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
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
|
"""Test script for the gzip module.
"""
import array
import functools
import io
import os
import pathlib
import struct
import sys
import unittest
from subprocess import PIPE, Popen
from test import support
from test.support import _4G, bigmemtest
from test.support.script_helper import assert_python_ok, assert_python_failure
gzip = support.import_module('gzip')
data1 = b""" int length=DEFAULTALLOC, err = Z_OK;
PyObject *RetVal;
int flushmode = Z_FINISH;
unsigned long start_total_out;
"""
data2 = b"""/* zlibmodule.c -- gzip-compatible data compression */
/* See http://www.gzip.org/zlib/
/* See http://www.winimage.com/zLibDll for Windows */
"""
TEMPDIR = os.path.abspath(support.TESTFN) + '-gzdir'
class UnseekableIO(io.BytesIO):
def seekable(self):
return False
def tell(self):
raise io.UnsupportedOperation
def seek(self, *args):
raise io.UnsupportedOperation
class BaseTest(unittest.TestCase):
filename = support.TESTFN
def setUp(self):
support.unlink(self.filename)
def tearDown(self):
support.unlink(self.filename)
class TestGzip(BaseTest):
def write_and_read_back(self, data, mode='b'):
b_data = bytes(data)
with gzip.GzipFile(self.filename, 'w'+mode) as f:
l = f.write(data)
self.assertEqual(l, len(b_data))
with gzip.GzipFile(self.filename, 'r'+mode) as f:
self.assertEqual(f.read(), b_data)
def test_write(self):
with gzip.GzipFile(self.filename, 'wb') as f:
f.write(data1 * 50)
# Try flush and fileno.
f.flush()
f.fileno()
if hasattr(os, 'fsync'):
os.fsync(f.fileno())
f.close()
# Test multiple close() calls.
f.close()
def test_write_read_with_pathlike_file(self):
filename = pathlib.Path(self.filename)
with gzip.GzipFile(filename, 'w') as f:
f.write(data1 * 50)
self.assertIsInstance(f.name, str)
with gzip.GzipFile(filename, 'a') as f:
f.write(data1)
with gzip.GzipFile(filename) as f:
d = f.read()
self.assertEqual(d, data1 * 51)
self.assertIsInstance(f.name, str)
# The following test_write_xy methods test that write accepts
# the corresponding bytes-like object type as input
# and that the data written equals bytes(xy) in all cases.
def test_write_memoryview(self):
self.write_and_read_back(memoryview(data1 * 50))
m = memoryview(bytes(range(256)))
data = m.cast('B', shape=[8,8,4])
self.write_and_read_back(data)
def test_write_bytearray(self):
self.write_and_read_back(bytearray(data1 * 50))
def test_write_array(self):
self.write_and_read_back(array.array('I', data1 * 40))
def test_write_incompatible_type(self):
# Test that non-bytes-like types raise TypeError.
# Issue #21560: attempts to write incompatible types
# should not affect the state of the fileobject
with gzip.GzipFile(self.filename, 'wb') as f:
with self.assertRaises(TypeError):
f.write('')
with self.assertRaises(TypeError):
f.write([])
f.write(data1)
with gzip.GzipFile(self.filename, 'rb') as f:
self.assertEqual(f.read(), data1)
def test_read(self):
self.test_write()
# Try reading.
with gzip.GzipFile(self.filename, 'r') as f:
d = f.read()
self.assertEqual(d, data1*50)
def test_read1(self):
self.test_write()
blocks = []
nread = 0
with gzip.GzipFile(self.filename, 'r') as f:
while True:
d = f.read1()
if not d:
break
blocks.append(d)
nread += len(d)
# Check that position was updated correctly (see issue10791).
self.assertEqual(f.tell(), nread)
self.assertEqual(b''.join(blocks), data1 * 50)
@bigmemtest(size=_4G, memuse=1)
def test_read_large(self, size):
# Read chunk size over UINT_MAX should be supported, despite zlib's
# limitation per low-level call
compressed = gzip.compress(data1, compresslevel=1)
f = gzip.GzipFile(fileobj=io.BytesIO(compressed), mode='rb')
self.assertEqual(f.read(size), data1)
def test_io_on_closed_object(self):
