92
47
self.assert_('@@' in lines[2][2:])
93
48
## "Unterminated hunk header for patch:\n%s" % "".join(lines)
95
def test_binary_lines(self):
96
self.assertRaises(BinaryFile, udiff_lines, [1023 * 'a' + '\x00'], [])
97
self.assertRaises(BinaryFile, udiff_lines, [], [1023 * 'a' + '\x00'])
98
udiff_lines([1023 * 'a' + '\x00'], [], allow_binary=True)
99
udiff_lines([], [1023 * 'a' + '\x00'], allow_binary=True)
101
def test_external_diff(self):
102
lines = external_udiff_lines(['boo\n'], ['goo\n'])
103
self.check_patch(lines)
104
self.assertEqual('\n', lines[-1])
106
def test_external_diff_no_fileno(self):
107
# Make sure that we can handle not having a fileno, even
108
# if the diff is large
109
lines = external_udiff_lines(['boo\n']*10000,
112
self.check_patch(lines)
114
def test_external_diff_binary_lang_c(self):
116
for lang in ('LANG', 'LC_ALL', 'LANGUAGE'):
117
old_env[lang] = osutils.set_or_unset_env(lang, 'C')
119
lines = external_udiff_lines(['\x00foobar\n'], ['foo\x00bar\n'])
120
# Older versions of diffutils say "Binary files", newer
121
# versions just say "Files".
122
self.assertContainsRe(lines[0],
123
'(Binary f|F)iles old and new differ\n')
124
self.assertEquals(lines[1:], ['\n'])
126
for lang, old_val in old_env.iteritems():
127
osutils.set_or_unset_env(lang, old_val)
129
def test_no_external_diff(self):
130
"""Check that NoDiff is raised when diff is not available"""
131
# Use os.environ['PATH'] to make sure no 'diff' command is available
132
orig_path = os.environ['PATH']
134
os.environ['PATH'] = ''
135
self.assertRaises(NoDiff, external_diff,
136
'old', ['boo\n'], 'new', ['goo\n'],
137
StringIO(), diff_opts=['-u'])
139
os.environ['PATH'] = orig_path
141
def test_internal_diff_default(self):
142
# Default internal diff encoding is utf8
144
internal_diff(u'old_\xb5', ['old_text\n'],
145
u'new_\xe5', ['new_text\n'], output)
146
lines = output.getvalue().splitlines(True)
147
self.check_patch(lines)
148
self.assertEquals(['--- old_\xc2\xb5\n',
149
'+++ new_\xc3\xa5\n',
157
def test_internal_diff_utf8(self):
159
internal_diff(u'old_\xb5', ['old_text\n'],
160
u'new_\xe5', ['new_text\n'], output,
161
path_encoding='utf8')
162
lines = output.getvalue().splitlines(True)
163
self.check_patch(lines)
164
self.assertEquals(['--- old_\xc2\xb5\n',
165
'+++ new_\xc3\xa5\n',
173
def test_internal_diff_iso_8859_1(self):
175
internal_diff(u'old_\xb5', ['old_text\n'],
176
u'new_\xe5', ['new_text\n'], output,
177
path_encoding='iso-8859-1')
178
lines = output.getvalue().splitlines(True)
179
self.check_patch(lines)
180
self.assertEquals(['--- old_\xb5\n',
189
def test_internal_diff_returns_bytes(self):
191
output = StringIO.StringIO()
192
internal_diff(u'old_\xb5', ['old_text\n'],
193
u'new_\xe5', ['new_text\n'], output)
194
self.failUnless(isinstance(output.getvalue(), str),
195
'internal_diff should return bytestrings')
198
class TestDiffFiles(TestCaseInTempDir):
200
def test_external_diff_binary(self):
201
"""The output when using external diff should use diff's i18n error"""
202
# Make sure external_diff doesn't fail in the current LANG
203
lines = external_udiff_lines(['\x00foobar\n'], ['foo\x00bar\n'])
205
cmd = ['diff', '-u', '--binary', 'old', 'new']
206
open('old', 'wb').write('\x00foobar\n')
207
open('new', 'wb').write('foo\x00bar\n')
208
pipe = subprocess.Popen(cmd, stdout=subprocess.PIPE,
209
stdin=subprocess.PIPE)
210
out, err = pipe.communicate()
