exact Unicode characters, but 'annotate' is
135
# a user function (non-scripting), so shouldn't die because of
136
# unrepresentable annotation characters. So encode using 'replace',
137
# and write them again.
138
to_file.write(anno.encode(encoding, 'replace'))
139
to_file.write('| %s\n' % (text,))
143
def _show_id_annotations(annotations, to_file, full):
147
max_origin_len = max(len(origin) for origin, text in annotations)
148
for origin, text in annotations:
149
if full or last_rev_id != origin:
153
to_file.write('%*s | %s' % (max_origin_len, this, text))
158
def _expand_annotations(annotations, branch, current_rev=None):
159
"""Expand a file's annotations into command line UI ready tuples.
161
Each tuple includes detailed information, such as the author name, and date
162
string for the commit, rather than just the revision id.
164
:param annotations: The annotations to expand.
165
:param revision_id_to_revno: A map from id to revision numbers.
166
:param branch: A locked branch to query for revision details.
168
repository = branch.repository
169
if current_rev is not None:
170
# This can probably become a function on MutableTree, get_revno_map
171
# there, or something.
172
last_revision = current_rev.revision_id
173
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
174
# XXX: The main difficulty is that we need to inject a single new node
175
# (current_rev) into the graph before it gets numbered, etc.
176
# Once KnownGraph gets an 'add_node()' function, we can use
177
# VF.get_known_graph_ancestry().
178
graph = repository.get_graph()
179
revision_graph = dict(((key, value) for key, value in
180
graph.iter_ancestry(current_rev.parent_ids) if value is not None))
181
revision_graph = _strip_NULL_ghosts(revision_graph)
182
revision_graph[last_revision] = current_rev.parent_ids
183
merge_sorted_revisions = tsort.merge_sort(
188
revision_id_to_revno = dict((rev_id, revno)
189
for seq_num, rev_id, depth, revno, end_of_merge in
190
merge_sorted_revisions)
192
revision_id_to_revno = branch.get_revision_id_to_revno_map()
60
193
last_origin = None
61
for origin, text in w.annotate_iter(rev_id):
194
revision_ids = set(o for o, t in annotations)
196
if CURRENT_REVISION in revision_ids:
197
revision_id_to_revno[CURRENT_REVISION] = (
198
"%d?" % (branch.revno() + 1),)
199
revisions[CURRENT_REVISION] = current_rev
200
revision_ids = [o for o in revision_ids if
201
repository.has_revision(o)]
202
revisions.update((r.revision_id, r) for r in
203
repository.get_revisions(revision_ids))
204
for origin, text in annotations:
62
205
text = text.rstrip('\r\n')
63
206
if origin == last_origin:
64
207
(revno_str, author, date_str) = ('','','')
66
209
last_origin = origin
67
line_rev_id = w.idx_to_name(origin)
68
if not branch.has_revision(line_rev_id):
210
if origin not in revisions:
69
211
(revno_str, author, date_str) = ('?','?','?')
72
revno_str = str(rh.index(line_rev_id) + 1)
75
rev = branch.get_revision(line_rev_id)
213
revno_str = '.'.join(str(i) for i in
214
revision_id_to_revno[origin])
215
rev = revisions[origin]
76
216
tz = rev.timezone or 0
77
date_str = time.strftime('%Y%m%d',
217
date_str = time.strftime('%Y%m%d',
78
218
time.gmtime(rev.timestamp + tz))
79
219
# a lazy way to get something like the email address
80
220
# TODO: Get real email address
81
author = rev.committer
221
author = rev.get_apparent_authors()[0]
83
223
author = extract_email_address(author)
224
except errors.NoEmailInUsername:
85
225
pass # use the whole name
86
yield (revno_str, author, date_str, line_rev_id, text)
89
if __name__ == '__main__':
90
from bzrlib.branch import Branch
91
from bzrlib.trace import enable_default_logging
92
from bzrlib.workingtree import WorkingTree
94
enable_default_logging()
95
b = Branch.open_containing(sys.argv[1])[0]
96
tree = WorkingTree(b.base, b)
97
rp = tree.relpath(sys.argv[1])
98
tree = b.revision_tree(b.last_revision())
99
file_id = tree.inventory.path2id(rp)
100
file_version = tree.inventory[file_id].revision
101
annotate_file(b, file_version, file_id, to_file = sys.stdout)
226
yield (revno_str, author, date_str, origin, text)
229
def reannotate(parents_lines, new_lines, new_revision_id,
230
_left_matching_blocks=None,
231
heads_provider=None):
232
"""Create a new annotated version from new lines and parent annotations.
