1
# Copyright (C) 2005, 2006 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
from cStringIO import StringIO
21
from tempfile import TemporaryFile
23
from bzrlib.diff import internal_diff, external_diff, show_diff_trees
24
from bzrlib.errors import BinaryFile, NoDiff
25
import bzrlib.patiencediff
26
from bzrlib.tests import (TestCase, TestCaseWithTransport,
27
TestCaseInTempDir, TestSkipped)
30
def udiff_lines(old, new, allow_binary=False):
32
internal_diff('old', old, 'new', new, output, allow_binary)
34
return output.readlines()
37
def external_udiff_lines(old, new, use_stringio=False):
39
# StringIO has no fileno, so it tests a different codepath
42
output = TemporaryFile()
44
external_diff('old', old, 'new', new, output, diff_opts=['-u'])
46
raise TestSkipped('external "diff" not present to test')
48
lines = output.readlines()
53
class TestDiff(TestCase):
55
def test_add_nl(self):
56
"""diff generates a valid diff for patches that add a newline"""
57
lines = udiff_lines(['boo'], ['boo\n'])
58
self.check_patch(lines)
59
self.assertEquals(lines[4], '\\ No newline at end of file\n')
60
## "expected no-nl, got %r" % lines[4]
62
def test_add_nl_2(self):
63
"""diff generates a valid diff for patches that change last line and
66
lines = udiff_lines(['boo'], ['goo\n'])
67
self.check_patch(lines)
68
self.assertEquals(lines[4], '\\ No newline at end of file\n')
69
## "expected no-nl, got %r" % lines[4]
71
def test_remove_nl(self):
72
"""diff generates a valid diff for patches that change last line and
75
lines = udiff_lines(['boo\n'], ['boo'])
76
self.check_patch(lines)
77
self.assertEquals(lines[5], '\\ No newline at end of file\n')
78
## "expected no-nl, got %r" % lines[5]
80
def check_patch(self, lines):
81
self.assert_(len(lines) > 1)
82
## "Not enough lines for a file header for patch:\n%s" % "".join(lines)
83
self.assert_(lines[0].startswith ('---'))
84
## 'No orig line for patch:\n%s' % "".join(lines)
85
self.assert_(lines[1].startswith ('+++'))
86
## 'No mod line for patch:\n%s' % "".join(lines)
87
self.assert_(len(lines) > 2)
88
## "No hunks for patch:\n%s" % "".join(lines)
89
self.assert_(lines[2].startswith('@@'))
90
## "No hunk header for patch:\n%s" % "".join(lines)
91
self.assert_('@@' in lines[2][2:])
92
## "Unterminated hunk header for patch:\n%s" % "".join(lines)
94
def test_binary_lines(self):
95
self.assertRaises(BinaryFile, udiff_lines, [1023 * 'a' + '\x00'], [])
96
self.assertRaises(BinaryFile, udiff_lines, [], [1023 * 'a' + '\x00'])
97
udiff_lines([1023 * 'a' + '\x00'], [], allow_binary=True)
98
udiff_lines([], [1023 * 'a' + '\x00'], allow_binary=True)
100
def test_external_diff(self):
101
lines = external_udiff_lines(['boo\n'], ['goo\n'])
102
self.check_patch(lines)
103
self.assertEqual('\n', lines[-1])
105
def test_external_diff_no_fileno(self):
106
# Make sure that we can handle not having a fileno, even
107