# Test that I/O operations on closed GzipFile objects raise a
# ValueError, just like the corresponding functions on file objects.
# Write to a file, open it for reading, then close it.
self.test_write()
f = gzip.GzipFile(self.filename, 'r')
fileobj = f.fileobj
self.assertFalse(fileobj.closed)
f.close()
self.assertTrue(fileobj.closed)
with self.assertRaises(ValueError):
f.read(1)
with self.assertRaises(ValueError):
f.seek(0)
with self.assertRaises(ValueError):
f.tell()
# Open the file for writing, then close it.
f = gzip.GzipFile(self.filename, 'w')
fileobj = f.fileobj
self.assertFalse(fileobj.closed)
f.close()
self.assertTrue(fileobj.closed)
with self.assertRaises(ValueError):
f.write(b'')
with self.assertRaises(ValueError):
f.flush()
def test_append(self):
self.test_write()
# Append to the previous file
with gzip.GzipFile(self.filename, 'ab') as f:
f.write(data2 * 15)
with gzip.GzipFile(self.filename, 'rb') as f:
d = f.read()
self.assertEqual(d, (data1*50) + (data2*15))
def test_many_append(self):
# Bug #1074261 was triggered when reading a file that contained
# many, many members. Create such a file and verify that reading it
# works.
with gzip.GzipFile(self.filename, 'wb', 9) as f:
f.write(b'a')
for i in range(0, 200):
with gzip.GzipFile(self.filename, "ab", 9) as f: # append
f.write(b'a')
# Try reading the file
with gzip.GzipFile(self.filename, "rb") as zgfile:
contents = b""
while 1:
ztxt = zgfile.read(8192)
contents += ztxt
if not ztxt: break
self.assertEqual(contents, b'a'*201)
def test_exclusive_write(self):
with gzip.GzipFile(self.filename, 'xb') as f:
f.write(data1 * 50)
with gzip.GzipFile(self.filename, 'rb') as f:
self.assertEqual(f.read(), data1 * 50)
with self.assertRaises(FileExistsError):
gzip.GzipFile(self.filename, 'xb')
def test_buffered_reader(self):
# Issue #7471: a GzipFile can be wrapped in a BufferedReader for
# performance.
self.test_write()
with gzip.GzipFile(self.filename, 'rb') as f:
with io.BufferedReader(f) as r:
lines = [line for line in r]
self.assertEqual(lines, 50 * data1.splitlines(keepends=True))
def test_readline(self):
self.test_write()
# Try .readline() with varying line lengths
with gzip.GzipFile(self.filename, 'rb') as f:
line_length = 0
while 1:
L = f.readline(line_length)
if not L and line_length != 0: break
self.assertTrue(len(L) <= line_length)
line_length = (line_length + 1) % 50
def test_readlines(self):
self.test_write()
# Try .readlines()
with gzip.GzipFile(self.filename, 'rb') as f:
L = f.readlines()
with gzip.GzipFile(self.filename, 'rb') as f:
while 1:
L = f.readlines(150)
if L == []: break
def test_seek_read(self):
self.test_write()
# Try seek, read test
with gzip.GzipFile(self.filename) as f:
