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
encoding = osutils.get_terminal_encoding()
78
# Handle the show_ids case
79
annotations = list(tree.annotate_iter(path, file_id))
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(annotations, branch,
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 = max_revid_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)
116
max_revid_len = max(len(x[3]) for x in annotation)
118
max_revno_len = min(max_revno_len, 12)
119
max_revno_len = max(max_revno_len, 3)
121
# Output the annotations
123
for (revno_str, author, date_str, line_rev_id, text) in annotation:
125
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
126
max_origin_len, author, date_str)
128
if len(revno_str) > max_revno_len:
129
revno_str = revno_str[:max_revno_len-1] + '>'
130
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
131
if anno.lstrip() == "" and full:
133
# GZ 2017-05-21: Writing both unicode annotation and bytes from file
134
# which the given to_file must cope with.
136
to_file.write('| %s\n' % (text.decode(encoding),))
140
def _show_id_annotations(annotations, to_file, full, encoding):
144
max_origin_len = max(len(origin) for origin, text in annotations)
145
for origin, text in annotations:
146
if full or last_rev_id != origin:
150
to_file.write('%*s | %s' % (max_origin_len, this.decode('utf-8'),
151
text.decode(encoding)))
156
def _expand_annotations(annotations, branch, current_rev=None):
157
"""Expand a file's annotations into command line UI ready tuples.
159
Each tuple includes detailed information, such as the author name, and date
160
string for the commit, rather than just the revision id.
162
:param annotations: The annotations to expand.
163
:param revision_id_to_revno: A map from id to revision numbers.
164
:param branch: A locked branch to query for revision details.
166
repository = branch.repository
167
revision_ids = set(o for o, t in annotations)
168
if current_rev is not None:
169
# This can probably become a function on MutableTree, get_revno_map
170
# there, or something.
171
last_revision = current_rev.revision_id
172
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
173
# XXX: The main difficulty is that we need to inject a single new node
174
# (current_rev) into the graph before it gets numbered, etc.
175
# Once KnownGraph gets an 'add_node()' function, we can use
176
# VF.get_known_graph_ancestry().
177
graph = repository.get_graph()
178
revision_graph = dict(((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 = dict((rev_id, revno)
188
for seq_num, rev_id, depth, revno, end_of_merge in
189
merge_sorted_revisions)
191
# TODO(jelmer): Only look up the revision ids that we need (i.e. those
192
# in revision_ids). Possibly add a HPSS call that can look those up
194
revision_id_to_revno = branch.get_revision_id_to_revno_map()
197
if CURRENT_REVISION in revision_ids:
198
revision_id_to_revno[CURRENT_REVISION] = (
199
"%d?" % (branch.revno() + 1),)
200
revisions[CURRENT_REVISION] = current_rev
203
repository.iter_revisions(revision_ids)
204
if entry[1] is not None)
205
for origin, text in annotations:
206
text = text.rstrip(b'\r\n')
207
if origin == last_origin:
208
(revno_str, author, date_str) = ('', '', '')
211
if origin not in revisions:
212
(revno_str, author, date_str) = ('?', '?', '?')
214
revno_str = '.'.join(str(i) for i in
215
revision_id_to_revno[origin])
216
rev = revisions[origin]
217
tz = rev.timezone or 0
218
date_str = time.strftime('%Y%m%d',
219
time.gmtime(rev.timestamp + tz))
220
# a lazy way to get something like the email address
221
# TODO: Get real email address
222
author = rev.get_apparent_authors()[0]
224
author = extract_email_address(author)
225
except NoEmailInUsername:
226
pass # use the whole name
227
yield (revno_str, author, date_str, origin, text)
230
def reannotate(parents_lines, new_lines, new_revision_id,
231
_left_matching_blocks=None,
232
heads_provider=None):
233
"""Create a new annotated version from new lines and parent annotations.
235
:param parents_lines: List of annotated lines for all parents
236
:param new_lines: The un-annotated new lines
237
:param new_revision_id: The revision-id to associate with new lines
238
(will often be CURRENT_REVISION)
239
:param left_matching_blocks: a hint about which areas are common
240
between the text and its left-hand-parent. The format is
241
the SequenceMatcher.get_matching_blocks format
242
(start_left, start_right, length_of_match).
243
:param heads_provider: An object which provides a .heads() call to resolve
244
if any revision ids are children of others.
245
If None, then any ancestry disputes will be resolved with
248
if len(parents_lines) == 0:
249
lines = [(new_revision_id, line) for line in new_lines]
250
elif len(parents_lines) == 1:
251
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
252
_left_matching_blocks)
253
elif len(parents_lines) == 2:
254
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
255
_left_matching_blocks)
256
lines = _reannotate_annotated(parents_lines[1], new_lines,
257
new_revision_id, left,
260
reannotations = [_reannotate(parents_lines[0], new_lines,
261
new_revision_id, _left_matching_blocks)]
262
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
263
for p in parents_lines[1:])
265
for annos in zip(*reannotations):
266
origins = set(a for a, l in annos)
267
if len(origins) == 1:
268
# All the parents agree, so just return the first one
269
lines.append(annos[0])
272
if len(origins) == 2 and new_revision_id in origins:
273
origins.remove(new_revision_id)
274
if len(origins) == 1:
275
lines.append((origins.pop(), line))
277
lines.append((new_revision_id, line))
281
def _reannotate(parent_lines, new_lines, new_revision_id,
282
matching_blocks=None):
284
if matching_blocks is None:
285
plain_parent_lines = [l for r, l in parent_lines]
286
matcher = patiencediff.PatienceSequenceMatcher(None,
287
plain_parent_lines, new_lines)
288
matching_blocks = matcher.get_matching_blocks()
290
for i, j, n in matching_blocks:
291
for line in new_lines[new_cur:j]:
292
lines.append((new_revision_id, line))
293
lines.extend(parent_lines[i:i+n])
298
def _get_matching_blocks(old, new):
299
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
300
return matcher.get_matching_blocks()
303
_break_annotation_tie = None
305
def _old_break_annotation_tie(annotated_lines):
306
"""Chose an attribution between several possible ones.
