1
from cStringIO import StringIO
3
from bzrlib.diff import internal_diff
4
from bzrlib.errors import BinaryFile
5
from bzrlib.patiencediff import (recurse_matches, SequenceMatcher, unique_lcs,
6
unified_diff, unified_diff_files)
7
from bzrlib.tests import TestCase, TestCaseInTempDir
10
def udiff_lines(old, new, allow_binary=False):
12
internal_diff('old', old, 'new', new, output, allow_binary)
14
return output.readlines()
17
class TestDiff(TestCase):
19
def test_add_nl(self):
20
"""diff generates a valid diff for patches that add a newline"""
21
lines = udiff_lines(['boo'], ['boo\n'])
22
self.check_patch(lines)
23
self.assertEquals(lines[4], '\\ No newline at end of file\n')
24
## "expected no-nl, got %r" % lines[4]
26
def test_add_nl_2(self):
27
"""diff generates a valid diff for patches that change last line and
30
lines = udiff_lines(['boo'], ['goo\n'])
31
self.check_patch(lines)
32
self.assertEquals(lines[4], '\\ No newline at end of file\n')
33
## "expected no-nl, got %r" % lines[4]
35
def test_remove_nl(self):
36
"""diff generates a valid diff for patches that change last line and
39
lines = udiff_lines(['boo\n'], ['boo'])
40
self.check_patch(lines)
41
self.assertEquals(lines[5], '\\ No newline at end of file\n')
42
## "expected no-nl, got %r" % lines[5]
44
def check_patch(self, lines):
45
self.assert_(len(lines) > 1)
46
## "Not enough lines for a file header for patch:\n%s" % "".join(lines)
47
self.assert_(lines[0].startswith ('---'))
48
## 'No orig line for patch:\n%s' % "".join(lines)
49
self.assert_(lines[1].startswith ('+++'))
50
## 'No mod line for patch:\n%s' % "".join(lines)
51
self.assert_(len(lines) > 2)
52
## "No hunks for patch:\n%s" % "".join(lines)
53
self.assert_(lines[2].startswith('@@'))
54
## "No hunk header for patch:\n%s" % "".join(lines)
55
self.assert_('@@' in lines[2][2:])
56
## "Unterminated hunk header for patch:\n%s" % "".join(lines)
58
def test_binary_lines(self):
59
self.assertRaises(BinaryFile, udiff_lines, [1023 * 'a' + '\x00'], [])
60
self.assertRaises(BinaryFile, udiff_lines, [], [1023 * 'a' + '\x00'])
61
udiff_lines([1023 * 'a' + '\x00'], [], allow_binary=True)
62
udiff_lines([], [1023 * 'a' + '\x00'], allow_binary=True)
65
class TestCDVDiffLib(TestCase):
67
def test_unique_lcs(self):
68
self.assertEquals(unique_lcs('', ''), [])
69
self.assertEquals(unique_lcs('a', 'a'), [(0,0)])
70
self.assertEquals(unique_lcs('a', 'b'), [])
71
self.assertEquals(unique_lcs('ab', 'ab'), [(0,0), (1,1)])
72
self.assertEquals(unique_lcs('abcde', 'cdeab'), [(2,0), (3,1), (4,2)])
73
self.assertEquals(unique_lcs('cdeab', 'abcde'), [(0,2), (1,3), (2,4)])
74
self.assertEquals(unique_lcs('abXde', 'abYde'), [(0,0), (1,1),
76
self.assertEquals(unique_lcs('acbac', 'abc'), [(2,1)])
78
def test_recurse_matches(self):
79
def test_one(a, b, matches):
81
recurse_matches(a, b, len(a), len(b), test_matches, 10)
82
self.assertEquals(test_matches, matches)
84
test_one(['a', None, 'b', None, 'c'], ['a', 'a', 'b', 'c', 'c'],
85
[(0, 0), (2, 2), (4, 4)])
86
test_one(['a', 'c', 'b', 'a', 'c'], ['a', 'b', 'c'],
87
[(0, 0), (2, 1), (4, 2)])
