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
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
|
# Copyright (C) 2006 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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
"""Whitebox tests for annotate functionality."""
import codecs
from cStringIO import StringIO
from bzrlib import (
annotate,
conflicts,
errors,
tests,
trace,
)
def annotation(text):
return [tuple(l.split(' ', 1)) for l in text.splitlines(True)]
parent_1 = annotation("""\
rev1 a
rev2 b
rev3 c
rev4 d
rev5 e
""")
parent_2 = annotation("""\
rev1 a
rev3 c
rev4 d
rev6 f
rev7 e
rev8 h
""")
expected_2_1 = annotation("""\
rev1 a
blahblah b
rev3 c
rev4 d
rev7 e
""")
# a: in both, same value, kept
# b: in 1, kept
# c: in both, same value, kept
# d: in both, same value, kept
# e: 1 and 2 disagree, so it goes to blahblah
# f: in 2, but not in new, so ignored
# g: not in 1 or 2, so it goes to blahblah
# h: only in parent 2, so 2 gets it
expected_1_2_2 = annotation("""\
rev1 a
rev2 b
rev3 c
rev4 d
blahblah e
blahblah g
rev8 h
""")
new_1 = """\
a
b
c
d
e
""".splitlines(True)
expected_1 = annotation("""\
blahblah a
blahblah b
blahblah c
blahblah d
blahblah e
""")
new_2 = """\
a
b
c
d
e
g
h
""".splitlines(True)
# For the 'duplicate' series, both sides introduce the same change, which then
# gets merged around. The last-modified should properly reflect this.
# We always change the fourth line so that the file is properly tracked as
# being modified in each revision. In reality, this probably would happen over
# many revisions, and it would be a different line that changes.
# BASE
# |\
# A B # line should be annotated as new for A and B
# |\|
# C D # line should 'converge' and say A
# |/
# E # D should supersede A and stay as D (not become E because C references
# A)
duplicate_base = annotation("""\
rev-base first
rev-base second
rev-base third
rev-base fourth-base
""")
duplicate_A = annotation("""\
rev-base first
rev-A alt-second
rev-base third
rev-A fourth-A
""")
duplicate_B = annotation("""\
rev-base first
rev-B alt-second
rev-base third
rev-B fourth-B
""")
duplicate_C = annotation("""\
rev-base first
rev-A alt-second
rev-base third
rev-C fourth-C
""")
duplicate_D = annotation("""\
rev-base first
rev-A alt-second
rev-base third
rev-D fourth-D
""")
duplicate_E = annotation("""\
rev-base first
rev-A alt-second
rev-base third
rev-E fourth-E
""")
class TestAnnotate(tests.TestCaseWithTransport):
def create_merged_trees(self):
"""create 2 trees with merges between them.
rev-1 --+
| |
rev-2 rev-1_1_1
| |
+------+
|
rev-3
"""
builder = self.make_branch_builder('branch')
builder.start_series()
self.addCleanup(builder.finish_series)
builder.build_snapshot('rev-1', None, [
('add', ('', 'root-id', 'directory', None)),
('add', ('a', 'a-id', 'file', 'first\n')),
], timestamp=1166046000.00, timezone=0, committer="joe@foo.com")
builder.build_snapshot('rev-2', ['rev-1'], [
('modify', ('a-id', 'first\nsecond\n')),
], timestamp=1166046001.00, timezone=0, committer="joe@foo.com")
builder.build_snapshot('rev-1_1_1', ['rev-1'], [
('modify', ('a-id', 'first\nthird\n')),
], timestamp=1166046002.00, timezone=0, committer="barry@foo.com")
builder.build_snapshot('rev-3', ['rev-2', 'rev-1_1_1'], [
('modify', ('a-id', 'first\nsecond\nthird\n')),
], timestamp=1166046003.00, timezone=0, committer="sal@foo.com")
return builder
def create_deeply_merged_trees(self):
"""Create some trees with a more complex merge history.
