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, file_id=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 file_id: The file_id to annotate (must match file path)
69
:param branch: Branch to use for revision revno lookups
76
encoding = osutils.get_terminal_encoding()
77
# Handle the show_ids case
78
annotations = list(tree.annotate_iter(path, file_id))
80
return _show_id_annotations(annotations, to_file, full, encoding)
82
if not getattr(tree, "get_revision_id", False):
83
# Create a virtual revision to represent the current tree state.
84
# Should get some more pending commit attributes, like pending tags,
86
current_rev = Revision(CURRENT_REVISION)
87
current_rev.parent_ids = tree.get_parent_ids()
89
current_rev.committer = branch.get_config_stack().get('email')
91
current_rev.committer = 'local user'
92
current_rev.message = "?"
93
current_rev.timestamp = round(time.time(), 3)
94
current_rev.timezone = osutils.local_time_offset()
97
annotation = list(_expand_annotations(
98
annotations, branch, current_rev))
99
_print_annotations(annotation, verbose, to_file, full, encoding)
102
def _print_annotations(annotation, verbose, to_file, full, encoding):
103
"""Print annotations to to_file.
105
:param to_file: The file to output the annotation to.
106
:param verbose: Show all details rather than truncating to ensure
107
reasonable text width.
108
:param full: XXXX Not sure what this does.
110
if len(annotation) == 0:
111
max_origin_len = max_revno_len = 0
113
max_origin_len = max(len(x[1]) for x in annotation)
114
max_revno_len = max(len(x[0]) 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.decode(encoding),))
138
def _show_id_annotations(annotations, to_file, full, encoding):
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' % (
149
max_origin_len, this.decode('utf-8'), text.decode(encoding)))
154
def _expand_annotations(annotations, branch, current_rev=None):
155
"""Expand a file's annotations into command line UI ready tuples.
157
Each tuple includes detailed information, such as the author name, and date
158
string for the commit, rather than just the revision id.
160
:param annotations: The annotations to expand.
161
:param revision_id_to_revno: A map from id to revision numbers.
162
:param branch: A locked branch to query for revision details.
164
repository = branch.repository
165
revision_ids = set(o for o, t in annotations)
166
if current_rev is not None:
167
# This can probably become a function on MutableTree, get_revno_map
168
# there, or something.
169
last_revision = current_rev.revision_id
170
# XXX: Partially Cloned from branch, uses the old_get_graph, eep.
171
# XXX: The main difficulty is that we need to inject a single new node
172
# (current_rev) into the graph before it gets numbered, etc.
173
# Once KnownGraph gets an 'add_node()' function, we can use
174
# VF.get_known_graph_ancestry().
175
graph = repository.get_graph()
177
key: value for key, value in
178
graph.iter_ancestry(current_rev.parent_ids) if value is not None}
179
revision_graph = _strip_NULL_ghosts(revision_graph)
180
revision_graph[last_revision] = current_rev.parent_ids
181
merge_sorted_revisions = tsort.merge_sort(
186
revision_id_to_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(
215
str(i) for i in 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(
287
None, 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
306
def _old_break_annotation_tie(annotated_lines):
307
"""Chose an attribution between several possible ones.
309
:param annotated_lines: A list of tuples ((file_id, rev_id), line) where
310
the lines are identical but the revids different while no parent
311
relation exist between them
313
:return : The "winning" line. This must be one with a revid that
314
guarantees that further criss-cross merges will converge. Failing to
315
do so have performance implications.
317
# sort lexicographically so that we always get a stable result.
319
# TODO: while 'sort' is the easiest (and nearly the only possible solution)
320
# with the current implementation, chosing the oldest revision is known to
321
# provide better results (as in matching user expectations). The most
322
# common use case being manual cherry-pick from an already existing
324
return sorted(annotated_lines)[0]
327
def _find_matching_unannotated_lines(output_lines, plain_child_lines,
328
child_lines, start_child, end_child,
329
right_lines, start_right, end_right,
330
heads_provider, revision_id):
331
"""Find lines in plain_right_lines that match the existing lines.