# if the diff is large
108
lines = external_udiff_lines(['boo\n']*10000,
111
self.check_patch(lines)
113
def test_external_diff_binary_lang_c(self):
114
orig_lang = os.environ.get('LANG')
115
orig_lc_all = os.environ.get('LC_ALL')
117
os.environ['LANG'] = 'C'
118
os.environ['LC_ALL'] = '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 name, value in [('LANG', orig_lang), ('LC_ALL', orig_lc_all)]:
130
os.environ[name] = value
132
def test_no_external_diff(self):
133
"""Check that NoDiff is raised when diff is not available"""
134
# Use os.environ['PATH'] to make sure no 'diff' command is available
135
orig_path = os.environ['PATH']
137
os.environ['PATH'] = ''
138
self.assertRaises(NoDiff, external_diff,
139
'old', ['boo\n'], 'new', ['goo\n'],
140
StringIO(), diff_opts=['-u'])
142
os.environ['PATH'] = orig_path
144
def test_internal_diff_default(self):
145
# Default internal diff encoding is utf8
147
internal_diff(u'old_\xb5', ['old_text\n'],
148
u'new_\xe5', ['new_text\n'], output)
149
lines = output.getvalue().splitlines(True)
150
self.check_patch(lines)
151
self.assertEquals(['--- old_\xc2\xb5\n',
152
'+++ new_\xc3\xa5\n',
160
def test_internal_diff_utf8(self):
162
internal_diff(u'old_\xb5', ['old_text\n'],
163
u'new_\xe5', ['new_text\n'], output,
164
path_encoding='utf8')
165
lines = output.getvalue().splitlines(True)
166
self.check_patch(lines)
167
self.assertEquals(['--- old_\xc2\xb5\n',
168
'+++ new_\xc3\xa5\n',
176
def test_internal_diff_iso_8859_1(self):
178
internal_diff(u'old_\xb5', ['old_text\n'],
179
u'new_\xe5', ['new_text\n'], output,
180
path_encoding='iso-8859-1')
181
lines = output.getvalue().splitlines(True)
182
self.check_patch(lines)
183
self.assertEquals(['--- old_\xb5\n',
192
def test_internal_diff_returns_bytes(self):
194
output = StringIO.StringIO()
195
internal_diff(u'old_\xb5', ['old_text\n'],
196
u'new_\xe5', ['new_text\n'], output)
197
self.failUnless(isinstance(output.getvalue(), str),
198
'internal_diff should return bytestrings')
201
class TestDiffFiles(TestCaseInTempDir):
203
def test_external_diff_binary(self):
204
"""The output when using external diff should use diff's i18n error"""
205
# Make sure external_diff doesn't fail in the current LANG
206
lines = external_udiff_lines(['\x00foobar\n'], ['foo\x00bar\n'])
208
cmd = ['diff', '-u', 'old', 'new']
209
open('old', 'wb').write('\x00foobar\n')
210
open('new', 'wb').write('foo\x00bar\n')
211
pipe = subprocess.Popen(cmd, stdout=subprocess.PIPE,
212
stdin=subprocess.PIPE)
213
out, err = pipe.communicate()
214
# Diff returns '2' on Binary files.
215
self.assertEqual(2, pipe.returncode)
216
# We should output whatever diff tells us, plus a trailing newline
217
self.assertEqual(out.splitlines(True) + ['\n'], lines)
220
class TestDiffDates(TestCaseWithTransport):
223
super(TestDiffDates, self).setUp()
224
self.wt = self.make_branch_and_tree('.')