while 1:
oldpos = f.tell()
line1 = f.readline()
if not line1: break
newpos = f.tell()
f.seek(oldpos) # negative seek
if len(line1)>10:
amount = 10
else:
amount = len(line1)
line2 = f.read(amount)
self.assertEqual(line1[:amount], line2)
f.seek(newpos) # positive seek
def test_seek_whence(self):
self.test_write()
# Try seek(whence=1), read test
with gzip.GzipFile(self.filename) as f:
f.read(10)
f.seek(10, whence=1)
y = f.read(10)
self.assertEqual(y, data1[20:30])
def test_seek_write(self):
# Try seek, write test
with gzip.GzipFile(self.filename, 'w') as f:
for pos in range(0, 256, 16):
f.seek(pos)
f.write(b'GZ\n')
def test_mode(self):
self.test_write()
with gzip.GzipFile(self.filename, 'r') as f:
self.assertEqual(f.myfileobj.mode, 'rb')
support.unlink(self.filename)
with gzip.GzipFile(self.filename, 'x') as f:
self.assertEqual(f.myfileobj.mode, 'xb')
def test_1647484(self):
for mode in ('wb', 'rb'):
with gzip.GzipFile(self.filename, mode) as f:
self.assertTrue(hasattr(f, "name"))
self.assertEqual(f.name, self.filename)
def test_paddedfile_getattr(self):
self.test_write()
with gzip.GzipFile(self.filename, 'rb') as f:
self.assertTrue(hasattr(f.fileobj, "name"))
self.assertEqual(f.fileobj.name, self.filename)
def test_mtime(self):
mtime = 123456789
with gzip.GzipFile(self.filename, 'w', mtime = mtime) as fWrite:
fWrite.write(data1)
with gzip.GzipFile(self.filename) as fRead:
self.assertTrue(hasattr(fRead, 'mtime'))
self.assertIsNone(fRead.mtime)
dataRead = fRead.read()
self.assertEqual(dataRead, data1)
self.assertEqual(fRead.mtime, mtime)
def test_metadata(self):
mtime = 123456789
with gzip.GzipFile(self.filename, 'w', mtime = mtime) as fWrite:
fWrite.write(data1)
with open(self.filename, 'rb') as fRead:
# see RFC 1952: http://www.faqs.org/rfcs/rfc1952.html
idBytes = fRead.read(2)
self.assertEqual(idBytes, b'\x1f\x8b') # gzip ID
cmByte = fRead.read(1)
self.assertEqual(cmByte, b'\x08') # deflate
flagsByte = fRead.read(1)
self.assertEqual(flagsByte, b'\x08') # only the FNAME flag is set
mtimeBytes = fRead.read(4)
self.assertEqual(mtimeBytes, struct.pack('<i', mtime)) # little-endian
xflByte = fRead.read(1)
self.assertEqual(xflByte, b'\x02') # maximum compression
osByte = fRead.read(1)
self.assertEqual(osByte, b'\xff') # OS "unknown" (OS-independent)
# Since the FNAME flag is set, the zero-terminated filename follows.
# RFC 1952 specifies that this is the name of the input file, if any.
# However, the gzip module defaults to storing the name of the output
# file in this field.
expected = self.filename.encode('Latin-1') + b'\x00'
nameBytes = fRead.read(len(expected))
self.assertEqual(nameBytes, expected)