211
# Diff returns '2' on Binary files.
212
self.assertEqual(2, pipe.returncode)
213
# We should output whatever diff tells us, plus a trailing newline
214
self.assertEqual(out.splitlines(True) + ['\n'], lines)
217
class TestShowDiffTreesHelper(TestCaseWithTransport):
218
"""Has a helper for running show_diff_trees"""
220
def get_diff(self, tree1, tree2, specific_files=None, working_tree=None):
222
if working_tree is not None:
223
extra_trees = (working_tree,)
226
show_diff_trees(tree1, tree2, output, specific_files=specific_files,
227
extra_trees=extra_trees, old_label='old/',
229
return output.getvalue()
232
class TestDiffDates(TestShowDiffTreesHelper):
235
super(TestDiffDates, self).setUp()
236
self.wt = self.make_branch_and_tree('.')
237
self.b = self.wt.branch
238
self.build_tree_contents([
239
('file1', 'file1 contents at rev 1\n'),
240
('file2', 'file2 contents at rev 1\n')
242
self.wt.add(['file1', 'file2'])
244
message='Revision 1',
245
timestamp=1143849600, # 2006-04-01 00:00:00 UTC
248
self.build_tree_contents([('file1', 'file1 contents at rev 2\n')])
250
message='Revision 2',
251
timestamp=1143936000, # 2006-04-02 00:00:00 UTC
254
self.build_tree_contents([('file2', 'file2 contents at rev 3\n')])
256
message='Revision 3',
257
timestamp=1144022400, # 2006-04-03 00:00:00 UTC
260
self.wt.remove(['file2'])
262
message='Revision 4',
263
timestamp=1144108800, # 2006-04-04 00:00:00 UTC
266
self.build_tree_contents([
267
('file1', 'file1 contents in working tree\n')
269
# set the date stamps for files in the working tree to known values
270
os.utime('file1', (1144195200, 1144195200)) # 2006-04-05 00:00:00 UTC
272
def test_diff_rev_tree_working_tree(self):
273
output = self.get_diff(self.wt.basis_tree(), self.wt)
274
# note that the date for old/file1 is from rev 2 rather than from
275
# the basis revision (rev 4)
276
self.assertEqualDiff(output, '''\
277
=== modified file 'file1'
278
--- old/file1\t2006-04-02 00:00:00 +0000
279
+++ new/file1\t2006-04-05 00:00:00 +0000
281
-file1 contents at rev 2
282
+file1 contents in working tree
286
def test_diff_rev_tree_rev_tree(self):
287
tree1 = self.b.repository.revision_tree('rev-2')
288
tree2 = self.b.repository.revision_tree('rev-3')
289
output = self.get_diff(tree1, tree2)
290
self.assertEqualDiff(output, '''\
291
=== modified file 'file2'
292
--- old/file2\t2006-04-01 00:00:00 +0000
293
+++ new/file2\t2006-04-03 00:00:00 +0000
295
-file2 contents at rev 1
296
+file2 contents at rev 3
300
def test_diff_add_files(self):
301
tree1 = self.b.repository.revision_tree(None)
302
tree2 = self.b.repository.revision_tree('rev-1')
303
output = self.get_diff(tree1, tree2)
304
# the files have the epoch time stamp for the tree in which
306
self.assertEqualDiff(output, '''\
307
=== added file 'file1'
308
--- old/file1\t1970-01-01 00:00:00 +0000
309
+++ new/file1\t2006-04-01 00:00:00 +0000
311
+file1 contents at rev 1
313
=== added file 'file2'
314
--- old/file2\t1970-01-01 00:00:00 +0000
315
+++ new/file2\t2006-04-01 00:00:00 +0000
317
+file2 contents at rev 1
321
def test_diff_remove_files(self):
322
tree1 = self.b.repository.revision_tree('rev-3')
323
tree2 = self.b.repository.revision_tree('rev-4')
324
output = self.get_diff(tree1, tree2)
325
# the file has the epoch time stamp for the tree in which
327
self.assertEqualDiff(output, '''\
328
=== removed file 'file2'
329
--- old/file2\t2006-04-03 00:00:00 +0000
330
+++ new/file2\t1970-01-01 00:00:00 +0000
332
-file2 contents at rev 3
336
def test_show_diff_specified(self):
337
"""A working tree filename can be used to identify a file"""
338
self.wt.rename_one('file1', 'file1b')
339
old_tree = self.b.repository.revision_tree('rev-1')
340
new_tree = self.b.repository.revision_tree('rev-4')
341
out = self.get_diff(old_tree, new_tree, specific_files=['file1b'],
342
working_tree=self.wt)
343
self.assertContainsRe(out, 'file1\t')
345
def test_recursive_diff(self):
346
"""Children of directories are matched"""
349
self.wt.add(['dir1', 'dir2'])
350
self.wt.rename_one('file1', 'dir1/file1')
351
old_tree = self.b.repository.revision_tree('rev-1')
352
new_tree = self.b.repository.revision_tree('rev-4')
353
out = self.get_diff(old_tree, new_tree, specific_files=['dir1'],
354
working_tree=self.wt)
355
self.assertContainsRe(out, 'file1\t')
356
out = self.get_diff(old_tree, new_tree, specific_files=['dir2'],
357
working_tree=self.wt)
358
self.assertNotContainsRe(out, 'file1\t')
362
class TestShowDiffTrees(TestShowDiffTreesHelper):
363
"""Direct tests for show_diff_trees"""
365
def test_modified_file(self):
366
"""Test when a file is modified."""