225
self.b = self.wt.branch
226
self.build_tree_contents([
227
('file1', 'file1 contents at rev 1\n'),
228
('file2', 'file2 contents at rev 1\n')
230
self.wt.add(['file1', 'file2'])
232
message='Revision 1',
233
timestamp=1143849600, # 2006-04-01 00:00:00 UTC
236
self.build_tree_contents([('file1', 'file1 contents at rev 2\n')])
238
message='Revision 2',
239
timestamp=1143936000, # 2006-04-02 00:00:00 UTC
242
self.build_tree_contents([('file2', 'file2 contents at rev 3\n')])
244
message='Revision 3',
245
timestamp=1144022400, # 2006-04-03 00:00:00 UTC
248
self.wt.remove(['file2'])
250
message='Revision 4',
251
timestamp=1144108800, # 2006-04-04 00:00:00 UTC
254
self.build_tree_contents([
255
('file1', 'file1 contents in working tree\n')
257
# set the date stamps for files in the working tree to known values
258
os.utime('file1', (1144195200, 1144195200)) # 2006-04-05 00:00:00 UTC
260
def get_diff(self, tree1, tree2, specific_files=None, working_tree=None):
262
if working_tree is not None:
263
extra_trees = (working_tree,)
266
show_diff_trees(tree1, tree2, output, specific_files=specific_files,
267
extra_trees=extra_trees, old_label='old/',
269
return output.getvalue()
271
def test_diff_rev_tree_working_tree(self):
272
output = self.get_diff(self.wt.basis_tree(), self.wt)
273
# note that the date for old/file1 is from rev 2 rather than from
274
# the basis revision (rev 4)
275
self.assertEqualDiff(output, '''\
276
=== modified file 'file1'
277
--- old/file1\t2006-04-02 00:00:00 +0000
278
+++ new/file1\t2006-04-05 00:00:00 +0000
280
-file1 contents at rev 2
281
+file1 contents in working tree
285
def test_diff_rev_tree_rev_tree(self):
286
tree1 = self.b.repository.revision_tree('rev-2')
287
tree2 = self.b.repository.revision_tree('rev-3')
288
output = self.get_diff(tree1, tree2)
289
self.assertEqualDiff(output, '''\
290
=== modified file 'file2'
291
--- old/file2\t2006-04-01 00:00:00 +0000
292
+++ new/file2\t2006-04-03 00:00:00 +0000
294
-file2 contents at rev 1
295
+file2 contents at rev 3
299
def test_diff_add_files(self):
300
tree1 = self.b.repository.revision_tree(None)
301
tree2 = self.b.repository.revision_tree('rev-1')
302
output = self.get_diff(tree1, tree2)
303
# the files have the epoch time stamp for the tree in which
305
self.assertEqualDiff(output, '''\
306
=== added file 'file1'
307
--- old/file1\t1970-01-01 00:00:00 +0000
308
+++ new/file1\t2006-04-01 00:00:00 +0000
310
+file1 contents at rev 1
312
=== added file 'file2'
313
--- old/file2\t1970-01-01 00:00:00 +0000
314
+++ new/file2\t2006-04-01 00:00:00 +0000
316
+file2 contents at rev 1
320
def test_diff_remove_files(self):
321
tree1 = self.b.repository.revision_tree('rev-3')
322
tree2 = self.b.repository.revision_tree('rev-4')
323
output = self.get_diff(tree1, tree2)
324
# the file has the epoch time stamp for the tree in which
326
self.assertEqualDiff(output, '''\
327
=== removed file 'file2'
328
--- old/file2\t2006-04-03 00:00:00 +0000
329
+++ new/file2\t1970-01-01 00:00:00 +0000
331
-file2 contents at rev 3
335
def test_show_diff_specified(self):
336
"""A working tree filename can be used to identify a file"""
337
self.wt.rename_one('file1', 'file1b')
338
old_tree = self.b.repository.revision_tree('rev-1')
339
