1
# Copyright (C) 2005-2010 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
from __future__ import absolute_import
21
# TODO: Choice of more or less verbose formats:
23
# interposed: show more details between blocks of modified lines
25
# TODO: Show which revision caused a line to merge into the parent
27
# TODO: perhaps abbreviate timescales depending on how recent they are
28
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
33
from .lazy_import import lazy_import
34
lazy_import(globals(), """
47
extract_email_address,
49
from .repository import _strip_NULL_ghosts
50
from .revision import (
56
def annotate_file_tree(tree, path, to_file, verbose=False, full=False,
57
show_ids=False, branch=None, file_id=None):
58
"""Annotate file_id in a tree.
60
The tree should already be read_locked() when annotate_file_tree is called.
62
:param tree: The tree to look for revision numbers and history from.
63
:param path: The path to annotate
64
:param to_file: The file to output the annotation to.
65
:param verbose: Show all details rather than truncating to ensure
66
reasonable text width.
67
:param full: XXXX Not sure what this does.
68
:param show_ids: Show revision ids in the annotation output.
69
:param file_id: The file_id to annotate (must match file path)
70
:param branch: Branch to use for revision revno lookups
77
# Handle the show_ids case
78
annotations = list(tree.annotate_iter(path, file_id))
80
return _show_id_annotations(annotations, to_file, full)
82
if not getattr(tree, "get_revision_id", False):
83
# Create a virtual revision to represent the current tree state.
84
# Should get some more pending commit attributes, like pending tags,
86
current_rev = Revision(CURRENT_REVISION)
87
current_rev.parent_ids = tree.get_parent_ids()
89
current_rev.committer = branch.get_config_stack().get('email')
91
current_rev.committer = 'local user'
92
current_rev.message = "?"
93
current_rev.timestamp = round(time.time(), 3)
94
current_rev.timezone = osutils.local_time_offset()
97
annotation = list(_expand_annotations(annotations, branch,
99
_print_annotations(annotation, verbose, to_file, full)
102
def _print_annotations(annotation, verbose, to_file, full):
103
"""Print annotations to to_file.
105
:param to_file: The file to output the annotation to.
106
:param verbose: Show all details rather than truncating to ensure
107
reasonable text width.
108
:param full: XXXX Not sure what this does.
110
if len(annotation) == 0:
111
max_origin_len = max_revno_len = max_revid_len = 0
113
max_origin_len = max(len(x[1]) for x in annotation)
114
max_revno_len = max(len(x[0]) for x in annotation)
115
max_revid_len = max(len(x[3]) for x in annotation)
117
max_revno_len = min(max_revno_len, 12)
118
max_revno_len = max(max_revno_len, 3)
120
# Output the annotations
122
for (revno_str, author, date_str, line_rev_id, text) in annotation:
124
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
125
max_origin_len, author, date_str)
127
if len(revno_str) > max_revno_len:
128
revno_str = revno_str[:max_revno_len-1] + '>'
129
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
130
if anno.lstrip() == "" and full:
132
# GZ 2017-05-21: Writing both unicode annotation and bytes from file
133
# which the given to_file must cope with.
135
to_file.write('| %s\n' % (text,))
139
def _show_id_annotations(annotations, to_file, full):
143
max_origin_len = max(len(origin) for origin, text in annotations)
144
for origin, text in annotations:
145
if full or last_rev_id != origin:
149
to_file.write('%*s | %s' % (max_origin_len, this, text))
154
def _expand_annotations(annotations, branch, current_rev=None):
155
"""Expand a file's annotations into command line UI ready tuples.
157
Each tuple includes detailed information, such as the author name, and date
158
string for the commit, rather than just the revision id.
160
:param annotations: The annotations to expand.
161
:param revision_id_to_revno: A map from id to revision numbers.
162
:param branch: A locked branch to query for revision details.