367
tree = self.make_branch_and_tree('tree')
368
self.build_tree_contents([('tree/file', 'contents\n')])
369
tree.add(['file'], ['file-id'])
370
tree.commit('one', rev_id='rev-1')
372
self.build_tree_contents([('tree/file', 'new contents\n')])
373
diff = self.get_diff(tree.basis_tree(), tree)
374
self.assertContainsRe(diff, "=== modified file 'file'\n")
375
self.assertContainsRe(diff, '--- old/file\t')
376
self.assertContainsRe(diff, '\\+\\+\\+ new/file\t')
377
self.assertContainsRe(diff, '-contents\n'
380
def test_modified_file_in_renamed_dir(self):
381
"""Test when a file is modified in a renamed directory."""
382
tree = self.make_branch_and_tree('tree')
383
self.build_tree(['tree/dir/'])
384
self.build_tree_contents([('tree/dir/file', 'contents\n')])
385
tree.add(['dir', 'dir/file'], ['dir-id', 'file-id'])
386
tree.commit('one', rev_id='rev-1')
388
tree.rename_one('dir', 'other')
389
self.build_tree_contents([('tree/other/file', 'new contents\n')])
390
diff = self.get_diff(tree.basis_tree(), tree)
391
self.assertContainsRe(diff, "=== renamed directory 'dir' => 'other'\n")
392
self.assertContainsRe(diff, "=== modified file 'other/file'\n")
393
# XXX: This is technically incorrect, because it used to be at another
394
# location. What to do?
395
self.assertContainsRe(diff, '--- old/dir/file\t')
396
self.assertContainsRe(diff, '\\+\\+\\+ new/other/file\t')
397
self.assertContainsRe(diff, '-contents\n'
400
def test_renamed_directory(self):
401
"""Test when only a directory is only renamed."""
402
tree = self.make_branch_and_tree('tree')
403
self.build_tree(['tree/dir/'])
404
self.build_tree_contents([('tree/dir/file', 'contents\n')])
405
tree.add(['dir', 'dir/file'], ['dir-id', 'file-id'])
406
tree.commit('one', rev_id='rev-1')
408
tree.rename_one('dir', 'newdir')
409
diff = self.get_diff(tree.basis_tree(), tree)
410
# Renaming a directory should be a single "you renamed this dir" even
411
# when there are files inside.
412
self.assertEqual("=== renamed directory 'dir' => 'newdir'\n", diff)
414
def test_renamed_file(self):
415
"""Test when a file is only renamed."""
416
tree = self.make_branch_and_tree('tree')
417
self.build_tree_contents([('tree/file', 'contents\n')])
418
tree.add(['file'], ['file-id'])
419
tree.commit('one', rev_id='rev-1')
421
tree.rename_one('file', 'newname')
422
diff = self.get_diff(tree.basis_tree(), tree)
423
self.assertContainsRe(diff, "=== renamed file 'file' => 'newname'\n")
424
# We shouldn't have a --- or +++ line, because there is no content
426
self.assertNotContainsRe(diff, '---')
428
def test_renamed_and_modified_file(self):
429
"""Test when a file is only renamed."""
