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, file_id, to_file, verbose=False, full=False,
57
show_ids=False, branch=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 file_id: The file_id 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 branch: Branch to use for revision revno lookups
76
# Handle the show_ids case
77
annotations = list(tree.annotate_iter(file_id))
79
return _show_id_annotations(annotations, to_file, full)
81
if not getattr(tree, "get_revision_id", False):
82
# Create a virtual revision to represent the current tree state.
83
# Should get some more pending commit attributes, like pending tags,
85
current_rev = Revision(CURRENT_REVISION)
86
current_rev.parent_ids = tree.get_parent_ids()
88
current_rev.committer = branch.get_config_stack().get('email')
90
current_rev.committer = 'local user'
91
current_rev.message = "?"
92
current_rev.timestamp = round(time.time(), 3)
93
current_rev.timezone = osutils.local_time_offset()
96
annotation = list(_expand_annotations(annotations, branch,
98
_print_annotations(annotation, verbose, to_file, full)
101
def _print_annotations(annotation, verbose, to_file, full):
102
"""Print annotations to to_file.
104
:param to_file: The file to output the annotation to.
105
:param verbose: Show all details rather than truncating to ensure
106
reasonable text width.
107
:param full: XXXX Not sure what this does.
109
if len(annotation) == 0:
110
max_origin_len = max_revno_len = max_revid_len = 0
112
max_origin_len = max(len(x[1]) for x in annotation)
113
max_revno_len = max(len(x[0]) for x in annotation)
114
max_revid_len = max(len(x[3]) for x in annotation)
116
max_revno_len = min(max_revno_len, 12)
117
max_revno_len = max(max_revno_len, 3)
119
# Output the annotations
121
for (revno_str, author, date_str, line_rev_id, text) in annotation:
123
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
124
max_origin_len, author, date_str)
126
if len(revno_str) > max_revno_len:
127
revno_str = revno_str[:max_revno_len-1] + '>'
128
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
129
if anno.lstrip() == "" and full:
131
# GZ 2017-05-21: Writing both unicode annotation and bytes from file
132
# which the given to_file must cope with.
134
to_file.write('| %s\n' % (text,))
138
def _show_id_annotations(annotations, to_file, full):
142
max_origin_len = max(len(origin) for origin, text in annotations)
143
for origin, text in annotations:
144
if full or last_rev_id != origin:
148
to_file.write('%*s | %s' % (max_origin_len, this, text))
153
def _expand_annotations(annotations, branch, current_rev=None):
154
"""Expand a file's annotations into command line UI ready tuples.
156
Each tuple includes detailed information, such as the author name, and date
157
string for the commit, rather than just the revision id.
159
:param annotations: The annotations to expand.
160
:param revision_id_to_revno: A map from id to revision numbers.
161
:param branch: A locked branch to query for revision details.
163
repository = branch.repository
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()
174
revision_graph = dict(((key, value) for key, value in
175
graph.iter_ancestry(current_rev.parent_ids) if value is not None))
176
revision_graph = _strip_NULL_ghosts(revision_graph)
177
revision_graph[last_revision] = current_rev.parent_ids
178
merge_sorted_revisions = tsort.merge_sort(
183
revision_id_to_revno = dict((rev_id, revno)
184
for seq_num, rev_id, depth, revno, end_of_merge in
185
merge_sorted_revisions)
187
revision_id_to_revno = branch.get_revision_id_to_revno_map()
189
revision_ids = set(o for o, t in annotations)
191
if CURRENT_REVISION in revision_ids:
192
revision_id_to_revno[CURRENT_REVISION] = (
193
"%d?" % (branch.revno() + 1),)
194
revisions[CURRENT_REVISION] = current_rev
197
repository.iter_revisions(revision_ids)
198
if entry[1] is not None)
199
for origin, text in annotations:
200
text = text.rstrip('\r\n')
201
if origin == last_origin:
202
(revno_str, author, date_str) = ('','','')
205
if origin not in revisions:
206
(revno_str, author, date_str) = ('?','?','?')