new_tree = self.b.repository.revision_tree('rev-4')
340
out = self.get_diff(old_tree, new_tree, specific_files=['file1b'],
341
working_tree=self.wt)
342
self.assertContainsRe(out, 'file1\t')
344
def test_recursive_diff(self):
345
"""Children of directories are matched"""
348
self.wt.add(['dir1', 'dir2'])
349
self.wt.rename_one('file1', 'dir1/file1')
350
old_tree = self.b.repository.revision_tree('rev-1')
351
new_tree = self.b.repository.revision_tree('rev-4')
352
out = self.get_diff(old_tree, new_tree, specific_files=['dir1'],
353
working_tree=self.wt)
354
self.assertContainsRe(out, 'file1\t')
355
out = self.get_diff(old_tree, new_tree, specific_files=['dir2'],
356
working_tree=self.wt)
357
self.assertNotContainsRe(out, 'file1\t')
360
class TestPatienceDiffLib(TestCase):
362
def test_unique_lcs(self):
363
unique_lcs = bzrlib.patiencediff.unique_lcs
364
self.assertEquals(unique_lcs('', ''), [])
365
self.assertEquals(unique_lcs('a', 'a'), [(0,0)])
366
self.assertEquals(unique_lcs('a', 'b'), [])
367
self.assertEquals(unique_lcs('ab', 'ab'), [(0,0), (1,1)])
368
self.assertEquals(unique_lcs('abcde', 'cdeab'), [(2,0), (3,1), (4,2)])
369
self.assertEquals(unique_lcs('cdeab', 'abcde'), [(0,2), (1,3), (2,4)])
370
self.assertEquals(unique_lcs('abXde', 'abYde'), [(0,0), (1,1),
372
self.assertEquals(unique_lcs('acbac', 'abc'), [(2,1)])
374
def test_recurse_matches(self):
375
def test_one(a, b, matches):
377
bzrlib.patiencediff.recurse_matches(a, b, 0, 0, len(a), len(b),
379
self.assertEquals(test_matches, matches)
381
test_one(['a', '', 'b', '', 'c'], ['a', 'a', 'b', 'c', 'c'],
382
[(0, 0), (2, 2), (4, 4)])
383
test_one(['a', 'c', 'b', 'a', 'c'], ['a', 'b', 'c'],
384
[(0, 0), (2, 1), (4, 2)])
386
# recurse_matches doesn't match non-unique
387
# lines surrounded by bogus text.
388
# The update has been done in patiencediff.SequenceMatcher instead
390
# This is what it could be
391
#test_one('aBccDe', 'abccde', [(0,0), (2,2), (3,3), (5,5)])
393
# This is what it currently gives:
394
test_one('aBccDe', 'abccde', [(0,0), (5,5)])
396
def test_matching_blocks(self):
397
def chk_blocks(a, b, expected_blocks):
398
# difflib always adds a signature of the total
399
# length, with no matching entries at the end
400
s = bzrlib.patiencediff.PatienceSequenceMatcher(None, a, b)
401
blocks = s.get_matching_blocks()
402
self.assertEquals((len(a), len(b), 0), blocks[-1])
403
self.assertEquals(expected_blocks, blocks[:-1])
405
# Some basic matching tests
406
chk_blocks('', '', [])
407
chk_blocks([], [], [])
408
chk_blocks('abcd', 'abcd', [(0, 0, 4)])
409
chk_blocks('abcd', 'abce', [(0, 0, 3)])
410
chk_blocks('eabc', 'abce', [(1, 0, 3)])
411
chk_blocks('eabce', 'abce', [(1, 0, 4)])
412
chk_blocks('abcde', 'abXde', [(0, 0, 2), (3, 3, 2)])
413
chk_blocks('abcde', 'abXYZde', [(0, 0, 2), (3, 5, 2)])
414
chk_blocks('abde', 'abXYZde', [(0, 0, 2), (2, 5, 2)])
415
# This may check too much, but it checks to see that
416
# a copied block stays attached to the previous section,
418
# difflib would tend to grab the trailing longest match
419
# which would make the diff not look right
420
chk_blocks('abcdefghijklmnop', 'abcdefxydefghijklmnop',
421
[(0, 0, 6), (6, 11, 10)])
423
# make sure it supports passing in lists
427
'how are you today?\n'],
429
'how are you today?\n'],