# Since no other flags were set, the header ends here.
# Rather than process the compressed data, let's seek to the trailer.
fRead.seek(os.stat(self.filename).st_size - 8)
crc32Bytes = fRead.read(4) # CRC32 of uncompressed data [data1]
self.assertEqual(crc32Bytes, b'\xaf\xd7d\x83')
isizeBytes = fRead.read(4)
self.assertEqual(isizeBytes, struct.pack('<i', len(data1)))
def test_with_open(self):
# GzipFile supports the context management protocol
with gzip.GzipFile(self.filename, "wb") as f:
f.write(b"xxx")
f = gzip.GzipFile(self.filename, "rb")
f.close()
try:
with f:
pass
except ValueError:
pass
else:
self.fail("__enter__ on a closed file didn't raise an exception")
try:
with gzip.GzipFile(self.filename, "wb") as f:
1/0
except ZeroDivisionError:
pass
else:
self.fail("1/0 didn't raise an exception")
def test_zero_padded_file(self):
with gzip.GzipFile(self.filename, "wb") as f:
f.write(data1 * 50)
# Pad the file with zeroes
with open(self.filename, "ab") as f:
f.write(b"\x00" * 50)
with gzip.GzipFile(self.filename, "rb") as f:
d = f.read()
self.assertEqual(d, data1 * 50, "Incorrect data in file")
def test_gzip_BadGzipFile_exception(self):
self.assertTrue(issubclass(gzip.BadGzipFile, OSError))
def test_bad_gzip_file(self):
with open(self.filename, 'wb') as file:
file.write(data1 * 50)
with gzip.GzipFile(self.filename, 'r') as file:
self.assertRaises(gzip.BadGzipFile, file.readlines)
def test_non_seekable_file(self):
uncompressed = data1 * 50
buf = UnseekableIO()
with gzip.GzipFile(fileobj=buf, mode="wb") as f:
f.write(uncompressed)
compressed = buf.getvalue()
buf = UnseekableIO(compressed)
with gzip.GzipFile(fileobj=buf, mode="rb") as f:
self.assertEqual(f.read(), uncompressed)
def test_peek(self):
uncompressed = data1 * 200
with gzip.GzipFile(self.filename, "wb") as f:
f.write(uncompressed)
def sizes():
while True:
for n in range(5, 50, 10):
yield n
with gzip.GzipFile(self.filename, "rb") as f:
f.max_read_chunk = 33
nread = 0
for n in sizes():
s = f.peek(n)
if s == b'':
break
self.assertEqual(f.read(len(s)), s)
nread += len(s)
self.assertEqual(f.read(100), b'')
self.assertEqual(nread, len(uncompressed))
def test_textio_readlines(self):
# Issue #10791: TextIOWrapper.readlines() fails when wrapping GzipFile.
lines = (data1 * 50).decode("ascii").splitlines(keepends=True)
self.test_write()
with gzip.GzipFile(self.filename, 'r') as f:
with io.TextIOWrapper(f, encoding="ascii") as t:
self.assertEqual(t.readlines(), lines)
def test_fileobj_from_fdopen(self):
# Issue #13781: Opening a GzipFile for writing fails when using a
# fileobj created with os.fdopen().
fd = os.open(self.filename, os.O_WRONLY | os.O_CREAT)
with os.fdopen(fd, "wb") as f:
with gzip.GzipFile(fileobj=f, mode="w") as g:
pass
def test_fileobj_mode(self):
gzip.GzipFile(self.filename, "wb").close()
with open(self.filename, "r+b") as f:
with gzip.GzipFile(fileobj=f, mode='r') as g:
self.assertEqual(g.mode, gzip.READ)
with gzip.GzipFile(fileobj=f, mode='w') as g:
self.assertEqual(g.mode, gzip.WRITE)
with gzip.GzipFile(fileobj=f, mode='a') as g:
self.assertEqual(g.mode, gzip.WRITE)
with gzip.GzipFile(fileobj=f, mode='x') as g:
self.assertEqual(g.mode, gzip.WRITE)
with self.assertRaises(ValueError):
gzip.GzipFile(fileobj=f, mode='z')
for mode in "rb", "r+b":
with open(self.filename, mode) as f:
with gzip.GzipFile(fileobj=f) as g:
self.assertEqual(g.mode, gzip.READ)
for mode in "wb", "ab", "xb":
if "x" in mode:
support.unlink(self.filename)
with open(self.filename, mode) as f:
with self.assertWarns(FutureWarning):
g = gzip.GzipFile(fileobj=f)
with g:
self.assertEqual(g.mode, gzip.WRITE)
def test_bytes_filename(self):
str_filename = self.filename
try:
bytes_filename = str_filename.encode("ascii")
except UnicodeEncodeError:
self.skipTest("Temporary file name needs to be ASCII")
with gzip.GzipFile(bytes_filename, "wb") as f:
f.write(data1 * 50)
with gzip.GzipFile(bytes_filename, "rb") as f:
self.assertEqual(f.read(), data1 * 50)
