1
# Copyright (C) 2005-2010 Canonical Ltd
1
# Copyright (C) 2004, 2005 by 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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
17
"""File annotate based on weave storage"""
19
from __future__ import absolute_import
21
19
# TODO: Choice of more or less verbose formats:
23
21
# interposed: show more details between blocks of modified lines
25
23
# TODO: Show which revision caused a line to merge into the parent
27
25
# TODO: perhaps abbreviate timescales depending on how recent they are
28
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
26
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
33
from .lazy_import import lazy_import
34
lazy_import(globals(), """
47
from .repository import _strip_NULL_ghosts
48
from .revision import (
54
def annotate_file_tree(tree, path, to_file, verbose=False, full=False,
55
show_ids=False, branch=None):
56
"""Annotate path in a tree.
58
The tree should already be read_locked() when annotate_file_tree is called.
60
:param tree: The tree to look for revision numbers and history from.
61
:param path: The path to annotate
62
:param to_file: The file to output the annotation to.
63
:param verbose: Show all details rather than truncating to ensure
64
reasonable text width.
65
:param full: XXXX Not sure what this does.
66
:param show_ids: Show revision ids in the annotation output.
67
:param branch: Branch to use for revision revno lookups
33
from bzrlib.config import extract_email_address
34
from bzrlib.errors import BzrError
37
def annotate_file(branch, rev_id, file_id, verbose=False, full=False,
71
39
if to_file is None:
72
40
to_file = sys.stdout
74
encoding = osutils.get_terminal_encoding()
75
# Handle the show_ids case
76
annotations = list(tree.annotate_iter(path))
78
return _show_id_annotations(annotations, to_file, full, encoding)
80
if not getattr(tree, "get_revision_id", False):
81
# Create a virtual revision to represent the current tree state.
82
# Should get some more pending commit attributes, like pending tags,
84
current_rev = Revision(CURRENT_REVISION)
85
current_rev.parent_ids = tree.get_parent_ids()
87
current_rev.committer = branch.get_config_stack().get('email')
88
except errors.NoWhoami:
89
current_rev.committer = 'local user'
90
current_rev.message = "?"
91
current_rev.timestamp = round(time.time(), 3)
92
current_rev.timezone = osutils.local_time_offset()
95
annotation = list(_expand_annotations(
96
annotations, branch, current_rev))
97
_print_annotations(annotation, verbose, to_file, full, encoding)
100
def _print_annotations(annotation, verbose, to_file, full, encoding):
101
"""Print annotations to to_file.
103
:param to_file: The file to output the annotation to.
104
:param verbose: Show all details rather than truncating to ensure
105
reasonable text width.
106
:param full: XXXX Not sure what this does.
108
if len(annotation) == 0:
109
max_origin_len = max_revno_len = 0
111
max_origin_len = max(len(x[1]) for x in annotation)
112
max_revno_len = max(len(x[0]) for x in annotation)
114
max_revno_len = min(max_revno_len, 12)
115
max_revno_len = max(max_revno_len, 3)
117
# Output the annotations
119
for (revno_str, author, date_str, line_rev_id, text) in annotation:
43
annotation = list(_annotate_file(branch, rev_id, file_id))
44
max_origin_len = max(len(origin) for origin in set(x[1] for x in annotation))
45
for (revno_str, author, date_str, line_rev_id, text ) in annotation:
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:
129
# GZ 2017-05-21: Writing both unicode annotation and bytes from file
130
# which the given to_file must cope with.
132
to_file.write('| %s\n' % (text.decode(encoding),))
136
def _show_id_annotations(annotations, to_file, full, encoding):
140
max_origin_len = max(len(origin) for origin, text in annotations)
141
for origin, text in annotations:
142
if full or last_rev_id != origin:
146
to_file.write('%*s | %s' % (
147
max_origin_len, this.decode('utf-8'), text.decode(encoding)))
152
def _expand_annotations(annotations, branch, current_rev=None):
153
"""Expand a file's annotations into command line UI ready tuples.
155
Each tuple includes detailed information, such as the author name, and date
156
string for the commit, rather than just the revision id.
158
:param annotations: The annotations to expand.
159
:param revision_id_to_revno: A map from id to revision numbers.
160
:param branch: A locked branch to query for revision details.
