18
18
"""Tests for tuned_gzip."""
21
from io import BytesIO
29
class TestToGzip(tests.TestCase):
21
# do not use bzrlib test cases here - this should be suitable for sending
23
from cStringIO import StringIO
24
from unittest import TestCase
28
from bzrlib import tuned_gzip
31
class FakeDecompress(object):
32
"""A fake decompressor for testing GzipFile."""
37
def decompress(self, buf):
38
"""Return an empty string as though we are at eof."""
39
# note that the zlib module *overwrites* unused data
40
# on writes after EOF.
41
self.unused_data = buf
45
class TestFakeDecompress(TestCase):
46
"""We use a fake decompressor to test GzipFile.
48
This class tests the behaviours we want from it.
51
def test_decompress(self):
52
# decompressing returns no data.
53
decompress = FakeDecompress()
54
self.assertEqual('', decompress.decompress('0'))
56
def test_unused_data(self):
57
# after decompressing, we have 1 unused byte.
58
# this is normally set by decompressors when they
59
# detect the end of a compressed stream.
60
decompress = FakeDecompress()
61
decompress.decompress('0')
62
self.assertEqual('0', decompress.unused_data)
63
# decompressing again (when the short read is read)
64
# will give us the latest input in the unused_data
65
# this is arguably a bug in zlib but ...
66
decompress.decompress('1234567')
67
self.assertEqual('1234567', decompress.unused_data)
70
class TestGzip(TestCase):
72
def test__read_short_remainder(self):
73
# a _read call at the end of a compressed hunk should
74
# read more bytes if there is less than 8 bytes (the
75
# gzip trailer) unread.
76
stream = StringIO('\0\0\0\0\0\0\0\0')
77
myfile = tuned_gzip.GzipFile(fileobj=stream)
78
# disable the _new_member check, we are microtesting.
79
myfile._new_member = False
80
myfile.crc = zlib.crc32('')
81
myfile.decompress = FakeDecompress()
84
# all the data should have been read now
85
self.assertEqual('', stream.read())
86
# and it should be new member time in the stream.
87
self.failUnless(myfile._new_member)
90
class TestToGzip(TestCase):
31
92
def assertToGzip(self, chunks):
32
raw_bytes = b''.join(chunks)
93
bytes = ''.join(chunks)
33
94
gzfromchunks = tuned_gzip.chunks_to_gzip(chunks)
34
decoded = gzip.GzipFile(fileobj=BytesIO(b''.join(gzfromchunks))).read()
35
lraw, ldecoded = len(raw_bytes), len(decoded)
36
self.assertEqual(lraw, ldecoded,
37
'Expecting data length %d, got %d' % (lraw, ldecoded))
38
self.assertEqual(raw_bytes, decoded)
95
gzfrombytes = tuned_gzip.bytes_to_gzip(bytes)
96
self.assertEqual(gzfrombytes, gzfromchunks)
97
decoded = tuned_gzip.GzipFile(fileobj=StringIO(gzfromchunks)).read()
98
self.assertEqual(bytes, decoded)
40
100
def test_single_chunk(self):
41
self.assertToGzip([b'a modest chunk\nwith some various\nbits\n'])
101
self.assertToGzip(['a modest chunk\nwith some various\nbits\n'])
43
103
def test_simple_text(self):
44
self.assertToGzip([b'some\n', b'strings\n', b'to\n', b'process\n'])
104
self.assertToGzip(['some\n', 'strings\n', 'to\n', 'process\n'])
46
106
def test_large_chunks(self):
47
self.assertToGzip([b'a large string\n' * 1024])
48
self.assertToGzip([b'a large string\n'] * 1024)
107
self.assertToGzip(['a large string\n'*1024])
108
self.assertToGzip(['a large string\n']*1024)
50
110
def test_enormous_chunks(self):
51
self.assertToGzip([b'a large string\n' * 1024 * 256])
52
self.assertToGzip([b'a large string\n'] * 1024 * 256)
111
self.assertToGzip(['a large string\n'*1024*256])
112
self.assertToGzip(['a large string\n']*1024*256)