# Sanity check that we are actually operating on the right file.
with gzip.GzipFile(str_filename, "rb") as f:
self.assertEqual(f.read(), data1 * 50)
def test_decompress_limited(self):
"""Decompressed data buffering should be limited"""
bomb = gzip.compress(b'\0' * int(2e6), compresslevel=9)
self.assertLess(len(bomb), io.DEFAULT_BUFFER_SIZE)
bomb = io.BytesIO(bomb)
decomp = gzip.GzipFile(fileobj=bomb)
self.assertEqual(decomp.read(1), b'\0')
max_decomp = 1 + io.DEFAULT_BUFFER_SIZE
self.assertLessEqual(decomp._buffer.raw.tell(), max_decomp,
"Excessive amount of data was decompressed")
# Testing compress/decompress shortcut functions
def test_compress(self):
for data in [data1, data2]:
for args in [(), (1,), (6,), (9,)]:
datac = gzip.compress(data, *args)
self.assertEqual(type(datac), bytes)
with gzip.GzipFile(fileobj=io.BytesIO(datac), mode="rb") as f:
self.assertEqual(f.read(), data)
def test_compress_mtime(self):
mtime = 123456789
for data in [data1, data2]:
for args in [(), (1,), (6,), (9,)]:
with self.subTest(data=data, args=args):
datac = gzip.compress(data, *args, mtime=mtime)
self.assertEqual(type(datac), bytes)
with gzip.GzipFile(fileobj=io.BytesIO(datac), mode="rb") as f:
f.read(1) # to set mtime attribute
self.assertEqual(f.mtime, mtime)
def test_decompress(self):
for data in (data1, data2):
buf = io.BytesIO()
with gzip.GzipFile(fileobj=buf, mode="wb") as f:
f.write(data)
self.assertEqual(gzip.decompress(buf.getvalue()), data)
# Roundtrip with compress
datac = gzip.compress(data)
self.assertEqual(gzip.decompress(datac), data)
def test_read_truncated(self):
data = data1*50
# Drop the CRC (4 bytes) and file size (4 bytes).
truncated = gzip.compress(data)[:-8]
with gzip.GzipFile(fileobj=io.BytesIO(truncated)) as f:
self.assertRaises(EOFError, f.read)
with gzip.GzipFile(fileobj=io.BytesIO(truncated)) as f:
self.assertEqual(f.read(len(data)), data)
self.assertRaises(EOFError, f.read, 1)
# Incomplete 10-byte header.
for i in range(2, 10):
with gzip.GzipFile(fileobj=io.BytesIO(truncated[:i])) as f:
self.assertRaises(EOFError, f.read, 1)
def test_read_with_extra(self):
# Gzip data with an extra field
gzdata = (b'\x1f\x8b\x08\x04\xb2\x17cQ\x02\xff'
b'\x05\x00Extra'
b'\x0bI-.\x01\x002\xd1Mx\x04\x00\x00\x00')
with gzip.GzipFile(fileobj=io.BytesIO(gzdata)) as f:
self.assertEqual(f.read(), b'Test')
def test_prepend_error(self):
# See issue #20875
with gzip.open(self.filename, "wb") as f:
f.write(data1)
with gzip.open(self.filename, "rb") as f:
f._buffer.raw._fp.prepend()
class TestOpen(BaseTest):
def test_binary_modes(self):
uncompressed = data1 * 50
with gzip.open(self.filename, "wb") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read())
self.assertEqual(file_data, uncompressed)
with gzip.open(self.filename, "rb") as f:
self.assertEqual(f.read(), uncompressed)
with gzip.open(self.filename, "ab") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read())
self.assertEqual(file_data, uncompressed * 2)
with self.assertRaises(FileExistsError):
gzip.open(self.filename, "xb")
support.unlink(self.filename)
with gzip.open(self.filename, "xb") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read())
self.assertEqual(file_data, uncompressed)
def test_pathlike_file(self):
filename = pathlib.Path(self.filename)
with gzip.open(filename, "wb") as f:
f.write(data1 * 50)
with gzip.open(filename, "ab") as f:
f.write(data1)
with gzip.open(filename) as f:
self.assertEqual(f.read(), data1 * 51)
def test_implicit_binary_modes(self):