208
revno_str = '.'.join(str(i) for i in
209
revision_id_to_revno[origin])
210
rev = revisions[origin]
211
tz = rev.timezone or 0
212
date_str = time.strftime('%Y%m%d',
213
osutils.gmtime(rev.timestamp + tz))
214
# a lazy way to get something like the email address
215
# TODO: Get real email address
216
author = rev.get_apparent_authors()[0]
218
author = extract_email_address(author)
219
except NoEmailInUsername:
220
pass # use the whole name
221
yield (revno_str, author, date_str, origin, text)
224
def reannotate(parents_lines, new_lines, new_revision_id,
225
_left_matching_blocks=None,
226
heads_provider=None):
227
"""Create a new annotated version from new lines and parent annotations.
229
:param parents_lines: List of annotated lines for all parents
230
:param new_lines: The un-annotated new lines
231
:param new_revision_id: The revision-id to associate with new lines
232
(will often be CURRENT_REVISION)
233
:param left_matching_blocks: a hint about which areas are common
234
between the text and its left-hand-parent. The format is
235
the SequenceMatcher.get_matching_blocks format
236
(start_left, start_right, length_of_match).
237
:param heads_provider: An object which provides a .heads() call to resolve
238
if any revision ids are children of others.
239
If None, then any ancestry disputes will be resolved with
242
if len(parents_lines) == 0:
243
lines = [(new_revision_id, line) for line in new_lines]
244
elif len(parents_lines) == 1:
245
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
246
_left_matching_blocks)
247
elif len(parents_lines) == 2:
248
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
249
_left_matching_blocks)
250
lines = _reannotate_annotated(parents_lines[1], new_lines,
251
new_revision_id, left,
254
reannotations = [_reannotate(parents_lines[0], new_lines,
255
new_revision_id, _left_matching_blocks)]
256
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
257
for p in parents_lines[1:])
259
for annos in zip(*reannotations):
260
origins = set(a for a, l in annos)
261
if len(origins) == 1:
262
# All the parents agree, so just return the first one
263
lines.append(annos[0])
266
if len(origins) == 2 and new_revision_id in origins:
267
origins.remove(new_revision_id)
268
if len(origins) == 1:
269
lines.append((origins.pop(), line))
271
lines.append((new_revision_id, line))
275
def _reannotate(parent_lines, new_lines, new_revision_id,
276
matching_blocks=None):
278
if matching_blocks is None:
279
plain_parent_lines = [l for r, l in parent_lines]
280
matcher = patiencediff.PatienceSequenceMatcher(None,
281
plain_parent_lines, new_lines)
282
matching_blocks = matcher.get_matching_blocks()
284
for i, j, n in matching_blocks:
285
for line in new_lines[new_cur:j]:
286
lines.append((new_revision_id, line))
287
lines.extend(parent_lines[i:i+n])
292
def _get_matching_blocks(old, new):
293
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
294
return matcher.get_matching_blocks()
297
_break_annotation_tie = None
299
def _old_break_annotation_tie(annotated_lines):
300
"""Chose an attribution between several possible ones.
302
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
303
the lines are identical but the revids different while no parent
304
relation exist between them
306
:return : The "winning" line. This must be one with a revid that
307
guarantees that further criss-cross merges will converge. Failing to
308
do so have performance implications.
310
# sort lexicographically so that we always get a stable result.
312
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
313
# with the current implementation, chosing the oldest revision is known to
314
# provide better results (as in matching user expectations). The most
315
# common use case being manual cherry-pick from an already existing
317
return sorted(annotated_lines)[0]
320
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
321
child_lines, start_child, end_child,
322
right_lines, start_right, end_right,
323
heads_provider, revision_id):
324
"""Find lines in plain_right_lines that match the existing lines.