234
:param parents_lines: List of annotated lines for all parents
235
:param new_lines: The un-annotated new lines
236
:param new_revision_id: The revision-id to associate with new lines
237
(will often be CURRENT_REVISION)
238
:param left_matching_blocks: a hint about which areas are common
239
between the text and its left-hand-parent. The format is
240
the SequenceMatcher.get_matching_blocks format
241
(start_left, start_right, length_of_match).
242
:param heads_provider: An object which provides a .heads() call to resolve
243
if any revision ids are children of others.
244
If None, then any ancestry disputes will be resolved with
247
if len(parents_lines) == 0:
248
lines = [(new_revision_id, line) for line in new_lines]
249
elif len(parents_lines) == 1:
250
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
251
_left_matching_blocks)
252
elif len(parents_lines) == 2:
253
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
254
_left_matching_blocks)
255
lines = _reannotate_annotated(parents_lines[1], new_lines,
256
new_revision_id, left,
259
reannotations = [_reannotate(parents_lines[0], new_lines,
260
new_revision_id, _left_matching_blocks)]
261
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
262
for p in parents_lines[1:])
264
for annos in zip(*reannotations):
265
origins = set(a for a, l in annos)
266
if len(origins) == 1:
267
# All the parents agree, so just return the first one
268
lines.append(annos[0])
271
if len(origins) == 2 and new_revision_id in origins:
272
origins.remove(new_revision_id)
273
if len(origins) == 1:
274
lines.append((origins.pop(), line))
276
lines.append((new_revision_id, line))
280
def _reannotate(parent_lines, new_lines, new_revision_id,
281
matching_blocks=None):
283
if matching_blocks is None:
284
plain_parent_lines = [l for r, l in parent_lines]
285
matcher = patiencediff.PatienceSequenceMatcher(None,
286
plain_parent_lines, new_lines)
287
matching_blocks = matcher.get_matching_blocks()
289
for i, j, n in matching_blocks:
290
for line in new_lines[new_cur:j]:
291
lines.append((new_revision_id, line))
292
lines.extend(parent_lines[i:i+n])
297
def _get_matching_blocks(old, new):
298
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
299
return matcher.get_matching_blocks()
302
_break_annotation_tie = None
304
def _old_break_annotation_tie(annotated_lines):
305
"""Chose an attribution between several possible ones.
307
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
308
the lines are identical but the revids different while no parent
309
relation exist between them
311
:return : The "winning" line. This must be one with a revid that
312
guarantees that further criss-cross merges will converge. Failing to
313
do so have performance implications.
315
# sort lexicographically so that we always get a stable result.
317
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
318
# with the current implementation, chosing the oldest revision is known to
319
# provide better results (as in matching user expectations). The most
320
# common use case being manual cherry-pick from an already existing
322
return sorted(annotated_lines)[0]
325
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
326
child_lines, start_child, end_child,
327
right_lines, start_right, end_right,
328
heads_provider, revision_id):
329
"""Find lines in plain_right_lines that match the existing lines.