# Test implicit binary modes (no "b" or "t" in mode string).
uncompressed = data1 * 50
with gzip.open(self.filename, "w") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read())
self.assertEqual(file_data, uncompressed)
with gzip.open(self.filename, "r") as f:
self.assertEqual(f.read(), uncompressed)
with gzip.open(self.filename, "a") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read())
self.assertEqual(file_data, uncompressed * 2)
with self.assertRaises(FileExistsError):
gzip.open(self.filename, "x")
support.unlink(self.filename)
with gzip.open(self.filename, "x") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read())
self.assertEqual(file_data, uncompressed)
def test_text_modes(self):
uncompressed = data1.decode("ascii") * 50
uncompressed_raw = uncompressed.replace("\n", os.linesep)
with gzip.open(self.filename, "wt") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read()).decode("ascii")
self.assertEqual(file_data, uncompressed_raw)
with gzip.open(self.filename, "rt") as f:
self.assertEqual(f.read(), uncompressed)
with gzip.open(self.filename, "at") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read()).decode("ascii")
self.assertEqual(file_data, uncompressed_raw * 2)
def test_fileobj(self):
uncompressed_bytes = data1 * 50
uncompressed_str = uncompressed_bytes.decode("ascii")
compressed = gzip.compress(uncompressed_bytes)
with gzip.open(io.BytesIO(compressed), "r") as f:
self.assertEqual(f.read(), uncompressed_bytes)
with gzip.open(io.BytesIO(compressed), "rb") as f:
self.assertEqual(f.read(), uncompressed_bytes)
with gzip.open(io.BytesIO(compressed), "rt") as f:
self.assertEqual(f.read(), uncompressed_str)
def test_bad_params(self):
# Test invalid parameter combinations.
with self.assertRaises(TypeError):
gzip.open(123.456)
with self.assertRaises(ValueError):
gzip.open(self.filename, "wbt")
with self.assertRaises(ValueError):
gzip.open(self.filename, "xbt")
with self.assertRaises(ValueError):
gzip.open(self.filename, "rb", encoding="utf-8")
with self.assertRaises(ValueError):
gzip.open(self.filename, "rb", errors="ignore")
with self.assertRaises(ValueError):
gzip.open(self.filename, "rb", newline="\n")
def test_encoding(self):
# Test non-default encoding.
uncompressed = data1.decode("ascii") * 50
uncompressed_raw = uncompressed.replace("\n", os.linesep)
with gzip.open(self.filename, "wt", encoding="utf-16") as f:
f.write(uncompressed)
with open(self.filename, "rb") as f:
file_data = gzip.decompress(f.read()).decode("utf-16")
self.assertEqual(file_data, uncompressed_raw)
with gzip.open(self.filename, "rt", encoding="utf-16") as f:
self.assertEqual(f.read(), uncompressed)
def test_encoding_error_handler(self):
# Test with non-default encoding error handler.
with gzip.open(self.filename, "wb") as f:
f.write(b"foo\xffbar")
with gzip.open(self.filename, "rt", encoding="ascii", errors="ignore") \
as f:
self.assertEqual(f.read(), "foobar")
def test_newline(self):