164
repository = branch.repository
165
if current_rev is not None:
166
# This can probably become a function on MutableTree, get_revno_map
167
# there, or something.
168
last_revision = current_rev.revision_id
169
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
170
# XXX: The main difficulty is that we need to inject a single new node
171
# (current_rev) into the graph before it gets numbered, etc.
172
# Once KnownGraph gets an 'add_node()' function, we can use
173
# VF.get_known_graph_ancestry().
174
graph = repository.get_graph()
175
revision_graph = dict(((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 = dict((rev_id, revno)
185
for seq_num, rev_id, depth, revno, end_of_merge in
186
merge_sorted_revisions)
188
revision_id_to_revno = branch.get_revision_id_to_revno_map()
190
revision_ids = set(o for o, t in annotations)
192
if CURRENT_REVISION in revision_ids:
193
revision_id_to_revno[CURRENT_REVISION] = (
194
"%d?" % (branch.revno() + 1),)
195
revisions[CURRENT_REVISION] = current_rev
198
repository.iter_revisions(revision_ids)
199
if entry[1] is not None)
200
for origin, text in annotations:
201
text = text.rstrip('\r\n')
202
if origin == last_origin:
203
(revno_str, author, date_str) = ('', '', '')
206
if origin not in revisions:
207
(revno_str, author, date_str) = ('?', '?', '?')
209
revno_str = '.'.join(str(i) for i in
210
revision_id_to_revno[origin])
211
rev = revisions[origin]
212
tz = rev.timezone or 0
213
date_str = time.strftime('%Y%m%d',
214
time.gmtime(rev.timestamp + tz))
215
# a lazy way to get something like the email address
216
# TODO: Get real email address
217
author = rev.get_apparent_authors()[0]
219
author = extract_email_address(author)
220
except NoEmailInUsername:
221
pass # use the whole name
222
yield (revno_str, author, date_str, origin, text)
225
def reannotate(parents_lines, new_lines, new_revision_id,
226
_left_matching_blocks=None,
227
heads_provider=None):
228
"""Create a new annotated version from new lines and parent annotations.
230
:param parents_lines: List of annotated lines for all parents
231
:param new_lines: The un-annotated new lines
232
:param new_revision_id: The revision-id to associate with new lines
233
(will often be CURRENT_REVISION)
234
:param left_matching_blocks: a hint about which areas are common
235
between the text and its left-hand-parent. The format is
236
the SequenceMatcher.get_matching_blocks format
237
(start_left, start_right, length_of_match).
238
:param heads_provider: An object which provides a .heads() call to resolve
239
if any revision ids are children of others.
240
If None, then any ancestry disputes will be resolved with
243
if len(parents_lines) == 0:
244
lines = [(new_revision_id, line) for line in new_lines]
245
elif len(parents_lines) == 1:
246
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
247
_left_matching_blocks)
248
elif len(parents_lines) == 2:
249
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
250
_left_matching_blocks)
251
lines = _reannotate_annotated(parents_lines[1], new_lines,
252
new_revision_id, left,
255
reannotations = [_reannotate(parents_lines[0], new_lines,
256
new_revision_id, _left_matching_blocks)]
257
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
258
for p in parents_lines[1:])
260
for annos in zip(*reannotations):
261
origins = set(a for a, l in annos)
262
if len(origins) == 1:
263
# All the parents agree, so just return the first one
264
lines.append(annos[0])
267
if len(origins) == 2 and new_revision_id in origins:
268
origins.remove(new_revision_id)
269
if len(origins) == 1:
270
lines.append((origins.pop(), line))
272
lines.append((new_revision_id, line))
276
def _reannotate(parent_lines, new_lines, new_revision_id,
277
matching_blocks=None):
279
if matching_blocks is None:
280
plain_parent_lines = [l for r, l in parent_lines]
281
matcher = patiencediff.PatienceSequenceMatcher(None,
282
plain_parent_lines, new_lines)
283
matching_blocks = matcher.get_matching_blocks()
285
for i, j, n in matching_blocks:
286
for line in new_lines[new_cur:j]:
287
lines.append((new_revision_id, line))
288
lines.extend(parent_lines[i:i+n])
293
def _get_matching_blocks(old, new):
294
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
295
return matcher.get_matching_blocks()
298
_break_annotation_tie = None
300
def _old_break_annotation_tie(annotated_lines):
301
"""Chose an attribution between several possible ones.
303
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
304
the lines are identical but the revids different while no parent
305
relation exist between them
307
:return : The "winning" line. This must be one with a revid that
308
guarantees that further criss-cross merges will converge. Failing to
309
do so have performance implications.