308
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
309
the lines are identical but the revids different while no parent
310
relation exist between them
312
:return : The "winning" line. This must be one with a revid that
313
guarantees that further criss-cross merges will converge. Failing to
314
do so have performance implications.
316
# sort lexicographically so that we always get a stable result.
318
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
319
# with the current implementation, chosing the oldest revision is known to
320
# provide better results (as in matching user expectations). The most
321
# common use case being manual cherry-pick from an already existing
323
return sorted(annotated_lines)[0]
326
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
327
child_lines, start_child, end_child,
328
right_lines, start_right, end_right,
329
heads_provider, revision_id):
330
"""Find lines in plain_right_lines that match the existing lines.
332
:param output_lines: Append final annotated lines to this list
333
:param plain_child_lines: The unannotated new lines for the child text
334
:param child_lines: Lines for the child text which have been annotated
337
:param start_child: Position in plain_child_lines and child_lines to start
339
:param end_child: Last position in plain_child_lines and child_lines to
341
:param right_lines: The annotated lines for the whole text for the right
343
:param start_right: Position in right_lines to start the match
344
:param end_right: Last position in right_lines to search for a match
345
:param heads_provider: When parents disagree on the lineage of a line, we
346
need to check if one side supersedes the other
347
:param revision_id: The label to give if a line should be labeled 'tip'
349
output_extend = output_lines.extend
350
output_append = output_lines.append
351
# We need to see if any of the unannotated lines match
352
plain_right_subset = [l for a, l in right_lines[start_right:end_right]]
353
plain_child_subset = plain_child_lines[start_child:end_child]
354
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
358
for right_idx, child_idx, match_len in match_blocks:
359
# All the lines that don't match are just passed along
360
if child_idx > last_child_idx:
361
output_extend(child_lines[start_child + last_child_idx
362
:start_child + child_idx])
363
for offset in range(match_len):
364
left = child_lines[start_child+child_idx+offset]
365
right = right_lines[start_right+right_idx+offset]
366
if left[0] == right[0]:
367
# The annotations match, just return the left one
369
elif left[0] == revision_id:
370
# The left parent marked this as unmatched, so let the
371
# right parent claim it
374
# Left and Right both claim this line
375
if heads_provider is None:
376
output_append((revision_id, left[1]))
378
heads = heads_provider.heads((left[0], right[0]))
380
output_append((next(iter(heads)), left[1]))
382
# Both claim different origins, get a stable result.
383
# If the result is not stable, there is a risk a
384
# performance degradation as criss-cross merges will
385
# flip-flop the attribution.
386
if _break_annotation_tie is None:
388
_old_break_annotation_tie([left, right]))
390
output_append(_break_annotation_tie([left, right]))
391
last_child_idx = child_idx + match_len
394
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
395
annotated_lines, heads_provider):
396
"""Update the annotations for a node based on another parent.
398
:param right_parent_lines: A list of annotated lines for the right-hand
400
:param new_lines: The unannotated new lines.
401
:param new_revision_id: The revision_id to attribute to lines which are not
402
present in either parent.
403
:param annotated_lines: A list of annotated lines. This should be the
404
annotation of new_lines based on parents seen so far.
405
:param heads_provider: When parents disagree on the lineage of a line, we
406
need to check if one side supersedes the other.
408
if len(new_lines) != len(annotated_lines):
409
raise AssertionError("mismatched new_lines and annotated_lines")
410
# First compare the newly annotated lines with the right annotated lines.
411
# Lines which were not changed in left or right should match. This tends to
412
# be the bulk of the lines, and they will need no further processing.
414
lines_extend = lines.extend
415
last_right_idx = 0 # The line just after the last match from the right side
417
matching_left_and_right = _get_matching_blocks(right_parent_lines,
419
for right_idx, left_idx, match_len in matching_left_and_right:
420
# annotated lines from last_left_idx to left_idx did not match the
421
# lines from last_right_idx to right_idx, the raw lines should be
422
# compared to determine what annotations need to be updated
423
if last_right_idx == right_idx or last_left_idx == left_idx:
424
# One of the sides is empty, so this is a pure insertion
425
lines_extend(annotated_lines[last_left_idx:left_idx])
427
# We need to see if any of the unannotated lines match
428
_find_matching_unannotated_lines(lines,
429
new_lines, annotated_lines,
430
last_left_idx, left_idx,
432
last_right_idx, right_idx,
435
last_right_idx = right_idx + match_len
436
last_left_idx = left_idx + match_len
437
# If left and right agree on a range, just push that into the output
438
lines_extend(annotated_lines[left_idx:left_idx + match_len])
443
from breezy._annotator_pyx import Annotator
444
except ImportError as e:
445
osutils.failed_to_load_extension(e)
446
from breezy._annotator_py import Annotator
added
removed
breezy/annotate.py
