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(), """
48
extract_email_address,
50
from .repository import _strip_NULL_ghosts
51
from .revision import (
57
def annotate_file_tree(tree, path, to_file, verbose=False, full=False,
58
show_ids=False, branch=None):
59
"""Annotate path in a tree.
61
The tree should already be read_locked() when annotate_file_tree is called.
63
:param tree: The tree to look for revision numbers and history from.
64
:param path: The path to annotate
65
:param to_file: The file to output the annotation to.
66
:param verbose: Show all details rather than truncating to ensure
67
reasonable text width.
68
:param full: XXXX Not sure what this does.
69
:param show_ids: Show revision ids in the annotation output.
70
:param branch: Branch to use for revision revno lookups
77
encoding = osutils.get_terminal_encoding()
78
# Handle the show_ids case
79
annotations = list(tree.annotate_iter(path))
81
return _show_id_annotations(annotations, to_file, full, encoding)
83
if not getattr(tree, "get_revision_id", False):
84
# Create a virtual revision to represent the current tree state.
85
# Should get some more pending commit attributes, like pending tags,
87
current_rev = Revision(CURRENT_REVISION)
88
current_rev.parent_ids = tree.get_parent_ids()
90
current_rev.committer = branch.get_config_stack().get('email')
92
current_rev.committer = 'local user'
93
current_rev.message = "?"
94
current_rev.timestamp = round(time.time(), 3)
95
current_rev.timezone = osutils.local_time_offset()
98
annotation = list(_expand_annotations(
99
annotations, branch, current_rev))
100
_print_annotations(annotation, verbose, to_file, full, encoding)
103
def _print_annotations(annotation, verbose, to_file, full, encoding):
104
"""Print annotations to to_file.
106
:param to_file: The file to output the annotation to.
107
:param verbose: Show all details rather than truncating to ensure
108
reasonable text width.
109
:param full: XXXX Not sure what this does.
111
if len(annotation) == 0:
112
max_origin_len = max_revno_len = 0
114
max_origin_len = max(len(x[1]) for x in annotation)
115
max_revno_len = max(len(x[0]) 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.decode(encoding),))
139
def _show_id_annotations(annotations, to_file, full, encoding):
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' % (
150
max_origin_len, this.decode('utf-8'), text.decode(encoding)))
155
def _expand_annotations(annotations, branch, current_rev=None):
156
"""Expand a file's annotations into command line UI ready tuples.
158
Each tuple includes detailed information, such as the author name, and date
159
string for the commit, rather than just the revision id.
161
:param annotations: The annotations to expand.
162
:param revision_id_to_revno: A map from id to revision numbers.
163
:param branch: A locked branch to query for revision details.
165
repository = branch.repository
166
revision_ids = set(o for o, t in annotations)
167
if current_rev is not None:
168
# This can probably become a function on MutableTree, get_revno_map
169
# there, or something.
170
last_revision = current_rev.revision_id
171
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
172
# XXX: The main difficulty is that we need to inject a single new node
173
# (current_rev) into the graph before it gets numbered, etc.
174
# Once KnownGraph gets an 'add_node()' function, we can use
175
# VF.get_known_graph_ancestry().
176
graph = repository.get_graph()
178
key: value for key, value in
179
graph.iter_ancestry(current_rev.parent_ids) if value is not None}
180
revision_graph = _strip_NULL_ghosts(revision_graph)
181
revision_graph[last_revision] = current_rev.parent_ids
182
merge_sorted_revisions = tsort.merge_sort(
187
revision_id_to_revno = {
189
for seq_num, rev_id, depth, revno, end_of_merge in
190
merge_sorted_revisions}
192
# TODO(jelmer): Only look up the revision ids that we need (i.e. those
193
# in revision_ids). Possibly add a HPSS call that can look those up
195
revision_id_to_revno = branch.get_revision_id_to_revno_map()
198
if CURRENT_REVISION in revision_ids:
199
revision_id_to_revno[CURRENT_REVISION] = (
200
"%d?" % (branch.revno() + 1),)
201
revisions[CURRENT_REVISION] = current_rev
204
repository.iter_revisions(revision_ids)
205
if entry[1] is not None)
206
for origin, text in annotations:
207
text = text.rstrip(b'\r\n')
208
if origin == last_origin:
209
(revno_str, author, date_str) = ('', '', '')
212
if origin not in revisions:
213
(revno_str, author, date_str) = ('?', '?', '?')