rev-1 --+
| |
rev-2 rev-1_1_1 --+
| | |
+------+ |
| | |
rev-3 rev-1_1_2 rev-1_2_1 ------+
| | | |
+------+ | |
| | |
rev-4 rev-1_2_2 rev-1_3_1
| | |
+-----------------+ |
| |
rev-5 |
| |
+--------------------------------+
|
rev-6
"""
builder = self.create_merged_trees()
builder.build_snapshot('rev-1_1_2', ['rev-1_1_1'], [])
builder.build_snapshot('rev-4', ['rev-3', 'rev-1_1_2'], [])
builder.build_snapshot('rev-1_2_1', ['rev-1_1_1'], [
('modify', ('a-id', 'first\nthird\nfourth\n')),
], timestamp=1166046003.00, timezone=0, committer="jerry@foo.com")
builder.build_snapshot('rev-1_2_2', ['rev-1_2_1'], [],
timestamp=1166046004.00, timezone=0, committer="jerry@foo.com")
builder.build_snapshot('rev-5', ['rev-4', 'rev-1_2_2'], [
('modify', ('a-id', 'first\nsecond\nthird\nfourth\n')),
], timestamp=1166046004.00, timezone=0, committer="jerry@foo.com")
builder.build_snapshot('rev-1_3_1', ['rev-1_2_1'], [
('modify', ('a-id', 'first\nthird\nfourth\nfifth\nsixth\n')),
], timestamp=1166046005.00, timezone=0, committer="george@foo.com")
builder.build_snapshot('rev-6', ['rev-5', 'rev-1_3_1'], [
('modify', ('a-id',
'first\nsecond\nthird\nfourth\nfifth\nsixth\n')),
])
return builder
def create_duplicate_lines_tree(self):
builder = self.make_branch_builder('branch')
builder.start_series()
self.addCleanup(builder.finish_series)
base_text = ''.join(l for r, l in duplicate_base)
a_text = ''.join(l for r, l in duplicate_A)
b_text = ''.join(l for r, l in duplicate_B)
c_text = ''.join(l for r, l in duplicate_C)
d_text = ''.join(l for r, l in duplicate_D)
e_text = ''.join(l for r, l in duplicate_E)
builder.build_snapshot('rev-base', None, [
('add', ('', 'root-id', 'directory', None)),
('add', ('file', 'file-id', 'file', base_text)),
])
builder.build_snapshot('rev-A', ['rev-base'], [
('modify', ('file-id', a_text))])
builder.build_snapshot('rev-B', ['rev-base'], [
('modify', ('file-id', b_text))])
builder.build_snapshot('rev-C', ['rev-A'], [
('modify', ('file-id', c_text))])
builder.build_snapshot('rev-D', ['rev-B', 'rev-A'], [
('modify', ('file-id', d_text))])
builder.build_snapshot('rev-E', ['rev-C', 'rev-D'], [
('modify', ('file-id', e_text))])
return builder
def assertRepoAnnotate(self, expected, repo, file_id, revision_id):
"""Assert that the revision is properly annotated."""
actual = list(repo.revision_tree(revision_id).annotate_iter(file_id))
if actual != expected:
# Create an easier to understand diff when the lines don't actually
# match
self.assertEqualDiff(''.join('\t'.join(l) for l in expected),
''.join('\t'.join(l) for l in actual))
def test_annotate_duplicate_lines(self):