89
# recurse_matches doesn't match non-unique
90
# lines surrounded by bogus text.
91
# The update has been done in patiencediff.SequenceMatcher instead
93
# This is what it could be
94
#test_one('aBccDe', 'abccde', [(0,0), (2,2), (3,3), (5,5)])
96
# This is what it currently gives:
97
test_one('aBccDe', 'abccde', [(0,0), (5,5)])
99
def test_matching_blocks(self):
100
def chk_blocks(a, b, matching):
101
# difflib always adds a signature of the total
102
# length, with no matching entries at the end
103
s = SequenceMatcher(None, a, b)
104
blocks = s.get_matching_blocks()
106
self.assertEquals(x, (len(a), len(b), 0))
107
self.assertEquals(matching, blocks)
109
# Some basic matching tests
110
chk_blocks('', '', [])
111
chk_blocks([], [], [])
112
chk_blocks('abcd', 'abcd', [(0, 0, 4)])
113
chk_blocks('abcd', 'abce', [(0, 0, 3)])
114
chk_blocks('eabc', 'abce', [(1, 0, 3)])
115
chk_blocks('eabce', 'abce', [(1, 0, 4)])
116
chk_blocks('abcde', 'abXde', [(0, 0, 2), (3, 3, 2)])
117
chk_blocks('abcde', 'abXYZde', [(0, 0, 2), (3, 5, 2)])
118
chk_blocks('abde', 'abXYZde', [(0, 0, 2), (2, 5, 2)])
119
# This may check too much, but it checks to see that
120
# a copied block stays attached to the previous section,
122
# difflib would tend to grab the trailing longest match
123
# which would make the diff not look right
124
chk_blocks('abcdefghijklmnop', 'abcdefxydefghijklmnop',
125
[(0, 0, 6), (6, 11, 10)])
127
# make sure it supports passing in lists
131
'how are you today?\n'],
133
'how are you today?\n'],
134
[(0, 0, 1), (2, 1, 1)])
136
chk_blocks('aBccDe', 'abccde', [(0,0,1), (2,2,2), (5,5,1)])
138
chk_blocks('aBcdEcdFg', 'abcdecdfg', [(0,0,1), (2,2,2),
141
chk_blocks('abbabbXd', 'cabbabxd', [(0,1,5), (7,7,1)])
142
chk_blocks('abbabbbb', 'cabbabbc', [(0,1,6)])
143
chk_blocks('bbbbbbbb', 'cbbbbbbc', [(0,1,6)])
145
def test_opcodes(self):
146
def chk_ops(a, b, codes):
147
s = SequenceMatcher(None, a, b)
148
self.assertEquals(codes, s.get_opcodes())
152
chk_ops('abcd', 'abcd', [('equal', 0,4, 0,4)])
153
chk_ops('abcd', 'abce', [('equal', 0,3, 0,3),
154
('replace', 3,4, 3,4)
156
chk_ops('eabc', 'abce', [('delete', 0,1, 0,0),
160
chk_ops('eabce', 'abce', [('delete', 0,1, 0,0),
163
chk_ops('abcde', 'abXde', [('equal', 0,2, 0,2),
164
('replace', 2,3, 2,3),
167
chk_ops('abcde', 'abXYZde', [('equal', 0,2, 0,2),
168
('replace', 2,3, 2,5),
171
chk_ops('abde', 'abXYZde', [('equal', 0,2, 0,2),
172
('insert', 2,2, 2,5),
175
chk_ops('abcdefghijklmnop', 'abcdefxydefghijklmnop',
176
[('equal', 0,6, 0,6),
177
('insert', 6,6, 6,11),
178
('equal', 6,16, 11,21)
183
, 'how are you today?\n'],
185
, 'how are you today?\n'],
186
[('equal', 0,1, 0,1),
187
('delete', 1,2, 1,1),
190
chk_ops('aBccDe', 'abccde',
191
[('equal', 0,1, 0,1),
192
('replace', 1,2, 1,2),
194
('replace', 4,5, 4,5),
197
chk_ops('aBcdEcdFg', 'abcdecdfg',
198
[('equal', 0,1, 0,1),
199
('replace', 1,2, 1,2),
201
('replace', 4,5, 4,5),
203
('replace', 7,8, 7,8),
207
def test_multiple_ranges(self):
208
# There was an earlier bug where we used a bad set of ranges,
209
# this triggers that specific bug, to make sure it doesn't regress
210
def chk_blocks(a, b, matching):
211
# difflib always adds a signature of the total
212
# length, with no matching entries at the end
213
s = SequenceMatcher(None, a, b)
214
blocks = s.get_matching_blocks()
216
self.assertEquals(x, (len(a), len(b), 0))
217
self.assertEquals(matching, blocks)
219
chk_blocks('abcdefghijklmnop'
220
, 'abcXghiYZQRSTUVWXYZijklmnop'
221
, [(0, 0, 3), (6, 4, 3), (9, 20, 7)])
223
chk_blocks('ABCd efghIjk L'
224
, 'AxyzBCn mo pqrstuvwI1 2 L'
225
, [(0,0,1), (1, 4, 2), (4, 7, 1), (9, 19, 1), (12, 23, 3)])
228
get added when you add a file in the directory.