430
tree = self.make_branch_and_tree('tree')
431
self.build_tree_contents([('tree/file', 'contents\n')])
432
tree.add(['file'], ['file-id'])
433
tree.commit('one', rev_id='rev-1')
435
tree.rename_one('file', 'newname')
436
self.build_tree_contents([('tree/newname', 'new contents\n')])
437
diff = self.get_diff(tree.basis_tree(), tree)
438
self.assertContainsRe(diff, "=== renamed file 'file' => 'newname'\n")
439
self.assertContainsRe(diff, '--- old/file\t')
440
self.assertContainsRe(diff, '\\+\\+\\+ new/newname\t')
441
self.assertContainsRe(diff, '-contents\n'
445
class TestPatienceDiffLib(TestCase):
447
def test_unique_lcs(self):
448
unique_lcs = bzrlib.patiencediff.unique_lcs
449
self.assertEquals(unique_lcs('', ''), [])
450
self.assertEquals(unique_lcs('a', 'a'), [(0,0)])
451
self.assertEquals(unique_lcs('a', 'b'), [])
452
self.assertEquals(unique_lcs('ab', 'ab'), [(0,0), (1,1)])
453
self.assertEquals(unique_lcs('abcde', 'cdeab'), [(2,0), (3,1), (4,2)])
454
self.assertEquals(unique_lcs('cdeab', 'abcde'), [(0,2), (1,3), (2,4)])
455
self.assertEquals(unique_lcs('abXde', 'abYde'), [(0,0), (1,1),
457
self.assertEquals(unique_lcs('acbac', 'abc'), [(2,1)])
459
def test_recurse_matches(self):
460
def test_one(a, b, matches):
462
bzrlib.patiencediff.recurse_matches(a, b, 0, 0, len(a), len(b),
464
self.assertEquals(test_matches, matches)
466
test_one(['a', '', 'b', '', 'c'], ['a', 'a', 'b', 'c', 'c'],
467
[(0, 0), (2, 2), (4, 4)])
468
test_one(['a', 'c', 'b', 'a', 'c'], ['a', 'b', 'c'],
469
[(0, 0), (2, 1), (4, 2)])
471
# recurse_matches doesn't match non-unique
472
# lines surrounded by bogus text.
473
# The update has been done in patiencediff.SequenceMatcher instead
475
# This is what it could be
476
#test_one('aBccDe', 'abccde', [(0,0), (2,2), (3,3), (5,5)])
478
# This is what it currently gives:
479
test_one('aBccDe', 'abccde', [(0,0), (5,5)])
481
def test_matching_blocks(self):
482
def chk_blocks(a, b, expected_blocks):
483
# difflib always adds a signature of the total
484
# length, with no matching entries at the end
485
s = bzrlib.patiencediff.PatienceSequenceMatcher(None, a, b)
486
blocks = s.get_matching_blocks()
487
self.assertEquals((len(a), len(b), 0), blocks[-1])
488
self.assertEquals(expected_blocks, blocks[:-1])
490
# Some basic matching tests
491
chk_blocks('', '', [])
492
chk_blocks([], [], [])
493
chk_blocks('abcd', 'abcd', [(0, 0, 4)])
494
chk_blocks('abcd', 'abce', [(0, 0, 3)])
495
chk_blocks('eabc', 'abce', [(1, 0, 3)])
496
chk_blocks('eabce', 'abce', [(1, 0, 4)])
497
chk_blocks('abcde', 'abXde', [(0, 0, 2), (3, 3, 2)])
498
chk_blocks('abcde', 'abXYZde', [(0, 0, 2), (3, 5, 2)])
499
chk_blocks('abde', 'abXYZde', [(0, 0, 2), (2, 5, 2)])
500
# This may check too much, but it checks to see that
501
# a copied block stays attached to the previous section,
503
# difflib would tend to grab the trailing longest match
504
# which would make the diff not look right
505
chk_blocks('abcdefghijklmnop', 'abcdefxydefghijklmnop',