# XXX: Should this be a per_repository test?
builder = self.create_duplicate_lines_tree()
repo = builder.get_branch().repository
repo.lock_read()
self.addCleanup(repo.unlock)
self.assertRepoAnnotate(duplicate_base, repo, 'file-id', 'rev-base')
self.assertRepoAnnotate(duplicate_A, repo, 'file-id', 'rev-A')
self.assertRepoAnnotate(duplicate_B, repo, 'file-id', 'rev-B')
self.assertRepoAnnotate(duplicate_C, repo, 'file-id', 'rev-C')
self.assertRepoAnnotate(duplicate_D, repo, 'file-id', 'rev-D')
self.assertRepoAnnotate(duplicate_E, repo, 'file-id', 'rev-E')
def test_annotate_shows_dotted_revnos(self):
builder = self.create_merged_trees()
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-3', 'a-id',
to_file=sio)
self.assertEqualDiff('1 joe@foo | first\n'
'2 joe@foo | second\n'
'1.1.1 barry@f | third\n',
sio.getvalue())
def test_annotate_limits_dotted_revnos(self):
"""Annotate should limit dotted revnos to a depth of 12"""
builder = self.create_deeply_merged_trees()
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-6', 'a-id',
to_file=sio, verbose=False, full=False)
self.assertEqualDiff('1 joe@foo | first\n'
'2 joe@foo | second\n'
'1.1.1 barry@f | third\n'
'1.2.1 jerry@f | fourth\n'
'1.3.1 george@ | fifth\n'
' | sixth\n',
sio.getvalue())
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-6', 'a-id',
to_file=sio, verbose=False, full=True)
self.assertEqualDiff('1 joe@foo | first\n'
'2 joe@foo | second\n'
'1.1.1 barry@f | third\n'
'1.2.1 jerry@f | fourth\n'
'1.3.1 george@ | fifth\n'
'1.3.1 george@ | sixth\n',
sio.getvalue())
# verbose=True shows everything, the full revno, user id, and date
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-6', 'a-id',
to_file=sio, verbose=True, full=False)
self.assertEqualDiff('1 joe@foo.com 20061213 | first\n'
'2 joe@foo.com 20061213 | second\n'
'1.1.1 barry@foo.com 20061213 | third\n'
'1.2.1 jerry@foo.com 20061213 | fourth\n'
'1.3.1 george@foo.com 20061213 | fifth\n'
' | sixth\n',
sio.getvalue())
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-6', 'a-id',
to_file=sio, verbose=True, full=True)
self.assertEqualDiff('1 joe@foo.com 20061213 | first\n'
'2 joe@foo.com 20061213 | second\n'
'1.1.1 barry@foo.com 20061213 | third\n'
'1.2.1 jerry@foo.com 20061213 | fourth\n'
'1.3.1 george@foo.com 20061213 | fifth\n'
'1.3.1 george@foo.com 20061213 | sixth\n',
sio.getvalue())
def test_annotate_uses_branch_context(self):
"""Dotted revnos should use the Branch context.
When annotating a non-mainline revision, the annotation should still
use dotted revnos from the mainline.
"""
builder = self.create_deeply_merged_trees()
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-1_3_1', 'a-id',
to_file=sio, verbose=False, full=False)
self.assertEqualDiff('1 joe@foo | first\n'
'1.1.1 barry@f | third\n'
'1.2.1 jerry@f | fourth\n'
'1.3.1 george@ | fifth\n'
' | sixth\n',
sio.getvalue())
def test_annotate_show_ids(self):
builder = self.create_deeply_merged_trees()
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-6', 'a-id',
to_file=sio, show_ids=True, full=False)
# It looks better with real revision ids :)
self.assertEqualDiff(' rev-1 | first\n'
' rev-2 | second\n'
'rev-1_1_1 | third\n'
'rev-1_2_1 | fourth\n'
'rev-1_3_1 | fifth\n'
' | sixth\n',
sio.getvalue())
sio = StringIO()
annotate.annotate_file(builder.get_branch(), 'rev-6', 'a-id',
to_file=sio, show_ids=True, full=True)
self.assertEqualDiff(' rev-1 | first\n'