333
:param output_lines: Append final annotated lines to this list
334
:param plain_child_lines: The unannotated new lines for the child text
335
:param child_lines: Lines for the child text which have been annotated
338
:param start_child: Position in plain_child_lines and child_lines to start
340
:param end_child: Last position in plain_child_lines and child_lines to
342
:param right_lines: The annotated lines for the whole text for the right
344
:param start_right: Position in right_lines to start the match
345
:param end_right: Last position in right_lines to search for a match
346
:param heads_provider: When parents disagree on the lineage of a line, we
347
need to check if one side supersedes the other
348
:param revision_id: The label to give if a line should be labeled 'tip'
350
output_extend = output_lines.extend
351
output_append = output_lines.append
352
# We need to see if any of the unannotated lines match
353
plain_right_subset = [l for a, l in right_lines[start_right:end_right]]
354
plain_child_subset = plain_child_lines[start_child:end_child]
355
match_blocks = _get_matching_blocks(plain_right_subset, plain_child_subset)
359
for right_idx, child_idx, match_len in match_blocks:
360
# All the lines that don't match are just passed along
361
if child_idx > last_child_idx:
362
output_extend(child_lines[start_child + last_child_idx:
363
start_child + child_idx])
364
for offset in range(match_len):
365
left = child_lines[start_child + child_idx + offset]
366
right = right_lines[start_right + right_idx + offset]
367
if left[0] == right[0]:
368
# The annotations match, just return the left one
370
elif left[0] == revision_id:
371
# The left parent marked this as unmatched, so let the
372
# right parent claim it
375
# Left and Right both claim this line
376
if heads_provider is None:
377
output_append((revision_id, left[1]))
379
heads = heads_provider.heads((left[0], right[0]))
381
output_append((next(iter(heads)), left[1]))
383
# Both claim different origins, get a stable result.
384
# If the result is not stable, there is a risk a
385
# performance degradation as criss-cross merges will
386
# flip-flop the attribution.
387
if _break_annotation_tie is None:
389
_old_break_annotation_tie([left, right]))
391
output_append(_break_annotation_tie([left, right]))
392
last_child_idx = child_idx + match_len
395
def _reannotate_annotated(right_parent_lines, new_lines, new_revision_id,
396
annotated_lines, heads_provider):
397
"""Update the annotations for a node based on another parent.
399
:param right_parent_lines: A list of annotated lines for the right-hand
401
:param new_lines: The unannotated new lines.
402
:param new_revision_id: The revision_id to attribute to lines which are not
403
present in either parent.
404
:param annotated_lines: A list of annotated lines. This should be the
405
annotation of new_lines based on parents seen so far.
406
:param heads_provider: When parents disagree on the lineage of a line, we
407
need to check if one side supersedes the other.
409
if len(new_lines) != len(annotated_lines):
410
raise AssertionError("mismatched new_lines and annotated_lines")
411
# First compare the newly annotated lines with the right annotated lines.
412
# Lines which were not changed in left or right should match. This tends to
413
# be the bulk of the lines, and they will need no further processing.
415
lines_extend = lines.extend
416
# The line just after the last match from the right side
419
matching_left_and_right = _get_matching_blocks(right_parent_lines,
421
for right_idx, left_idx, match_len in matching_left_and_right:
422
# annotated lines from last_left_idx to left_idx did not match the
423
# lines from last_right_idx to right_idx, the raw lines should be
424
# compared to determine what annotations need to be updated
425
if last_right_idx == right_idx or last_left_idx == left_idx:
426
# One of the sides is empty, so this is a pure insertion
427
lines_extend(annotated_lines[last_left_idx:left_idx])
429
# We need to see if any of the unannotated lines match
430
_find_matching_unannotated_lines(lines,
431
new_lines, annotated_lines,
432
last_left_idx, left_idx,
434
last_right_idx, right_idx,
437
last_right_idx = right_idx + match_len
438
last_left_idx = left_idx + match_len
439
# If left and right agree on a range, just push that into the output
440
lines_extend(annotated_lines[left_idx:left_idx + match_len])
445
from breezy._annotator_pyx import Annotator
446
except ImportError as e:
447
osutils.failed_to_load_extension(e)
448
from breezy._annotator_py import Annotator # noqa: F401