430
[(0, 0, 1), (2, 1, 1)])
432
# non unique lines surrounded by non-matching lines
434
chk_blocks('aBccDe', 'abccde', [(0,0,1), (5,5,1)])
436
# But they only need to be locally unique
437
chk_blocks('aBcDec', 'abcdec', [(0,0,1), (2,2,1), (4,4,2)])
439
# non unique blocks won't be matched
440
chk_blocks('aBcdEcdFg', 'abcdecdfg', [(0,0,1), (8,8,1)])
442
# but locally unique ones will
443
chk_blocks('aBcdEeXcdFg', 'abcdecdfg', [(0,0,1), (2,2,2),
444
(5,4,1), (7,5,2), (10,8,1)])
446
chk_blocks('abbabbXd', 'cabbabxd', [(7,7,1)])
447
chk_blocks('abbabbbb', 'cabbabbc', [])
448
chk_blocks('bbbbbbbb', 'cbbbbbbc', [])
450
def test_opcodes(self):
451
def chk_ops(a, b, expected_codes):
452
s = bzrlib.patiencediff.PatienceSequenceMatcher(None, a, b)
453
self.assertEquals(expected_codes, s.get_opcodes())
457
chk_ops('abcd', 'abcd', [('equal', 0,4, 0,4)])
458
chk_ops('abcd', 'abce', [('equal', 0,3, 0,3),
459
('replace', 3,4, 3,4)
461
chk_ops('eabc', 'abce', [('delete', 0,1, 0,0),
465
chk_ops('eabce', 'abce', [('delete', 0,1, 0,0),
468
chk_ops('abcde', 'abXde', [('equal', 0,2, 0,2),
469
('replace', 2,3, 2,3),
472
chk_ops('abcde', 'abXYZde', [('equal', 0,2, 0,2),
473
('replace', 2,3, 2,5),
476
chk_ops('abde', 'abXYZde', [('equal', 0,2, 0,2),
477
('insert', 2,2, 2,5),
480
chk_ops('abcdefghijklmnop', 'abcdefxydefghijklmnop',
481
[('equal', 0,6, 0,6),
482
('insert', 6,6, 6,11),
483
('equal', 6,16, 11,21)
488
, 'how are you today?\n'],
490
, 'how are you today?\n'],
491
[('equal', 0,1, 0,1),
492
('delete', 1,2, 1,1),
495
chk_ops('aBccDe', 'abccde',
496
[('equal', 0,1, 0,1),
497
('replace', 1,5, 1,5),
500
chk_ops('aBcDec', 'abcdec',
501
[('equal', 0,1, 0,1),
502
('replace', 1,2, 1,2),
504
('replace', 3,4, 3,4),
507
chk_ops('aBcdEcdFg', 'abcdecdfg',
508
[('equal', 0,1, 0,1),
509
('replace', 1,8, 1,8),
512
chk_ops('aBcdEeXcdFg', 'abcdecdfg',
513
[('equal', 0,1, 0,1),
514
('replace', 1,2, 1,2),
516
('delete', 4,5, 4,4),
518
('delete', 6,7, 5,5),
520
('replace', 9,10, 7,8),
521
('equal', 10,11, 8,9)
524
def test_multiple_ranges(self):
525
# There was an earlier bug where we used a bad set of ranges,
526
# this triggers that specific bug, to make sure it doesn't regress
527
def chk_blocks(a, b, expected_blocks):
528
# difflib always adds a signature of the total
529
# length, with no matching entries at the end
530
s = bzrlib.patiencediff.PatienceSequenceMatcher(None, a, b)
531
blocks = s.get_matching_blocks()
533
self.assertEquals(x, (len(a), len(b), 0))
534
self.assertEquals(expected_blocks, blocks)
536
chk_blocks('abcdefghijklmnop'
537
, 'abcXghiYZQRSTUVWXYZijklmnop'
538
, [(0, 0, 3), (6, 4, 3), (9, 20, 7)])
540
chk_blocks('ABCd efghIjk L'
541
, 'AxyzBCn mo pqrstuvwI1 2 L'
542
, [(0,0,1), (1, 4, 2), (9, 19, 1), (12, 23, 3)])
544
# These are rot13 code snippets.
546
trg nqqrq jura lbh nqq n svyr va gur qverpgbel.
548
gnxrf_netf = ['svyr*']
549
gnxrf_bcgvbaf = ['ab-erphefr']
551
qrs eha(frys, svyr_yvfg, ab_erphefr=Snyfr):
552
sebz omeyvo.nqq vzcbeg fzneg_nqq, nqq_ercbegre_cevag, nqq_ercbegre_ahyy
554
ercbegre = nqq_ercbegre_ahyy
556
ercbegre = nqq_ercbegre_cevag
557
fzneg_nqq(svyr_yvfg, abg ab_erphefr, ercbegre)
560
pynff pzq_zxqve(Pbzznaq):
561
'''.splitlines(True), '''\
562
trg nqqrq jura lbh nqq n svyr va gur qverpgbel.