' rev-2 | second\n'
'rev-1_1_1 | third\n'
'rev-1_2_1 | fourth\n'
'rev-1_3_1 | fifth\n'
'rev-1_3_1 | sixth\n',
sio.getvalue())
def test_annotate_unicode_author(self):
tree1 = self.make_branch_and_tree('tree1')
self.build_tree_contents([('tree1/a', 'adi\xc3\xb3s')])
tree1.add(['a'], ['a-id'])
tree1.commit('a', rev_id='rev-1',
committer=u'Pepe P\xe9rez <pperez@ejemplo.com>',
timestamp=1166046000.00, timezone=0)
self.build_tree_contents([('tree1/b', 'bye')])
tree1.add(['b'], ['b-id'])
tree1.commit('b', rev_id='rev-2',
committer=u'p\xe9rez',
timestamp=1166046000.00, timezone=0)
tree1.lock_read()
self.addCleanup(tree1.unlock)
# this passes if no exception is raised
to_file = StringIO()
annotate.annotate_file(tree1.branch, 'rev-1', 'a-id', to_file=to_file)
sio = StringIO()
to_file = codecs.getwriter('ascii')(sio)
to_file.encoding = 'ascii' # codecs does not set it
annotate.annotate_file(tree1.branch, 'rev-2', 'b-id', to_file=to_file)
self.assertEqualDiff('2 p?rez | bye\n', sio.getvalue())
# test now with to_file.encoding = None
to_file = tests.StringIOWrapper()
to_file.encoding = None
annotate.annotate_file(tree1.branch, 'rev-2', 'b-id', to_file=to_file)
self.assertContainsRe('2 p.rez | bye\n', to_file.getvalue())
# and when it does not exist
to_file = StringIO()
annotate.annotate_file(tree1.branch, 'rev-2', 'b-id', to_file=to_file)
self.assertContainsRe('2 p.rez | bye\n', to_file.getvalue())
def test_annotate_author_or_committer(self):
tree1 = self.make_branch_and_tree('tree1')
self.build_tree_contents([('tree1/a', 'hello')])
tree1.add(['a'], ['a-id'])
tree1.commit('a', rev_id='rev-1',
committer='Committer <committer@example.com>',
timestamp=1166046000.00, timezone=0)
self.build_tree_contents([('tree1/b', 'bye')])
tree1.add(['b'], ['b-id'])
tree1.commit('b', rev_id='rev-2',
committer='Committer <committer@example.com>',
authors=['Author <author@example.com>'],
timestamp=1166046000.00, timezone=0)
tree1.lock_read()
self.addCleanup(tree1.unlock)
to_file = StringIO()
annotate.annotate_file(tree1.branch, 'rev-1', 'a-id', to_file=to_file)
self.assertEqual('1 committ | hello\n', to_file.getvalue())
to_file = StringIO()
annotate.annotate_file(tree1.branch, 'rev-2', 'b-id', to_file=to_file)
self.assertEqual('2 author@ | bye\n', to_file.getvalue())
class TestReannotate(tests.TestCase):
def annotateEqual(self, expected, parents, newlines, revision_id,
blocks=None):
annotate_list = list(annotate.reannotate(parents, newlines,
revision_id, blocks))
self.assertEqual(len(expected), len(annotate_list))
for e, a in zip(expected, annotate_list):
self.assertEqual(e, a)
def test_reannotate(self):
self.annotateEqual(parent_1, [parent_1], new_1, 'blahblah')
self.annotateEqual(expected_2_1, [parent_2], new_1, 'blahblah')
self.annotateEqual(expected_1_2_2, [parent_1, parent_2], new_2,
'blahblah')
def test_reannotate_no_parents(self):
self.annotateEqual(expected_1, [], new_1, 'blahblah')
def test_reannotate_left_matching_blocks(self):
"""Ensure that left_matching_blocks has an impact.
In this case, the annotation is ambiguous, so the hint isn't actually
lying.
"""
parent = [('rev1', 'a\n')]
new_text = ['a\n', 'a\n']
blocks = [(0, 0, 1), (1, 2, 0)]
self.annotateEqual([('rev1', 'a\n'), ('rev2', 'a\n')], [parent],
new_text, 'rev2', blocks)
blocks = [(0, 1, 1), (1, 2, 0)]
self.annotateEqual([('rev2', 'a\n'), ('rev1', 'a\n')], [parent],
new_text, 'rev2', blocks)
|