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
|
# -*- coding: utf-8 -*-
#
# Copyright (C) 2019 Radim Rehurek <me@radimrehurek.com>
#
# This code is distributed under the terms and conditions
# from the MIT License (MIT).
#
import gzip
import os
import os.path as P
import subprocess
from unittest import mock
import sys
import pytest
import smart_open.hdfs
CURR_DIR = P.dirname(P.abspath(__file__))
if sys.platform.startswith("win"):
pytest.skip("these tests don't work under Windows", allow_module_level=True)
#
# We want our mocks to emulate the real implementation as close as possible,
# so we use a Popen call during each test. If we mocked using io.BytesIO, then
# it is possible the mocks would behave differently to what we expect in real
# use.
#
# Since these tests use cat, they will not work in an environment without cat,
# such as Windows. The main line of this test submodule contains a simple
# cat implementation. We need this because Windows' analog, type, does
# weird stuff with line endings (inserts CRLF). Also, I don't know of a way
# to get type to echo standard input.
#
def cat(path=None):
command = [sys.executable, P.abspath(__file__)]
if path:
command.append(path)
return subprocess.Popen(command, stdin=subprocess.PIPE, stdout=subprocess.PIPE)
CAP_PATH = P.join(CURR_DIR, 'test_data', 'crime-and-punishment.txt')
with open(CAP_PATH, encoding='utf-8') as fin:
CRIME_AND_PUNISHMENT = fin.read()
def test_sanity_read_bytes():
with open(CAP_PATH, 'rb') as fin:
lines = [line for line in fin]
assert len(lines) == 3
def test_sanity_read_text():
with open(CAP_PATH, 'r', encoding='utf-8') as fin:
text = fin.read()
expected = 'В начале июля, в чрезвычайно жаркое время'
assert text[:len(expected)] == expected
@pytest.mark.parametrize('schema', [('hdfs', ), ('viewfs', )])
def test_read(schema):
with mock.patch('subprocess.Popen', return_value=cat(CAP_PATH)):
reader = smart_open.hdfs.CliRawInputBase(f'{schema}://dummy/url')
as_bytes = reader.read()
#
# Not 100% sure why this is necessary on Windows platforms, but the
# tests fail without it. It may be a bug, but I don't have time to
# investigate right now.
#
as_text = as_bytes.decode('utf-8').replace(os.linesep, '\n')
assert as_text == CRIME_AND_PUNISHMENT
@pytest.mark.parametrize('schema', [('hdfs', ), ('viewfs', )])
def test_read_75(schema):
with mock.patch('subprocess.Popen', return_value=cat(CAP_PATH)):
reader = smart_open.hdfs.CliRawInputBase(f'{schema}://dummy/url')
as_bytes = reader.read(75)
as_text = as_bytes.decode('utf-8').replace(os.linesep, '\n')
assert as_text == CRIME_AND_PUNISHMENT[:len(as_text)]
@pytest.mark.parametrize('schema', [('hdfs', ), ('viewfs', )])
def test_unzip(schema):
with mock.patch('subprocess.Popen', return_value=cat(CAP_PATH + '.gz')):
with gzip.GzipFile(fileobj=smart_open.hdfs.CliRawInputBase(f'{schema}://dummy/url')) as fin:
as_bytes = fin.read()
as_text = as_bytes.decode('utf-8')
assert as_text == CRIME_AND_PUNISHMENT
@pytest.mark.parametrize('schema', [('hdfs', ), ('viewfs', )])
def test_context_manager(schema):
with mock.patch('subprocess.Popen', return_value=cat(CAP_PATH)):
with smart_open.hdfs.CliRawInputBase(f'{schema}://dummy/url') as fin:
as_bytes = fin.read()
as_text = as_bytes.decode('utf-8').replace('\r\n', '\n')
assert as_text == CRIME_AND_PUNISHMENT
@pytest.mark.parametrize('schema', [('hdfs', ), ('viewfs', )])
def test_write(schema):
expected = 'мы в ответе за тех, кого приручили'
mocked_cat = cat()
with mock.patch('subprocess.Popen', return_value=mocked_cat):
with smart_open.hdfs.CliRawOutputBase(f'{schema}://dummy/url') as fout:
fout.write(expected.encode('utf-8'))
actual = mocked_cat.stdout.read().decode('utf-8')
assert actual == expected
@pytest.mark.parametrize('schema', [('hdfs', ), ('viewfs', )])
def test_write_zip(schema):
expected = 'мы в ответе за тех, кого приручили'
mocked_cat = cat()
with mock.patch('subprocess.Popen', return_value=mocked_cat):
with smart_open.hdfs.CliRawOutputBase(f'{schema}://dummy/url') as fout:
with gzip.GzipFile(fileobj=fout, mode='wb') as gz_fout:
gz_fout.write(expected.encode('utf-8'))
with gzip.GzipFile(fileobj=mocked_cat.stdout) as fin:
actual = fin.read().decode('utf-8')
assert actual == expected
def main():
try:
path = sys.argv[1]
except IndexError:
bytez = sys.stdin.buffer.read()
else:
with open(path, 'rb') as fin:
bytez = fin.read()
sys.stdout.buffer.write(bytez)
sys.stdout.flush()
if __name__ == '__main__':
main()
|
