1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
|
# Copyright (C) 2004, 2005 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
"""File annotate based on weave storage"""
# TODO: Choice of more or less verbose formats:
#
# interposed: show more details between blocks of modified lines
# TODO: Show which revision caused a line to merge into the parent
# TODO: perhaps abbreviate timescales depending on how recent they are
# e.g. "3:12 Tue", "13 Oct", "Oct 2005", etc.
import sys
import time
from bzrlib import (
errors,
patiencediff,
tsort,
)
from bzrlib.config import extract_email_address
def annotate_file(branch, rev_id, file_id, verbose=False, full=False,
to_file=None, show_ids=False):
if to_file is None:
to_file = sys.stdout
prevanno=''
last_rev_id = None
if show_ids:
w = branch.repository.weave_store.get_weave(file_id,
branch.repository.get_transaction())
annotations = list(w.annotate_iter(rev_id))
max_origin_len = max(len(origin) for origin, text in annotations)
for origin, text in annotations:
if full or last_rev_id != origin:
this = origin
else:
this = ''
to_file.write('%*s | %s' % (max_origin_len, this, text))
last_rev_id = origin
return
annotation = list(_annotate_file(branch, rev_id, file_id))
if len(annotation) == 0:
max_origin_len = max_revno_len = max_revid_len = 0
else:
max_origin_len = max(len(x[1]) for x in annotation)
max_revno_len = max(len(x[0]) for x in annotation)
max_revid_len = max(len(x[3]) for x in annotation)
if not verbose:
max_revno_len = min(max_revno_len, 12)
max_revno_len = max(max_revno_len, 3)
for (revno_str, author, date_str, line_rev_id, text) in annotation:
if verbose:
anno = '%-*s %-*s %8s ' % (max_revno_len, revno_str,
max_origin_len, author, date_str)
else:
if len(revno_str) > max_revno_len:
revno_str = revno_str[:max_revno_len-1] + '>'
anno = "%-*s %-7s " % (max_revno_len, revno_str, author[:7])
if anno.lstrip() == "" and full: anno = prevanno
print >>to_file, '%s| %s' % (anno, text)
prevanno=anno
def _annotate_file(branch, rev_id, file_id ):
"""Yield the origins for each line of a file.
This includes detailed information, such as the committer name, and
date string for the commit, rather than just the revision id.
"""
rh = branch.revision_history()
revision_graph = branch.repository.get_revision_graph(rev_id)
merge_sorted_revisions = tsort.merge_sort(
revision_graph,
rev_id,
None,
generate_revno=True)
revision_id_to_revno = dict((rev_id, revno)
for seq_num, rev_id, depth, revno, end_of_merge
in merge_sorted_revisions)
w = branch.repository.weave_store.get_weave(file_id,
branch.repository.get_transaction())
last_origin = None
annotations = list(w.annotate_iter(rev_id))
revision_ids = set(o for o, t in annotations)
revision_ids = [o for o in revision_ids if
branch.repository.has_revision(o)]
revisions = dict((r.revision_id, r) for r in
branch.repository.get_revisions(revision_ids))
for origin, text in annotations:
text = text.rstrip('\r\n')
if origin == last_origin:
(revno_str, author, date_str) = ('','','')
else:
last_origin = origin
if origin not in revisions:
(revno_str, author, date_str) = ('?','?','?')
else:
revno_str = '.'.join(str(i) for i in
revision_id_to_revno[origin])
rev = revisions[origin]
tz = rev.timezone or 0
date_str = time.strftime('%Y%m%d',
time.gmtime(rev.timestamp + tz))
# a lazy way to get something like the email address
# TODO: Get real email address
author = rev.committer
try:
author = extract_email_address(author)
except errors.NoEmailInUsername:
pass # use the whole name
yield (revno_str, author, date_str, origin, text)
def reannotate(parents_lines, new_lines, new_revision_id):
"""Create a new annotated version from new lines and parent annotations.
:param parents_lines: List of annotated lines for all parents
:param new_lines: The un-annotated new lines
:param new_revision_id: The revision-id to associate with new lines
(will often be CURRENT_REVISION)
"""
if len(parents_lines) == 1:
for data in _reannotate(parents_lines[0], new_lines, new_revision_id):
yield data
else:
reannotations = [list(_reannotate(p, new_lines, new_revision_id)) for
p in parents_lines]
for annos in zip(*reannotations):
origins = set(a for a, l in annos)
line = annos[0][1]
if len(origins) == 1:
yield iter(origins).next(), line
elif len(origins) == 2 and new_revision_id in origins:
yield (x for x in origins if x != new_revision_id).next(), line
else:
yield new_revision_id, line
def _reannotate(parent_lines, new_lines, new_revision_id):
plain_parent_lines = [l for r, l in parent_lines]
matcher = patiencediff.PatienceSequenceMatcher(None, plain_parent_lines,
new_lines)
new_cur = 0
for i, j, n in matcher.get_matching_blocks():
for line in new_lines[new_cur:j]:
yield new_revision_id, line
for data in parent_lines[i:i+n]:
yield data
new_cur = j + n
|