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
|
import asyncio
import zlib
from concurrent.futures import Executor
from typing import Optional, cast
try:
try:
import brotlicffi as brotli
except ImportError:
import brotli
HAS_BROTLI = True
except ImportError: # pragma: no cover
HAS_BROTLI = False
MAX_SYNC_CHUNK_SIZE = 1024
def encoding_to_mode(
encoding: Optional[str] = None,
suppress_deflate_header: bool = False,
) -> int:
if encoding == "gzip":
return 16 + zlib.MAX_WBITS
return -zlib.MAX_WBITS if suppress_deflate_header else zlib.MAX_WBITS
class ZlibBaseHandler:
def __init__(
self,
mode: int,
executor: Optional[Executor] = None,
max_sync_chunk_size: Optional[int] = MAX_SYNC_CHUNK_SIZE,
):
self._mode = mode
self._executor = executor
self._max_sync_chunk_size = max_sync_chunk_size
class ZLibCompressor(ZlibBaseHandler):
def __init__(
self,
encoding: Optional[str] = None,
suppress_deflate_header: bool = False,
level: Optional[int] = None,
wbits: Optional[int] = None,
strategy: int = zlib.Z_DEFAULT_STRATEGY,
executor: Optional[Executor] = None,
max_sync_chunk_size: Optional[int] = MAX_SYNC_CHUNK_SIZE,
):
super().__init__(
mode=(
encoding_to_mode(encoding, suppress_deflate_header)
if wbits is None
else wbits
),
executor=executor,
max_sync_chunk_size=max_sync_chunk_size,
)
if level is None:
self._compressor = zlib.compressobj(wbits=self._mode, strategy=strategy)
else:
self._compressor = zlib.compressobj(
wbits=self._mode, strategy=strategy, level=level
)
self._compress_lock = asyncio.Lock()
def compress_sync(self, data: bytes) -> bytes:
return self._compressor.compress(data)
async def compress(self, data: bytes) -> bytes:
"""Compress the data and returned the compressed bytes.
Note that flush() must be called after the last call to compress()
If the data size is large than the max_sync_chunk_size, the compression
will be done in the executor. Otherwise, the compression will be done
in the event loop.
"""
async with self._compress_lock:
# To ensure the stream is consistent in the event
# there are multiple writers, we need to lock
# the compressor so that only one writer can
# compress at a time.
if (
self._max_sync_chunk_size is not None
and len(data) > self._max_sync_chunk_size
):
return await asyncio.get_running_loop().run_in_executor(
self._executor, self._compressor.compress, data
)
return self.compress_sync(data)
def flush(self, mode: int = zlib.Z_FINISH) -> bytes:
return self._compressor.flush(mode)
class ZLibDecompressor(ZlibBaseHandler):
def __init__(
self,
encoding: Optional[str] = None,
suppress_deflate_header: bool = False,
executor: Optional[Executor] = None,
max_sync_chunk_size: Optional[int] = MAX_SYNC_CHUNK_SIZE,
):
super().__init__(
mode=encoding_to_mode(encoding, suppress_deflate_header),
executor=executor,
max_sync_chunk_size=max_sync_chunk_size,
)
self._decompressor = zlib.decompressobj(wbits=self._mode)
def decompress_sync(self, data: bytes, max_length: int = 0) -> bytes:
return self._decompressor.decompress(data, max_length)
async def decompress(self, data: bytes, max_length: int = 0) -> bytes:
"""Decompress the data and return the decompressed bytes.
If the data size is large than the max_sync_chunk_size, the decompression
will be done in the executor. Otherwise, the decompression will be done
in the event loop.
"""
if (
self._max_sync_chunk_size is not None
and len(data) > self._max_sync_chunk_size
):
return await asyncio.get_running_loop().run_in_executor(
self._executor, self._decompressor.decompress, data, max_length
)
return self.decompress_sync(data, max_length)
def flush(self, length: int = 0) -> bytes:
return (
self._decompressor.flush(length)
if length > 0
else self._decompressor.flush()
)
@property
def eof(self) -> bool:
return self._decompressor.eof
@property
def unconsumed_tail(self) -> bytes:
return self._decompressor.unconsumed_tail
@property
def unused_data(self) -> bytes:
return self._decompressor.unused_data
class BrotliDecompressor:
# Supports both 'brotlipy' and 'Brotli' packages
# since they share an import name. The top branches
# are for 'brotlipy' and bottom branches for 'Brotli'
def __init__(self) -> None:
if not HAS_BROTLI:
raise RuntimeError(
"The brotli decompression is not available. "
"Please install `Brotli` module"
)
self._obj = brotli.Decompressor()
def decompress_sync(self, data: bytes) -> bytes:
if hasattr(self._obj, "decompress"):
return cast(bytes, self._obj.decompress(data))
return cast(bytes, self._obj.process(data))
def flush(self) -> bytes:
if hasattr(self._obj, "flush"):
return cast(bytes, self._obj.flush())
return b""
|