506
[(0, 0, 6), (6, 11, 10)])
508
# make sure it supports passing in lists
512
'how are you today?\n'],
514
'how are you today?\n'],
515
[(0, 0, 1), (2, 1, 1)])
517
# non unique lines surrounded by non-matching lines
519
chk_blocks('aBccDe', 'abccde', [(0,0,1), (5,5,1)])
521
# But they only need to be locally unique
522
chk_blocks('aBcDec', 'abcdec', [(0,0,1), (2,2,1), (4,4,2)])
524
# non unique blocks won't be matched
525
chk_blocks('aBcdEcdFg', 'abcdecdfg', [(0,0,1), (8,8,1)])
527
# but locally unique ones will
528
chk_blocks('aBcdEeXcdFg', 'abcdecdfg', [(0,0,1), (2,2,2),
529
(5,4,1), (7,5,2), (10,8,1)])
531
chk_blocks('abbabbXd', 'cabbabxd', [(7,7,1)])
532
chk_blocks('abbabbbb', 'cabbabbc', [])
533
chk_blocks('bbbbbbbb', 'cbbbbbbc', [])
535
def test_opcodes(self):
536
def chk_ops(a, b, expected_codes):
537
s = bzrlib.patiencediff.PatienceSequenceMatcher(None, a, b)
538
self.assertEquals(expected_codes, s.get_opcodes())
542
chk_ops('abcd', 'abcd', [('equal', 0,4, 0,4)])
543
chk_ops('abcd', 'abce', [('equal', 0,3, 0,3),
544
('replace', 3,4, 3,4)
546
chk_ops('eabc', 'abce', [('delete', 0,1, 0,0),
550
chk_ops('eabce', 'abce', [('delete', 0,1, 0,0),
553
chk_ops('abcde', 'abXde', [('equal', 0,2, 0,2),
554
('replace', 2,3, 2,3),
557
chk_ops('abcde', 'abXYZde', [('equal', 0,2, 0,2),
558
('replace', 2,3, 2,5),
561
chk_ops('abde', 'abXYZde', [('equal', 0,2, 0,2),
562
('insert', 2,2, 2,5),
565
chk_ops('abcdefghijklmnop', 'abcdefxydefghijklmnop',
566
[('equal', 0,6, 0,6),
567
('insert', 6,6, 6,11),
568
('equal', 6,16, 11,21)
573
, 'how are you today?\n'],
575
, 'how are you today?\n'],
576
[('equal', 0,1, 0,1),
577
('delete', 1,2, 1,1),
580
chk_ops('aBccDe', 'abccde',
581
[('equal', 0,1, 0,1),
582
('replace', 1,5, 1,5),
585
chk_ops('aBcDec', 'abcdec',
586
[('equal', 0,1, 0,1),
587
('replace', 1,2, 1,2),
589
('replace', 3,4, 3,4),
592
chk_ops('aBcdEcdFg', 'abcdecdfg',
593
[('equal', 0,1, 0,1),
594
('replace', 1,8, 1,8),
597
chk_ops('aBcdEeXcdFg', 'abcdecdfg',
598
[('equal', 0,1, 0,1),
599
('replace', 1,2, 1,2),
601
('delete', 4,5, 4,4),
603
('delete', 6,7, 5,5),
605
('replace', 9,10, 7,8),
606
('equal', 10,11, 8,9)
609
def test_multiple_ranges(self):
610
# There was an earlier bug where we used a bad set of ranges,
611
# this triggers that specific bug, to make sure it doesn't regress
612
def chk_blocks(a, b, expected_blocks):
613
# difflib always adds a signature of the total
614
# length, with no matching entries at the end
615
s = bzrlib.patiencediff.PatienceSequenceMatcher(None, a, b)
616
blocks = s.get_matching_blocks()
618
self.assertEquals(x, (len(a), len(b), 0))
619
self.assertEquals(expected_blocks, blocks)
621
chk_blocks('abcdefghijklmnop'
622
, 'abcXghiYZQRSTUVWXYZijklmnop'
623
, [(0, 0, 3), (6, 4, 3), (9, 20, 7)])
625
chk_blocks('ABCd efghIjk L'
626
, 'AxyzBCn mo pqrstuvwI1 2 L'
627
, [(0,0,1), (1, 4, 2), (9, 19, 1), (12, 23, 3)])
629
# These are rot13 code snippets.
631
trg nqqrq jura lbh nqq n svyr va gur qverpgbel.