564
--qel-eha jvyy fubj juvpu svyrf jbhyq or nqqrq, ohg abg npghnyyl
567
gnxrf_netf = ['svyr*']
568
gnxrf_bcgvbaf = ['ab-erphefr', 'qel-eha']
570
qrs eha(frys, svyr_yvfg, ab_erphefr=Snyfr, qel_eha=Snyfr):
575
# Guvf vf cbvagyrff, ohg V'q engure abg envfr na reebe
576
npgvba = omeyvo.nqq.nqq_npgvba_ahyy
578
npgvba = omeyvo.nqq.nqq_npgvba_cevag
580
npgvba = omeyvo.nqq.nqq_npgvba_nqq
582
npgvba = omeyvo.nqq.nqq_npgvba_nqq_naq_cevag
584
omeyvo.nqq.fzneg_nqq(svyr_yvfg, abg ab_erphefr, npgvba)
587
pynff pzq_zxqve(Pbzznaq):
589
, [(0,0,1), (1, 4, 2), (9, 19, 1), (12, 23, 3)])
591
def test_patience_unified_diff(self):
592
txt_a = ['hello there\n',
594
'how are you today?\n']
595
txt_b = ['hello there\n',
596
'how are you today?\n']
597
unified_diff = bzrlib.patiencediff.unified_diff
598
psm = bzrlib.patiencediff.PatienceSequenceMatcher
599
self.assertEquals([ '--- \n',
604
' how are you today?\n'
606
, list(unified_diff(txt_a, txt_b,
607
sequencematcher=psm)))
608
txt_a = map(lambda x: x+'\n', 'abcdefghijklmnop')
609
txt_b = map(lambda x: x+'\n', 'abcdefxydefghijklmnop')
610
# This is the result with LongestCommonSubstring matching
611
self.assertEquals(['--- \n',
613
'@@ -1,6 +1,11 @@\n',
625
, list(unified_diff(txt_a, txt_b)))
626
# And the patience diff
627
self.assertEquals(['--- \n',
629
'@@ -4,6 +4,11 @@\n',
642
, list(unified_diff(txt_a, txt_b,
643
sequencematcher=psm)))
646
class TestPatienceDiffLibFiles(TestCaseInTempDir):
648
def test_patience_unified_diff_files(self):
649
txt_a = ['hello there\n',
651
'how are you today?\n']
652
txt_b = ['hello there\n',
653
'how are you today?\n']
654
open('a1', 'wb').writelines(txt_a)
655
open('b1', 'wb').writelines(txt_b)
657
unified_diff_files = bzrlib.patiencediff.unified_diff_files
658
psm = bzrlib.patiencediff.PatienceSequenceMatcher
659
self.assertEquals(['--- a1 \n',
664
' how are you today?\n',
666
, list(unified_diff_files('a1', 'b1',
667
sequencematcher=psm)))
669
txt_a = map(lambda x: x+'\n', 'abcdefghijklmnop')
670
txt_b = map(lambda x: x+'\n', 'abcdefxydefghijklmnop')
671
open('a2', 'wb').writelines(txt_a)
672
open('b2', 'wb').writelines(txt_b)
674
# This is the result with LongestCommonSubstring matching
675
self.assertEquals(['--- a2 \n',
677
'@@ -1,6 +1,11 @@\n',
689
, list(unified_diff_files('a2', 'b2')))
691
# And the patience diff
692
self.assertEquals(['--- a2 \n',
694
'@@ -4,6 +4,11 @@\n',
707
, list(unified_diff_files('a2', 'b2',
708
sequencematcher=psm)))
added
removed
bzrlib/tests/test_diff.py