331
:param output_lines: Append final annotated lines to this list
332
:param plain_child_lines: The unannotated new lines for the child text
333
:param child_lines: Lines for the child text which have been annotated
336
:param start_child: Position in plain_child_lines and child_lines to start
338
:param end_child: Last position in plain_child_lines and child_lines to
340
:param right_lines: The annotated lines for the whole text for the right
342
:param start_right: Position in right_lines to start the match
343
:param end_right: Last position in right_lines to search for a match
344
:param heads_provider: When parents disagree on the lineage of a line, we
345
need to check if one side supersedes the other
346
:param revision_id: The label to give if a line should be labeled 'tip'
348
output_extend = output_lines.extend
349
output_append = output_lines.append
350
# We need to see if any of the unannotated lines match
351
plain_right_subset = [l for a,l in right_lines[start_right:end_right]]
352
plain_child_subset = plain_child_lines[start_child:end_child]
353
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
357
for right_idx, child_idx, match_len in match_blocks:
358
# All the lines that don't match are just passed along
359
if child_idx > last_child_idx:
360
output_extend(child_lines[start_child + last_child_idx
361
:start_child + child_idx])
362
for offset in xrange(match_len):
363
left = child_lines[start_child+child_idx+offset]
364
right = right_lines[start_right+right_idx+offset]
365
if left[0] == right[0]:
366
# The annotations match, just return the left one
368
elif left[0] == revision_id:
369
# The left parent marked this as unmatched, so let the
370
# right parent claim it
373
# Left and Right both claim this line
374
if heads_provider is None:
375
output_append((revision_id, left[1]))
377
heads = heads_provider.heads((left[0], right[0]))
379
output_append((iter(heads).next(), left[1]))
381
# Both claim different origins, get a stable result.
382
# If the result is not stable, there is a risk a
383
# performance degradation as criss-cross merges will
384
# flip-flop the attribution.
385
if _break_annotation_tie is None:
387
_old_break_annotation_tie([left, right]))
389
output_append(_break_annotation_tie([left, right]))
390
last_child_idx = child_idx + match_len
393
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
394
annotated_lines, heads_provider):
395
"""Update the annotations for a node based on another parent.
397
:param right_parent_lines: A list of annotated lines for the right-hand
399
:param new_lines: The unannotated new lines.
400
:param new_revision_id: The revision_id to attribute to lines which are not
401
present in either parent.
402
:param annotated_lines: A list of annotated lines. This should be the
403
annotation of new_lines based on parents seen so far.
404
:param heads_provider: When parents disagree on the lineage of a line, we
405
need to check if one side supersedes the other.
407
if len(new_lines) != len(annotated_lines):
408
raise AssertionError("mismatched new_lines and annotated_lines")
409
# First compare the newly annotated lines with the right annotated lines.
410
# Lines which were not changed in left or right should match. This tends to
411
# be the bulk of the lines, and they will need no further processing.
413
lines_extend = lines.extend
414
last_right_idx = 0 # The line just after the last match from the right side
416
matching_left_and_right = _get_matching_blocks(right_parent_lines,
418
for right_idx, left_idx, match_len in matching_left_and_right:
419
# annotated lines from last_left_idx to left_idx did not match the
420
# lines from last_right_idx to right_idx, the raw lines should be
421
# compared to determine what annotations need to be updated
422
if last_right_idx == right_idx or last_left_idx == left_idx:
423
# One of the sides is empty, so this is a pure insertion
424
lines_extend(annotated_lines[last_left_idx:left_idx])
426
# We need to see if any of the unannotated lines match
427
_find_matching_unannotated_lines(lines,
428
new_lines, annotated_lines,
429
last_left_idx, left_idx,
431
last_right_idx, right_idx,
434
last_right_idx = right_idx + match_len
435
last_left_idx = left_idx + match_len
436
# If left and right agree on a range, just push that into the output
437
lines_extend(annotated_lines[left_idx:left_idx + match_len])
442
from bzrlib._annotator_pyx import Annotator
443
except ImportError, e:
444
osutils.failed_to_load_extension(e)
445
from bzrlib._annotator_py import Annotator
added
removed
bzrlib/annotate.py