633
gnxrf_netf = ['svyr*']
634
gnxrf_bcgvbaf = ['ab-erphefr']
636
qrs eha(frys, svyr_yvfg, ab_erphefr=Snyfr):
637
sebz omeyvo.nqq vzcbeg fzneg_nqq, nqq_ercbegre_cevag, nqq_ercbegre_ahyy
639
ercbegre = nqq_ercbegre_ahyy
641
ercbegre = nqq_ercbegre_cevag
642
fzneg_nqq(svyr_yvfg, abg ab_erphefr, ercbegre)
645
pynff pzq_zxqve(Pbzznaq):
646
'''.splitlines(True), '''\
647
trg nqqrq jura lbh nqq n svyr va gur qverpgbel.
649
--qel-eha jvyy fubj juvpu svyrf jbhyq or nqqrq, ohg abg npghnyyl
652
gnxrf_netf = ['svyr*']
653
gnxrf_bcgvbaf = ['ab-erphefr', 'qel-eha']
655
qrs eha(frys, svyr_yvfg, ab_erphefr=Snyfr, qel_eha=Snyfr):
660
# Guvf vf cbvagyrff, ohg V'q engure abg envfr na reebe
661
npgvba = omeyvo.nqq.nqq_npgvba_ahyy
663
npgvba = omeyvo.nqq.nqq_npgvba_cevag
665
npgvba = omeyvo.nqq.nqq_npgvba_nqq
667
npgvba = omeyvo.nqq.nqq_npgvba_nqq_naq_cevag
669
omeyvo.nqq.fzneg_nqq(svyr_yvfg, abg ab_erphefr, npgvba)
672
pynff pzq_zxqve(Pbzznaq):
674
, [(0,0,1), (1, 4, 2), (9, 19, 1), (12, 23, 3)])
676
def test_patience_unified_diff(self):
677
txt_a = ['hello there\n',
679
'how are you today?\n']
680
txt_b = ['hello there\n',
681
'how are you today?\n']
682
unified_diff = bzrlib.patiencediff.unified_diff
683
psm = bzrlib.patiencediff.PatienceSequenceMatcher
684
self.assertEquals([ '--- \n',
689
' how are you today?\n'
691
, list(unified_diff(txt_a, txt_b,
692
sequencematcher=psm)))
693
txt_a = map(lambda x: x+'\n', 'abcdefghijklmnop')
694
txt_b = map(lambda x: x+'\n', 'abcdefxydefghijklmnop')
695
# This is the result with LongestCommonSubstring matching
696
self.assertEquals(['--- \n',
698
'@@ -1,6 +1,11 @@\n',
710
, list(unified_diff(txt_a, txt_b)))
711
# And the patience diff
712
self.assertEquals(['--- \n',
714
'@@ -4,6 +4,11 @@\n',
727
, list(unified_diff(txt_a, txt_b,
728
sequencematcher=psm)))
731
class TestPatienceDiffLibFiles(TestCaseInTempDir):
733
def test_patience_unified_diff_files(self):
734
txt_a = ['hello there\n',
736
'how are you today?\n']
737
txt_b = ['hello there\n',
738
'how are you today?\n']
739
open('a1', 'wb').writelines(txt_a)
740
open('b1', 'wb').writelines(txt_b)
742
unified_diff_files = bzrlib.patiencediff.unified_diff_files
743
psm = bzrlib.patiencediff.PatienceSequenceMatcher
744
self.assertEquals(['--- a1 \n',
749
' how are you today?\n',
751
, list(unified_diff_files('a1', 'b1',
752
sequencematcher=psm)))
754
txt_a = map(lambda x: x+'\n', 'abcdefghijklmnop')
755
txt_b = map(lambda x: x+'\n', 'abcdefxydefghijklmnop')
756
open('a2', 'wb').writelines(txt_a)
757
open('b2', 'wb').writelines(txt_b)
759
# This is the result with LongestCommonSubstring matching
760
self.assertEquals(['--- a2 \n',
762
'@@ -1,6 +1,11 @@\n',
774
, list(unified_diff_files('a2', 'b2')))
776
# And the patience diff
777
self.assertEquals(['--- a2 \n',
779
'@@ -4,6 +4,11 @@\n',
792
, list(unified_diff_files('a2', 'b2',
793
sequencematcher=psm)))
added
removed
bzrlib/tests/test_diff.py
