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(), """
46
extract_email_address,
48
from .repository import _strip_NULL_ghosts
49
from .revision import (
55
def annotate_file_tree(tree, path, to_file, verbose=False, full=False,
56
show_ids=False, branch=None):
57
"""Annotate file_id in a tree.
59
The tree should already be read_locked() when annotate_file_tree is called.
61
:param tree: The tree to look for revision numbers and history from.
62
:param path: The path to annotate
63
:param to_file: The file to output the annotation to.
64
:param verbose: Show all details rather than truncating to ensure
65
reasonable text width.
66
:param full: XXXX Not sure what this does.
67
:param show_ids: Show revision ids in the annotation output.
68
:param branch: Branch to use for revision revno lookups
75
encoding = osutils.get_terminal_encoding()
76
# Handle the show_ids case
77
annotations = list(tree.annotate_iter(path))
79
return _show_id_annotations(annotations, to_file, full, encoding)
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(
97
annotations, branch, current_rev))
98
_print_annotations(annotation, verbose, to_file, full, encoding)
101
def _print_annotations(annotation, verbose, to_file, full, encoding):
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 = 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)
115
max_revno_len = min(max_revno_len, 12)
116
max_revno_len = max(max_revno_len, 3)
118
# Output the annotations
120
for (revno_str, author, date_str, line_rev_id, text) in annotation:
122
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
123
max_origin_len, author, date_str)
125
if len(revno_str) > max_revno_len:
126
revno_str = revno_str[:max_revno_len - 1] + '>'
127
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
128
if anno.lstrip() == "" and full:
130
# GZ 2017-05-21: Writing both unicode annotation and bytes from file
131
# which the given to_file must cope with.
133
to_file.write('| %s\n' % (text.decode(encoding),))
137
def _show_id_annotations(annotations, to_file, full, encoding):
141
max_origin_len = max(len(origin) for origin, text in annotations)
142
for origin, text in annotations:
143
if full or last_rev_id != origin:
147
to_file.write('%*s | %s' % (
148
max_origin_len, this.decode('utf-8'), text.decode(encoding)))
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
revision_ids = set(o for o, t in annotations)
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()
176
key: value for key, value in
177
graph.iter_ancestry(current_rev.parent_ids) if value is not None}
178
revision_graph = _strip_NULL_ghosts(revision_graph)
179
revision_graph[last_revision] = current_rev.parent_ids
180
merge_sorted_revisions = tsort.merge_sort(
185
revision_id_to_revno = {
187
for seq_num, rev_id, depth, revno, end_of_merge in
188
merge_sorted_revisions}
190
# TODO(jelmer): Only look up the revision ids that we need (i.e. those
191
# in revision_ids). Possibly add a HPSS call that can look those up
193
revision_id_to_revno = branch.get_revision_id_to_revno_map()
196
if CURRENT_REVISION in revision_ids:
197
revision_id_to_revno[CURRENT_REVISION] = (
198
"%d?" % (branch.revno() + 1),)
199
revisions[CURRENT_REVISION] = current_rev
202
repository.iter_revisions(revision_ids)
203
if entry[1] is not None)
204
for origin, text in annotations:
205
text = text.rstrip(b'\r\n')
206
if origin == last_origin:
207
(revno_str, author, date_str) = ('', '', '')
210
if origin not in revisions:
211
(revno_str, author, date_str) = ('?', '?', '?')
213
revno_str = '.'.join(
214
str(i) for i in revision_id_to_revno[origin])
215
rev = revisions[origin]
216
tz = rev.timezone or 0
217
date_str = time.strftime('%Y%m%d',
218
time.gmtime(rev.timestamp + tz))
219
# a lazy way to get something like the email address
220
# TODO: Get real email address
221
author = rev.get_apparent_authors()[0]
223
author = extract_email_address(author)
224
except NoEmailInUsername:
225
pass # use the whole name
226
yield (revno_str, author, date_str, origin, text)
229
def reannotate(parents_lines, new_lines, new_revision_id,
230
_left_matching_blocks=None,
231
heads_provider=None):
232
"""Create a new annotated version from new lines and parent annotations.
234
:param parents_lines: List of annotated lines for all parents
235
:param new_lines: The un-annotated new lines
236
:param new_revision_id: The revision-id to associate with new lines
237
(will often be CURRENT_REVISION)
238
:param left_matching_blocks: a hint about which areas are common
239
between the text and its left-hand-parent. The format is
240
the SequenceMatcher.get_matching_blocks format
241
(start_left, start_right, length_of_match).
242
:param heads_provider: An object which provides a .heads() call to resolve
243
if any revision ids are children of others.
244
If None, then any ancestry disputes will be resolved with
247
if len(parents_lines) == 0:
248
lines = [(new_revision_id, line) for line in new_lines]
249
elif len(parents_lines) == 1:
250
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
251
_left_matching_blocks)
252
elif len(parents_lines) == 2:
253
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
254
_left_matching_blocks)
255
lines = _reannotate_annotated(parents_lines[1], new_lines,
256
new_revision_id, left,
259
reannotations = [_reannotate(parents_lines[0], new_lines,
260
new_revision_id, _left_matching_blocks)]
261
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
262
for p in parents_lines[1:])
264
for annos in zip(*reannotations):
265
origins = set(a for a, l in annos)
266
if len(origins) == 1:
267
# All the parents agree, so just return the first one
268
lines.append(annos[0])
271
if len(origins) == 2 and new_revision_id in origins:
272
origins.remove(new_revision_id)
273
if len(origins) == 1:
274
lines.append((origins.pop(), line))
276
lines.append((new_revision_id, line))
280
def _reannotate(parent_lines, new_lines, new_revision_id,
281
matching_blocks=None):
283
if matching_blocks is None:
284
plain_parent_lines = [l for r, l in parent_lines]
285
matcher = patiencediff.PatienceSequenceMatcher(
286
None, plain_parent_lines, new_lines)
287
matching_blocks = matcher.get_matching_blocks()
289
for i, j, n in matching_blocks:
290
for line in new_lines[new_cur:j]:
291
lines.append((new_revision_id, line))
292
lines.extend(parent_lines[i:i + n])
297
def _get_matching_blocks(old, new):
298
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
299
return matcher.get_matching_blocks()
302
_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
# The line just after the last match from the right side
418
matching_left_and_right = _get_matching_blocks(right_parent_lines,
420
for right_idx, left_idx, match_len in matching_left_and_right:
421
# annotated lines from last_left_idx to left_idx did not match the
422
# lines from last_right_idx to right_idx, the raw lines should be
423
# compared to determine what annotations need to be updated
424
if last_right_idx == right_idx or last_left_idx == left_idx:
425
# One of the sides is empty, so this is a pure insertion
426
lines_extend(annotated_lines[last_left_idx:left_idx])
428
# We need to see if any of the unannotated lines match
429
_find_matching_unannotated_lines(lines,
430
new_lines, annotated_lines,
431
last_left_idx, left_idx,
433
last_right_idx, right_idx,
436
last_right_idx = right_idx + match_len
437
last_left_idx = left_idx + match_len
438
# If left and right agree on a range, just push that into the output
439
lines_extend(annotated_lines[left_idx:left_idx + match_len])
444
from breezy._annotator_pyx import Annotator
445
except ImportError as e:
446
osutils.failed_to_load_extension(e)
447
from breezy._annotator_py import Annotator # noqa: F401