162
repository = branch.repository
163
revision_ids = set(o for o, t in annotations)
164
if current_rev is not None:
165
# This can probably become a function on MutableTree, get_revno_map
166
# there, or something.
167
last_revision = current_rev.revision_id
168
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
169
# XXX: The main difficulty is that we need to inject a single new node
170
# (current_rev) into the graph before it gets numbered, etc.
171
# Once KnownGraph gets an 'add_node()' function, we can use
172
# VF.get_known_graph_ancestry().
173
graph = repository.get_graph()
175
key: value for key, value in
176
graph.iter_ancestry(current_rev.parent_ids) if value is not None}
177
revision_graph = _strip_NULL_ghosts(revision_graph)
178
revision_graph[last_revision] = current_rev.parent_ids
179
merge_sorted_revisions = tsort.merge_sort(
184
revision_id_to_revno = {
186
for seq_num, rev_id, depth, revno, end_of_merge in
187
merge_sorted_revisions}
189
# TODO(jelmer): Only look up the revision ids that we need (i.e. those
190
# in revision_ids). Possibly add a HPSS call that can look those up
192
revision_id_to_revno = branch.get_revision_id_to_revno_map()
47
anno = '%5s %-*s %8s ' % (revno_str, max_origin_len, author, date_str)
49
anno = "%5s %-7s " % ( revno_str, author[:7] )
51
if anno.lstrip() == "" and full: anno = prevanno
52
print >>to_file, '%s| %s' % (anno, text)
55
def _annotate_file(branch, rev_id, file_id ):
57
rh = branch.revision_history()
58
w = branch.repository.weave_store.get_weave(file_id,
59
branch.repository.get_transaction())
193
60
last_origin = None
195
if CURRENT_REVISION in revision_ids:
196
revision_id_to_revno[CURRENT_REVISION] = (
197
"%d?" % (branch.revno() + 1),)
198
revisions[CURRENT_REVISION] = current_rev
201
repository.iter_revisions(revision_ids)
202
if entry[1] is not None)
203
for origin, text in annotations:
204
text = text.rstrip(b'\r\n')
61
for origin, text in w.annotate_iter(rev_id):
62
text = text.rstrip('\r\n')
205
63
if origin == last_origin:
206
(revno_str, author, date_str) = ('', '', '')
64
(revno_str, author, date_str) = ('','','')
208
66
last_origin = origin
209
if origin not in revisions:
210
(revno_str, author, date_str) = ('?', '?', '?')
67
if not branch.repository.has_revision(origin):
68
(revno_str, author, date_str) = ('?','?','?')
212
revno_str = '.'.join(
213
str(i) for i in revision_id_to_revno[origin])
214
rev = revisions[origin]
71
revno_str = str(rh.index(origin) + 1)
74
rev = branch.repository.get_revision(origin)
215
75
tz = rev.timezone or 0
216
date_str = time.strftime('%Y%m%d',
76
date_str = time.strftime('%Y%m%d',
217
77
time.gmtime(rev.timestamp + tz))
218
78
# a lazy way to get something like the email address
219
79
# TODO: Get real email address
220
author = rev.get_apparent_authors()[0]
221
_, email = config.parse_username(author)
80
author = rev.committer
82
author = extract_email_address(author)
84
pass # use the whole name
224
85
yield (revno_str, author, date_str, origin, text)
227
def reannotate(parents_lines, new_lines, new_revision_id,
228
_left_matching_blocks=None,
229
heads_provider=None):
230
"""Create a new annotated version from new lines and parent annotations.
232
:param parents_lines: List of annotated lines for all parents
233
:param new_lines: The un-annotated new lines
234
:param new_revision_id: The revision-id to associate with new lines
235
(will often be CURRENT_REVISION)
236
:param left_matching_blocks: a hint about which areas are common
237
between the text and its left-hand-parent. The format is
238
the SequenceMatcher.get_matching_blocks format
239
(start_left, start_right, length_of_match).
240
:param heads_provider: An object which provides a .heads() call to resolve
241
if any revision ids are children of others.
242
If None, then any ancestry disputes will be resolved with
245
if len(parents_lines) == 0:
246
lines = [(new_revision_id, line) for line in new_lines]
247
elif len(parents_lines) == 1:
248
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
249
_left_matching_blocks)
250
elif len(parents_lines) == 2:
251
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
252
_left_matching_blocks)
253
lines = _reannotate_annotated(parents_lines[1], new_lines,
254
new_revision_id, left,
257
reannotations = [_reannotate(parents_lines[0], new_lines,
258
new_revision_id, _left_matching_blocks)]
259
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
260
for p in parents_lines[1:])
262
for annos in zip(*reannotations):
263
origins = set(a for a, l in annos)
264
if len(origins) == 1:
265
# All the parents agree, so just return the first one
266
lines.append(annos[0])
269
if len(origins) == 2 and new_revision_id in origins:
270
origins.remove(new_revision_id)
271
if len(origins) == 1:
272
lines.append((origins.pop(), line))
274
lines.append((new_revision_id, line))
278
def _reannotate(parent_lines, new_lines, new_revision_id,
279
matching_blocks=None):
281
if matching_blocks is None:
282
plain_parent_lines = [l for r, l in parent_lines]
283
matcher = patiencediff.PatienceSequenceMatcher(
284
None, plain_parent_lines, new_lines)
285
matching_blocks = matcher.get_matching_blocks()
287
for i, j, n in matching_blocks:
288
for line in new_lines[new_cur:j]:
289
lines.append((new_revision_id, line))
290
lines.extend(parent_lines[i:i + n])
295
def _get_matching_blocks(old, new):
296
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
297
return matcher.get_matching_blocks()
300
_break_annotation_tie = None
303
def _old_break_annotation_tie(annotated_lines):
304
"""Chose an attribution between several possible ones.