230
takes_args = ['file*']
231
takes_options = ['no-recurse']
233
def run(self, file_list, no_recurse=False):
234
from bzrlib.add import smart_add, add_reporter_print, add_reporter_null
236
reporter = add_reporter_null
238
reporter = add_reporter_print
239
smart_add(file_list, not no_recurse, reporter)
242
class cmd_mkdir(Command):
245
get added when you add a file in the directory.
247
--dry-run will show which files would be added, but not actually
250
takes_args = ['file*']
251
takes_options = ['no-recurse', 'dry-run']
253
def run(self, file_list, no_recurse=False, dry_run=False):
258
# This is pointless, but I'd rather not raise an error
259
action = bzrlib.add.add_action_null
261
action = bzrlib.add.add_action_print
263
action = bzrlib.add.add_action_add
265
action = bzrlib.add.add_action_add_and_print
267
bzrlib.add.smart_add(file_list, not no_recurse, action)
270
class cmd_mkdir(Command):
272
, [(0,0,1), (1, 4, 2), (9, 19, 1), (12, 23, 3)])
274
def test_cdv_unified_diff(self):
275
txt_a = ['hello there\n',
277
'how are you today?\n']
278
txt_b = ['hello there\n',
279
'how are you today?\n']
280
self.assertEquals([ '--- \n',
285
' how are you today?\n'
287
, list(unified_diff(txt_a, txt_b
288
, sequencematcher=SequenceMatcher)))
289
txt_a = map(lambda x: x+'\n', 'abcdefghijklmnop')
290
txt_b = map(lambda x: x+'\n', 'abcdefxydefghijklmnop')
291
# This is the result with LongestCommonSubstring matching
292
self.assertEquals(['--- \n',
294
'@@ -1,6 +1,11 @@\n',
306
, list(unified_diff(txt_a, txt_b)))
308
self.assertEquals(['--- \n',
310
'@@ -4,6 +4,11 @@\n',
323
, list(unified_diff(txt_a, txt_b,
324
sequencematcher=SequenceMatcher)))
327
class TestCDVDiffLibFiles(TestCaseInTempDir):
329
def test_cdv_unified_diff_files(self):
330
txt_a = ['hello there\n',
332
'how are you today?\n']
333
txt_b = ['hello there\n',
334
'how are you today?\n']
335
open('a1', 'wb').writelines(txt_a)
336
open('b1', 'wb').writelines(txt_b)
338
self.assertEquals(['--- a1 \n',
343
' how are you today?\n',
345
, list(unified_diff_files('a1', 'b1',
346
sequencematcher=SequenceMatcher)))
348
txt_a = map(lambda x: x+'\n', 'abcdefghijklmnop')
349
txt_b = map(lambda x: x+'\n', 'abcdefxydefghijklmnop')
350
open('a2', 'wb').writelines(txt_a)
351
open('b2', 'wb').writelines(txt_b)
353
# This is the result with LongestCommonSubstring matching
354
self.assertEquals(['--- a2 \n',
356
'@@ -1,6 +1,11 @@\n',
368
, list(unified_diff_files('a2', 'b2')))
371
self.assertEquals(['--- a2 \n',
373
'@@ -4,6 +4,11 @@\n',
386
, list(unified_diff_files('a2', 'b2',
387
sequencematcher=SequenceMatcher)))
added
removed
bzrlib/tests/test_diff.py