215
revno_str = '.'.join(
216
str(i) for i in revision_id_to_revno[origin])
217
rev = revisions[origin]
218
tz = rev.timezone or 0
219
date_str = time.strftime('%Y%m%d',
220
time.gmtime(rev.timestamp + tz))
221
# a lazy way to get something like the email address
222
# TODO: Get real email address
223
author = rev.get_apparent_authors()[0]
225
author = extract_email_address(author)
226
except NoEmailInUsername:
227
pass # use the whole name
228
yield (revno_str, author, date_str, origin, text)
231
def reannotate(parents_lines, new_lines, new_revision_id,
232
_left_matching_blocks=None,
233
heads_provider=None):
234
"""Create a new annotated version from new lines and parent annotations.
236
:param parents_lines: List of annotated lines for all parents
237
:param new_lines: The un-annotated new lines
238
:param new_revision_id: The revision-id to associate with new lines
239
(will often be CURRENT_REVISION)
240
:param left_matching_blocks: a hint about which areas are common
241
between the text and its left-hand-parent. The format is
242
the SequenceMatcher.get_matching_blocks format
243
(start_left, start_right, length_of_match).
244
:param heads_provider: An object which provides a .heads() call to resolve
245
if any revision ids are children of others.
246
If None, then any ancestry disputes will be resolved with
249
if len(parents_lines) == 0:
250
lines = [(new_revision_id, line) for line in new_lines]
251
elif len(parents_lines) == 1:
252
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
253
_left_matching_blocks)
254
elif len(parents_lines) == 2:
255
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
256
_left_matching_blocks)
257
lines = _reannotate_annotated(parents_lines[1], new_lines,
258
new_revision_id, left,
261
reannotations = [_reannotate(parents_lines[0], new_lines,
262
new_revision_id, _left_matching_blocks)]
263
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
264
for p in parents_lines[1:])
266
for annos in zip(*reannotations):
267
origins = set(a for a, l in annos)
268
if len(origins) == 1:
269
# All the parents agree, so just return the first one
270
lines.append(annos[0])
273
if len(origins) == 2 and new_revision_id in origins:
274
origins.remove(new_revision_id)
275
if len(origins) == 1:
276
lines.append((origins.pop(), line))
278
lines.append((new_revision_id, line))
282
def _reannotate(parent_lines, new_lines, new_revision_id,
283
matching_blocks=None):
285
if matching_blocks is None:
286
plain_parent_lines = [l for r, l in parent_lines]
287
matcher = patiencediff.PatienceSequenceMatcher(
288
None, plain_parent_lines, new_lines)
289
matching_blocks = matcher.get_matching_blocks()
291
for i, j, n in matching_blocks:
292
for line in new_lines[new_cur:j]:
293
lines.append((new_revision_id, line))
294
lines.extend(parent_lines[i:i + n])
299
def _get_matching_blocks(old, new):
300
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
301
return matcher.get_matching_blocks()
304
_break_annotation_tie = None
307
def _old_break_annotation_tie(annotated_lines):
308
"""Chose an attribution between several possible ones.
310
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
311
the lines are identical but the revids different while no parent
312
relation exist between them
314
:return : The "winning" line. This must be one with a revid that
315
guarantees that further criss-cross merges will converge. Failing to
316
do so have performance implications.
318
# sort lexicographically so that we always get a stable result.
320
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
321
# with the current implementation, chosing the oldest revision is known to
322
# provide better results (as in matching user expectations). The most
323
# common use case being manual cherry-pick from an already existing
325
return sorted(annotated_lines)[0]
328
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
329
child_lines, start_child, end_child,
330
right_lines, start_right, end_right,
331
heads_provider, revision_id):
332
"""Find lines in plain_right_lines that match the existing lines.
