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(), """
44
from .config import extract_email_address
45
from .repository import _strip_NULL_ghosts
46
from .revision import (
52
def annotate_file_tree(tree, file_id, to_file, verbose=False, full=False,
53
show_ids=False, branch=None):
54
"""Annotate file_id in a tree.
56
The tree should already be read_locked() when annotate_file_tree is called.
58
:param tree: The tree to look for revision numbers and history from.
59
:param file_id: The file_id to annotate.
60
:param to_file: The file to output the annotation to.
61
:param verbose: Show all details rather than truncating to ensure
62
reasonable text width.
63
:param full: XXXX Not sure what this does.
64
:param show_ids: Show revision ids in the annotation output.
65
:param branch: Branch to use for revision revno lookups
72
# Handle the show_ids case
73
annotations = list(tree.annotate_iter(file_id))
75
return _show_id_annotations(annotations, to_file, full)
77
if not getattr(tree, "get_revision_id", False):
78
# Create a virtual revision to represent the current tree state.
79
# Should get some more pending commit attributes, like pending tags,
81
current_rev = Revision(CURRENT_REVISION)
82
current_rev.parent_ids = tree.get_parent_ids()
84
current_rev.committer = branch.get_config_stack().get('email')
85
except errors.NoWhoami:
86
current_rev.committer = 'local user'
87
current_rev.message = "?"
88
current_rev.timestamp = round(time.time(), 3)
89
current_rev.timezone = osutils.local_time_offset()
92
annotation = list(_expand_annotations(annotations, branch,
94
_print_annotations(annotation, verbose, to_file, full)
97
def _print_annotations(annotation, verbose, to_file, full):
98
"""Print annotations to to_file.
100
:param to_file: The file to output the annotation to.
101
:param verbose: Show all details rather than truncating to ensure
102
reasonable text width.
103
:param full: XXXX Not sure what this does.
105
if len(annotation) == 0:
106
max_origin_len = max_revno_len = max_revid_len = 0
108
max_origin_len = max(len(x[1]) for x in annotation)
109
max_revno_len = max(len(x[0]) for x in annotation)
110
max_revid_len = max(len(x[3]) for x in annotation)
112
max_revno_len = min(max_revno_len, 12)
113
max_revno_len = max(max_revno_len, 3)
115
# Output the annotations
117
for (revno_str, author, date_str, line_rev_id, text) in annotation:
119
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
120
max_origin_len, author, date_str)
122
if len(revno_str) > max_revno_len:
123
revno_str = revno_str[:max_revno_len-1] + '>'
124
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
125
if anno.lstrip() == "" and full:
127
# GZ 2017-05-21: Writing both unicode annotation and bytes from file
128
# which the given to_file must cope with.
130
to_file.write('| %s\n' % (text,))
134
def _show_id_annotations(annotations, to_file, full):
138
max_origin_len = max(len(origin) for origin, text in annotations)
139
for origin, text in annotations:
140
if full or last_rev_id != origin:
144
to_file.write('%*s | %s' % (max_origin_len, this, text))
149
def _expand_annotations(annotations, branch, current_rev=None):
150
"""Expand a file's annotations into command line UI ready tuples.
152
Each tuple includes detailed information, such as the author name, and date
153
string for the commit, rather than just the revision id.
155
:param annotations: The annotations to expand.
156
:param revision_id_to_revno: A map from id to revision numbers.
157
:param branch: A locked branch to query for revision details.
159
repository = branch.repository
160
if current_rev is not None:
161
# This can probably become a function on MutableTree, get_revno_map
162
# there, or something.
163
last_revision = current_rev.revision_id
164
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
165
# XXX: The main difficulty is that we need to inject a single new node
166
# (current_rev) into the graph before it gets numbered, etc.
167
# Once KnownGraph gets an 'add_node()' function, we can use
168
# VF.get_known_graph_ancestry().
169
graph = repository.get_graph()
170
revision_graph = dict(((key, value) for key, value in
171
graph.iter_ancestry(current_rev.parent_ids) if value is not None))
172
revision_graph = _strip_NULL_ghosts(revision_graph)
173
revision_graph[last_revision] = current_rev.parent_ids
174
merge_sorted_revisions = tsort.merge_sort(
179
revision_id_to_revno = dict((rev_id, revno)
180
for seq_num, rev_id, depth, revno, end_of_merge in
181
merge_sorted_revisions)
183
revision_id_to_revno = branch.get_revision_id_to_revno_map()
185
revision_ids = set(o for o, t in annotations)
187
if CURRENT_REVISION in revision_ids:
188
revision_id_to_revno[CURRENT_REVISION] = (
189
"%d?" % (branch.revno() + 1),)
190
revisions[CURRENT_REVISION] = current_rev
191
revision_ids = [o for o in revision_ids if
192
repository.has_revision(o)]
193
revisions.update((r.revision_id, r) for r in
194
repository.get_revisions(revision_ids))
195
for origin, text in annotations:
196
text = text.rstrip('\r\n')
197
if origin == last_origin:
198
(revno_str, author, date_str) = ('','','')
201
if origin not in revisions:
202
(revno_str, author, date_str) = ('?','?','?')
204
revno_str = '.'.join(str(i) for i in
205
revision_id_to_revno[origin])
206
rev = revisions[origin]
207
tz = rev.timezone or 0
208
date_str = time.strftime('%Y%m%d',
209
osutils.gmtime(rev.timestamp + tz))
210
# a lazy way to get something like the email address
211
# TODO: Get real email address
212
author = rev.get_apparent_authors()[0]
214
author = extract_email_address(author)
215
except errors.NoEmailInUsername:
216
pass # use the whole name
217
yield (revno_str, author, date_str, origin, text)
220
def reannotate(parents_lines, new_lines, new_revision_id,
221
_left_matching_blocks=None,
222
heads_provider=None):
223
"""Create a new annotated version from new lines and parent annotations.
225
:param parents_lines: List of annotated lines for all parents
226
:param new_lines: The un-annotated new lines
227
:param new_revision_id: The revision-id to associate with new lines
228
(will often be CURRENT_REVISION)
229
:param left_matching_blocks: a hint about which areas are common
230
between the text and its left-hand-parent. The format is
231
the SequenceMatcher.get_matching_blocks format
232
(start_left, start_right, length_of_match).
233
:param heads_provider: An object which provides a .heads() call to resolve
234
if any revision ids are children of others.
235
If None, then any ancestry disputes will be resolved with
238
if len(parents_lines) == 0:
239
lines = [(new_revision_id, line) for line in new_lines]
240
elif len(parents_lines) == 1:
241
lines = _reannotate(parents_lines[0], new_lines, new_revision_id,
242
_left_matching_blocks)
243
elif len(parents_lines) == 2:
244
left = _reannotate(parents_lines[0], new_lines, new_revision_id,
245
_left_matching_blocks)
246
lines = _reannotate_annotated(parents_lines[1], new_lines,
247
new_revision_id, left,
250
reannotations = [_reannotate(parents_lines[0], new_lines,
251
new_revision_id, _left_matching_blocks)]
252
reannotations.extend(_reannotate(p, new_lines, new_revision_id)
253
for p in parents_lines[1:])
255
for annos in zip(*reannotations):
256
origins = set(a for a, l in annos)
257
if len(origins) == 1:
258
# All the parents agree, so just return the first one
259
lines.append(annos[0])
262
if len(origins) == 2 and new_revision_id in origins:
263
origins.remove(new_revision_id)
264
if len(origins) == 1:
265
lines.append((origins.pop(), line))
267
lines.append((new_revision_id, line))
271
def _reannotate(parent_lines, new_lines, new_revision_id,
272
matching_blocks=None):
274
if matching_blocks is None:
275
plain_parent_lines = [l for r, l in parent_lines]
276
matcher = patiencediff.PatienceSequenceMatcher(None,
277
plain_parent_lines, new_lines)
278
matching_blocks = matcher.get_matching_blocks()
280
for i, j, n in matching_blocks:
281
for line in new_lines[new_cur:j]:
282
lines.append((new_revision_id, line))
283
lines.extend(parent_lines[i:i+n])
288
def _get_matching_blocks(old, new):
289
matcher = patiencediff.PatienceSequenceMatcher(None, old, new)
290
return matcher.get_matching_blocks()
293
_break_annotation_tie = None
295
def _old_break_annotation_tie(annotated_lines):
296
"""Chose an attribution between several possible ones.
298
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
299
the lines are identical but the revids different while no parent
300
relation exist between them
302
:return : The "winning" line. This must be one with a revid that
303
guarantees that further criss-cross merges will converge. Failing to
304
do so have performance implications.