326
:param output_lines: Append final annotated lines to this list
327
:param plain_child_lines: The unannotated new lines for the child text
328
:param child_lines: Lines for the child text which have been annotated
331
:param start_child: Position in plain_child_lines and child_lines to start
333
:param end_child: Last position in plain_child_lines and child_lines to
335
:param right_lines: The annotated lines for the whole text for the right
337
:param start_right: Position in right_lines to start the match
338
:param end_right: Last position in right_lines to search for a match
339
:param heads_provider: When parents disagree on the lineage of a line, we
340
need to check if one side supersedes the other
341
:param revision_id: The label to give if a line should be labeled 'tip'
343
output_extend = output_lines.extend
344
output_append = output_lines.append
345
# We need to see if any of the unannotated lines match
346
plain_right_subset = [l for a,l in right_lines[start_right:end_right]]
347
plain_child_subset = plain_child_lines[start_child:end_child]
348
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
352
for right_idx, child_idx, match_len in match_blocks:
353
# All the lines that don't match are just passed along
354
if child_idx > last_child_idx:
355
output_extend(child_lines[start_child + last_child_idx
356
:start_child + child_idx])
357
for offset in range(match_len):
358
left = child_lines[start_child+child_idx+offset]
359
right = right_lines[start_right+right_idx+offset]
360
if left[0] == right[0]:
361
# The annotations match, just return the left one
363
elif left[0] == revision_id:
364
# The left parent marked this as unmatched, so let the
365
# right parent claim it
368
# Left and Right both claim this line
369
if heads_provider is None:
370
output_append((revision_id, left[1]))
372
heads = heads_provider.heads((left[0], right[0]))
374
output_append((next(iter(heads)), left[1]))
376
# Both claim different origins, get a stable result.
377
# If the result is not stable, there is a risk a
378
# performance degradation as criss-cross merges will
379
# flip-flop the attribution.
380
if _break_annotation_tie is None:
382
_old_break_annotation_tie([left, right]))
384
output_append(_break_annotation_tie([left, right]))
385
last_child_idx = child_idx + match_len
388
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
389
annotated_lines, heads_provider):
390
"""Update the annotations for a node based on another parent.
392
:param right_parent_lines: A list of annotated lines for the right-hand
394
:param new_lines: The unannotated new lines.
395
:param new_revision_id: The revision_id to attribute to lines which are not
396
present in either parent.
397
:param annotated_lines: A list of annotated lines. This should be the
398
annotation of new_lines based on parents seen so far.
399
:param heads_provider: When parents disagree on the lineage of a line, we
400
need to check if one side supersedes the other.
402
if len(new_lines) != len(annotated_lines):
403
raise AssertionError("mismatched new_lines and annotated_lines")
404
# First compare the newly annotated lines with the right annotated lines.
405
# Lines which were not changed in left or right should match. This tends to
406
# be the bulk of the lines, and they will need no further processing.
408
lines_extend = lines.extend
409
last_right_idx = 0 # The line just after the last match from the right side
411
matching_left_and_right = _get_matching_blocks(right_parent_lines,
413
for right_idx, left_idx, match_len in matching_left_and_right:
414
# annotated lines from last_left_idx to left_idx did not match the
415
# lines from last_right_idx to right_idx, the raw lines should be
416
# compared to determine what annotations need to be updated
417
if last_right_idx == right_idx or last_left_idx == left_idx:
418
# One of the sides is empty, so this is a pure insertion
419
lines_extend(annotated_lines[last_left_idx:left_idx])
421
# We need to see if any of the unannotated lines match
422
_find_matching_unannotated_lines(lines,
423
new_lines, annotated_lines,
424
last_left_idx, left_idx,
426
last_right_idx, right_idx,
429
last_right_idx = right_idx + match_len
430
last_left_idx = left_idx + match_len
431
# If left and right agree on a range, just push that into the output
432
lines_extend(annotated_lines[left_idx:left_idx + match_len])
437
from breezy._annotator_pyx import Annotator
438
except ImportError as e:
439
osutils.failed_to_load_extension(e)
440
from breezy._annotator_py import Annotator