306
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
307
the lines are identical but the revids different while no parent
308
relation exist between them
310
:return : The "winning" line. This must be one with a revid that
311
guarantees that further criss-cross merges will converge. Failing to
312
do so have performance implications.
314
# sort lexicographically so that we always get a stable result.
316
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
317
# with the current implementation, chosing the oldest revision is known to
318
# provide better results (as in matching user expectations). The most
319
# common use case being manual cherry-pick from an already existing
321
return sorted(annotated_lines)[0]
324
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
325
child_lines, start_child, end_child,
326
right_lines, start_right, end_right,
327
heads_provider, revision_id):
328
"""Find lines in plain_right_lines that match the existing lines.
330
:param output_lines: Append final annotated lines to this list
331
:param plain_child_lines: The unannotated new lines for the child text
332
:param child_lines: Lines for the child text which have been annotated
335
:param start_child: Position in plain_child_lines and child_lines to start
337
:param end_child: Last position in plain_child_lines and child_lines to
339
:param right_lines: The annotated lines for the whole text for the right
341
:param start_right: Position in right_lines to start the match
342
:param end_right: Last position in right_lines to search for a match
343
:param heads_provider: When parents disagree on the lineage of a line, we
344
need to check if one side supersedes the other
345
:param revision_id: The label to give if a line should be labeled 'tip'
347
output_extend = output_lines.extend
348
output_append = output_lines.append
349
# We need to see if any of the unannotated lines match
350
plain_right_subset = [l for a, l in right_lines[start_right:end_right]]
351
plain_child_subset = plain_child_lines[start_child:end_child]
352
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
356
for right_idx, child_idx, match_len in match_blocks:
357
# All the lines that don't match are just passed along
358
if child_idx > last_child_idx:
359
output_extend(child_lines[start_child + last_child_idx:
360
start_child + child_idx])
361
for offset in range(match_len):
362
left = child_lines[start_child + child_idx + offset]
363
right = right_lines[start_right + right_idx + offset]
364
if left[0] == right[0]:
365
# The annotations match, just return the left one
367
elif left[0] == revision_id:
368
# The left parent marked this as unmatched, so let the
369
# right parent claim it
372
# Left and Right both claim this line
373
if heads_provider is None:
374
output_append((revision_id, left[1]))
376
heads = heads_provider.heads((left[0], right[0]))
378
output_append((next(iter(heads)), left[1]))
380
# Both claim different origins, get a stable result.
381
# If the result is not stable, there is a risk a
382
# performance degradation as criss-cross merges will
383
# flip-flop the attribution.
384
if _break_annotation_tie is None:
386
_old_break_annotation_tie([left, right]))
388
output_append(_break_annotation_tie([left, right]))
389
last_child_idx = child_idx + match_len
392
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
393
annotated_lines, heads_provider):
394
"""Update the annotations for a node based on another parent.
396
:param right_parent_lines: A list of annotated lines for the right-hand
398
:param new_lines: The unannotated new lines.
399
:param new_revision_id: The revision_id to attribute to lines which are not
400
present in either parent.
401
:param annotated_lines: A list of annotated lines. This should be the
402
annotation of new_lines based on parents seen so far.
403
:param heads_provider: When parents disagree on the lineage of a line, we
404
need to check if one side supersedes the other.
406
if len(new_lines) != len(annotated_lines):
407
raise AssertionError("mismatched new_lines and annotated_lines")
408
# First compare the newly annotated lines with the right annotated lines.
409
# Lines which were not changed in left or right should match. This tends to
410
# be the bulk of the lines, and they will need no further processing.
412
lines_extend = lines.extend
413
# 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 breezy._annotator_pyx import Annotator
443
except ImportError as e:
444
osutils.failed_to_load_extension(e)
445
from breezy._annotator_py import Annotator # noqa: F401