311
# sort lexicographically so that we always get a stable result.
313
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
314
# with the current implementation, chosing the oldest revision is known to
315
# provide better results (as in matching user expectations). The most
316
# common use case being manual cherry-pick from an already existing
318
return sorted(annotated_lines)[0]
321
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
322
child_lines, start_child, end_child,
323
right_lines, start_right, end_right,
324
heads_provider, revision_id):
325
"""Find lines in plain_right_lines that match the existing lines.
327
:param output_lines: Append final annotated lines to this list
328
:param plain_child_lines: The unannotated new lines for the child text
329
:param child_lines: Lines for the child text which have been annotated
332
:param start_child: Position in plain_child_lines and child_lines to start
334
:param end_child: Last position in plain_child_lines and child_lines to
336
:param right_lines: The annotated lines for the whole text for the right
338
:param start_right: Position in right_lines to start the match
339
:param end_right: Last position in right_lines to search for a match
340
:param heads_provider: When parents disagree on the lineage of a line, we
341
need to check if one side supersedes the other
342
:param revision_id: The label to give if a line should be labeled 'tip'
344
output_extend = output_lines.extend
345
output_append = output_lines.append
346
# We need to see if any of the unannotated lines match
347
plain_right_subset = [l for a, l in right_lines[start_right:end_right]]
348
plain_child_subset = plain_child_lines[start_child:end_child]
349
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
353
for right_idx, child_idx, match_len in match_blocks:
354
# All the lines that don't match are just passed along
355
if child_idx > last_child_idx:
356
output_extend(child_lines[start_child + last_child_idx
357
:start_child + child_idx])
358
for offset in range(match_len):
359
left = child_lines[start_child+child_idx+offset]
360
right = right_lines[start_right+right_idx+offset]
361
if left[0] == right[0]:
362
# The annotations match, just return the left one
364
elif left[0] == revision_id:
365
# The left parent marked this as unmatched, so let the
366
# right parent claim it
369
# Left and Right both claim this line
370
if heads_provider is None:
371
output_append((revision_id, left[1]))
373
heads = heads_provider.heads((left[0], right[0]))
375
output_append((next(iter(heads)), left[1]))
377
# Both claim different origins, get a stable result.
378
# If the result is not stable, there is a risk a
379
# performance degradation as criss-cross merges will
380
# flip-flop the attribution.
381
if _break_annotation_tie is None:
383
_old_break_annotation_tie([left, right]))
385
output_append(_break_annotation_tie([left, right]))
386
last_child_idx = child_idx + match_len
389
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
390
annotated_lines, heads_provider):
391
"""Update the annotations for a node based on another parent.
393
:param right_parent_lines: A list of annotated lines for the right-hand
395
:param new_lines: The unannotated new lines.
396
:param new_revision_id: The revision_id to attribute to lines which are not
397
present in either parent.
398
:param annotated_lines: A list of annotated lines. This should be the
399
annotation of new_lines based on parents seen so far.
400
:param heads_provider: When parents disagree on the lineage of a line, we
401
need to check if one side supersedes the other.
403
if len(new_lines) != len(annotated_lines):
404
raise AssertionError("mismatched new_lines and annotated_lines")
405
# First compare the newly annotated lines with the right annotated lines.
406
# Lines which were not changed in left or right should match. This tends to
407
# be the bulk of the lines, and they will need no further processing.
409
lines_extend = lines.extend
410
last_right_idx = 0 # The line just after the last match from the right side
412
matching_left_and_right = _get_matching_blocks(right_parent_lines,
414
for right_idx, left_idx, match_len in matching_left_and_right:
415
# annotated lines from last_left_idx to left_idx did not match the
416
# lines from last_right_idx to right_idx, the raw lines should be
417
# compared to determine what annotations need to be updated
418
if last_right_idx == right_idx or last_left_idx == left_idx:
419
# One of the sides is empty, so this is a pure insertion
420
lines_extend(annotated_lines[last_left_idx:left_idx])
422
# We need to see if any of the unannotated lines match
423
_find_matching_unannotated_lines(lines,
424
new_lines, annotated_lines,
425
last_left_idx, left_idx,
427
last_right_idx, right_idx,
430
last_right_idx = right_idx + match_len
431
last_left_idx = left_idx + match_len
432
# If left and right agree on a range, just push that into the output
433
lines_extend(annotated_lines[left_idx:left_idx + match_len])
438
from breezy._annotator_pyx import Annotator
439
except ImportError as e:
440
osutils.failed_to_load_extension(e)
441
from breezy._annotator_py import Annotator