306
# sort lexicographically so that we always get a stable result.
308
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
309
# with the current implementation, chosing the oldest revision is known to
310
# provide better results (as in matching user expectations). The most
311
# common use case being manual cherry-pick from an already existing
313
return sorted(annotated_lines)[0]
316
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
317
child_lines, start_child, end_child,
318
right_lines, start_right, end_right,
319
heads_provider, revision_id):
320
"""Find lines in plain_right_lines that match the existing lines.
322
:param output_lines: Append final annotated lines to this list
323
:param plain_child_lines: The unannotated new lines for the child text
324
:param child_lines: Lines for the child text which have been annotated
327
:param start_child: Position in plain_child_lines and child_lines to start
329
:param end_child: Last position in plain_child_lines and child_lines to
331
:param right_lines: The annotated lines for the whole text for the right
333
:param start_right: Position in right_lines to start the match
334
:param end_right: Last position in right_lines to search for a match
335
:param heads_provider: When parents disagree on the lineage of a line, we
336
need to check if one side supersedes the other
337
:param revision_id: The label to give if a line should be labeled 'tip'
339
output_extend = output_lines.extend
340
output_append = output_lines.append
341
# We need to see if any of the unannotated lines match
342
plain_right_subset = [l for a,l in right_lines[start_right:end_right]]
343
plain_child_subset = plain_child_lines[start_child:end_child]
344
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
348
for right_idx, child_idx, match_len in match_blocks:
349
# All the lines that don't match are just passed along
350
if child_idx > last_child_idx:
351
output_extend(child_lines[start_child + last_child_idx
352
:start_child + child_idx])
353
for offset in range(match_len):
354
left = child_lines[start_child+child_idx+offset]
355
right = right_lines[start_right+right_idx+offset]
356
if left[0] == right[0]:
357
# The annotations match, just return the left one
359
elif left[0] == revision_id:
360
# The left parent marked this as unmatched, so let the
361
# right parent claim it
364
# Left and Right both claim this line
365
if heads_provider is None:
366
output_append((revision_id, left[1]))
368
heads = heads_provider.heads((left[0], right[0]))
370
output_append((next(iter(heads)), left[1]))
372
# Both claim different origins, get a stable result.
373
# If the result is not stable, there is a risk a
374
# performance degradation as criss-cross merges will
375
# flip-flop the attribution.
376
if _break_annotation_tie is None:
378
_old_break_annotation_tie([left, right]))
380
output_append(_break_annotation_tie([left, right]))
381
last_child_idx = child_idx + match_len
384
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
385
annotated_lines, heads_provider):
386
"""Update the annotations for a node based on another parent.
388
:param right_parent_lines: A list of annotated lines for the right-hand
390
:param new_lines: The unannotated new lines.
391
:param new_revision_id: The revision_id to attribute to lines which are not
392
present in either parent.
393
:param annotated_lines: A list of annotated lines. This should be the
394
annotation of new_lines based on parents seen so far.
395
:param heads_provider: When parents disagree on the lineage of a line, we
396
need to check if one side supersedes the other.
398
if len(new_lines) != len(annotated_lines):
399
raise AssertionError("mismatched new_lines and annotated_lines")
400
# First compare the newly annotated lines with the right annotated lines.
401
# Lines which were not changed in left or right should match. This tends to
402
# be the bulk of the lines, and they will need no further processing.
404
lines_extend = lines.extend
405
last_right_idx = 0 # The line just after the last match from the right side
407
matching_left_and_right = _get_matching_blocks(right_parent_lines,
409
for right_idx, left_idx, match_len in matching_left_and_right:
410
# annotated lines from last_left_idx to left_idx did not match the
411
# lines from last_right_idx to right_idx, the raw lines should be
412
# compared to determine what annotations need to be updated
413
if last_right_idx == right_idx or last_left_idx == left_idx:
414
# One of the sides is empty, so this is a pure insertion
415
lines_extend(annotated_lines[last_left_idx:left_idx])
417
# We need to see if any of the unannotated lines match
418
_find_matching_unannotated_lines(lines,
419
new_lines, annotated_lines,
420
last_left_idx, left_idx,
422
last_right_idx, right_idx,
425
last_right_idx = right_idx + match_len
426
last_left_idx = left_idx + match_len
427
# If left and right agree on a range, just push that into the output
428
lines_extend(annotated_lines[left_idx:left_idx + match_len])
433
from breezy._annotator_pyx import Annotator
434
except ImportError as e:
435
osutils.failed_to_load_extension(e)
436
from breezy._annotator_py import Annotator