334
:param output_lines: Append final annotated lines to this list
335
:param plain_child_lines: The unannotated new lines for the child text
336
:param child_lines: Lines for the child text which have been annotated
339
:param start_child: Position in plain_child_lines and child_lines to start
341
:param end_child: Last position in plain_child_lines and child_lines to
343
:param right_lines: The annotated lines for the whole text for the right
345
:param start_right: Position in right_lines to start the match
346
:param end_right: Last position in right_lines to search for a match
347
:param heads_provider: When parents disagree on the lineage of a line, we
348
need to check if one side supersedes the other
349
:param revision_id: The label to give if a line should be labeled 'tip'
351
output_extend = output_lines.extend
352
output_append = output_lines.append
353
# We need to see if any of the unannotated lines match
354
plain_right_subset = [l for a, l in right_lines[start_right:end_right]]
355
plain_child_subset = plain_child_lines[start_child:end_child]
356
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
360
for right_idx, child_idx, match_len in match_blocks:
361
# All the lines that don't match are just passed along
362
if child_idx > last_child_idx:
363
output_extend(child_lines[start_child + last_child_idx:
364
start_child + child_idx])
365
for offset in range(match_len):
366
left = child_lines[start_child + child_idx + offset]
367
right = right_lines[start_right + right_idx + offset]
368
if left[0] == right[0]:
369
# The annotations match, just return the left one
371
elif left[0] == revision_id:
372
# The left parent marked this as unmatched, so let the
373
# right parent claim it
376
# Left and Right both claim this line
377
if heads_provider is None:
378
output_append((revision_id, left[1]))
380
heads = heads_provider.heads((left[0], right[0]))
382
output_append((next(iter(heads)), left[1]))
384
# Both claim different origins, get a stable result.
385
# If the result is not stable, there is a risk a
386
# performance degradation as criss-cross merges will
387
# flip-flop the attribution.
388
if _break_annotation_tie is None:
390
_old_break_annotation_tie([left, right]))
392
output_append(_break_annotation_tie([left, right]))
393
last_child_idx = child_idx + match_len
396
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
397
annotated_lines, heads_provider):
398
"""Update the annotations for a node based on another parent.
400
:param right_parent_lines: A list of annotated lines for the right-hand
402
:param new_lines: The unannotated new lines.
403
:param new_revision_id: The revision_id to attribute to lines which are not
404
present in either parent.
405
:param annotated_lines: A list of annotated lines. This should be the
406
annotation of new_lines based on parents seen so far.
407
:param heads_provider: When parents disagree on the lineage of a line, we
408
need to check if one side supersedes the other.
410
if len(new_lines) != len(annotated_lines):
411
raise AssertionError("mismatched new_lines and annotated_lines")
412
# First compare the newly annotated lines with the right annotated lines.
413
# Lines which were not changed in left or right should match. This tends to
414
# be the bulk of the lines, and they will need no further processing.
416
lines_extend = lines.extend
417
# The line just after the last match from the right side
420
matching_left_and_right = _get_matching_blocks(right_parent_lines,
422
for right_idx, left_idx, match_len in matching_left_and_right:
423
# annotated lines from last_left_idx to left_idx did not match the
424
# lines from last_right_idx to right_idx, the raw lines should be
425
# compared to determine what annotations need to be updated
426
if last_right_idx == right_idx or last_left_idx == left_idx:
427
# One of the sides is empty, so this is a pure insertion
428
lines_extend(annotated_lines[last_left_idx:left_idx])
430
# We need to see if any of the unannotated lines match
431
_find_matching_unannotated_lines(lines,
432
new_lines, annotated_lines,
433
last_left_idx, left_idx,
435
last_right_idx, right_idx,
438
last_right_idx = right_idx + match_len
439
last_left_idx = left_idx + match_len
440
# If left and right agree on a range, just push that into the output
441
lines_extend(annotated_lines[left_idx:left_idx + match_len])
446
from breezy._annotator_pyx import Annotator
447
except ImportError as e:
448
osutils.failed_to_load_extension(e)
449
from breezy._annotator_py import Annotator # noqa: F401
added
removed
breezy/annotate.py