# Test with explicit newline (universal newline mode disabled).
uncompressed = data1.decode("ascii") * 50
with gzip.open(self.filename, "wt", newline="\n") as f:
f.write(uncompressed)
with gzip.open(self.filename, "rt", newline="\r") as f:
self.assertEqual(f.readlines(), [uncompressed])
def create_and_remove_directory(directory):
def decorator(function):
@functools.wraps(function)
def wrapper(*args, **kwargs):
os.makedirs(directory)
try:
return function(*args, **kwargs)
finally:
support.rmtree(directory)
return wrapper
return decorator
class TestCommandLine(unittest.TestCase):
data = b'This is a simple test with gzip'
def test_decompress_stdin_stdout(self):
with io.BytesIO() as bytes_io:
with gzip.GzipFile(fileobj=bytes_io, mode='wb') as gzip_file:
gzip_file.write(self.data)
args = sys.executable, '-m', 'gzip', '-d'
with Popen(args, stdin=PIPE, stdout=PIPE, stderr=PIPE) as proc:
out, err = proc.communicate(bytes_io.getvalue())
self.assertEqual(err, b'')
self.assertEqual(out, self.data)
@create_and_remove_directory(TEMPDIR)
def test_decompress_infile_outfile(self):
gzipname = os.path.join(TEMPDIR, 'testgzip.gz')
self.assertFalse(os.path.exists(gzipname))
with gzip.open(gzipname, mode='wb') as fp:
fp.write(self.data)
rc, out, err = assert_python_ok('-m', 'gzip', '-d', gzipname)
with open(os.path.join(TEMPDIR, "testgzip"), "rb") as gunziped:
self.assertEqual(gunziped.read(), self.data)
self.assertTrue(os.path.exists(gzipname))
self.assertEqual(rc, 0)
self.assertEqual(out, b'')
self.assertEqual(err, b'')
def test_decompress_infile_outfile_error(self):
rc, out, err = assert_python_ok('-m', 'gzip', '-d', 'thisisatest.out')
self.assertIn(b"filename doesn't end in .gz:", out)
self.assertEqual(rc, 0)
self.assertEqual(err, b'')
@create_and_remove_directory(TEMPDIR)
def test_compress_stdin_outfile(self):
args = sys.executable, '-m', 'gzip'
with Popen(args, stdin=PIPE, stdout=PIPE, stderr=PIPE) as proc:
out, err = proc.communicate(self.data)
self.assertEqual(err, b'')
self.assertEqual(out[:2], b"\x1f\x8b")
@create_and_remove_directory(TEMPDIR)
def test_compress_infile_outfile_default(self):
local_testgzip = os.path.join(TEMPDIR, 'testgzip')
gzipname = local_testgzip + '.gz'
self.assertFalse(os.path.exists(gzipname))
with open(local_testgzip, 'wb') as fp:
fp.write(self.data)
rc, out, err = assert_python_ok('-m', 'gzip', local_testgzip)
self.assertTrue(os.path.exists(gzipname))
self.assertEqual(out, b'')
self.assertEqual(err, b'')
@create_and_remove_directory(TEMPDIR)
def test_compress_infile_outfile(self):
for compress_level in ('--fast', '--best'):
with self.subTest(compress_level=compress_level):
local_testgzip = os.path.join(TEMPDIR, 'testgzip')
gzipname = local_testgzip + '.gz'
self.assertFalse(os.path.exists(gzipname))
with open(local_testgzip, 'wb') as fp:
fp.write(self.data)
rc, out, err = assert_python_ok('-m', 'gzip', compress_level, local_testgzip)
self.assertTrue(os.path.exists(gzipname))
self.assertEqual(out, b'')
self.assertEqual(err, b'')
os.remove(gzipname)
self.assertFalse(os.path.exists(gzipname))
def test_compress_fast_best_are_exclusive(self):
rc, out, err = assert_python_failure('-m', 'gzip', '--fast', '--best')
self.assertIn(b"error: argument --best: not allowed with argument --fast", err)
self.assertEqual(out, b'')
def test_decompress_cannot_have_flags_compression(self):
rc, out, err = assert_python_failure('-m', 'gzip', '--fast', '-d')
self.assertIn(b'error: argument -d/--decompress: not allowed with argument --fast', err)
self.assertEqual(out, b'')
def test_main(verbose=None):
support.run_unittest(TestGzip, TestOpen, TestCommandLine)
if __name__ == "__main__":
test_main(verbose=True)
|