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
|
"""Lazy ZIP over HTTP"""
__all__ = ["HTTPRangeRequestUnsupported", "dist_from_wheel_url"]
from bisect import bisect_left, bisect_right
from contextlib import contextmanager
from tempfile import NamedTemporaryFile
from typing import Any, Dict, Iterator, List, Optional, Tuple
from zipfile import BadZipfile, ZipFile
from pip._vendor.pkg_resources import Distribution
from pip._vendor.requests.models import CONTENT_CHUNK_SIZE, Response
from pip._internal.network.session import PipSession
from pip._internal.network.utils import HEADERS, raise_for_status, response_chunks
from pip._internal.utils.wheel import pkg_resources_distribution_for_wheel
class HTTPRangeRequestUnsupported(Exception):
pass
def dist_from_wheel_url(name: str, url: str, session: PipSession) -> Distribution:
"""Return a pkg_resources.Distribution from the given wheel URL.
This uses HTTP range requests to only fetch the potion of the wheel
containing metadata, just enough for the object to be constructed.
If such requests are not supported, HTTPRangeRequestUnsupported
is raised.
"""
with LazyZipOverHTTP(url, session) as wheel:
# For read-only ZIP files, ZipFile only needs methods read,
# seek, seekable and tell, not the whole IO protocol.
zip_file = ZipFile(wheel) # type: ignore
# After context manager exit, wheel.name
# is an invalid file by intention.
return pkg_resources_distribution_for_wheel(zip_file, name, wheel.name)
class LazyZipOverHTTP:
"""File-like object mapped to a ZIP file over HTTP.
This uses HTTP range requests to lazily fetch the file's content,
which is supposed to be fed to ZipFile. If such requests are not
supported by the server, raise HTTPRangeRequestUnsupported
during initialization.
"""
def __init__(
self, url: str, session: PipSession, chunk_size: int = CONTENT_CHUNK_SIZE
) -> None:
head = session.head(url, headers=HEADERS)
raise_for_status(head)
assert head.status_code == 200
self._session, self._url, self._chunk_size = session, url, chunk_size
self._length = int(head.headers["Content-Length"])
self._file = NamedTemporaryFile()
self.truncate(self._length)
self._left: List[int] = []
self._right: List[int] = []
if "bytes" not in head.headers.get("Accept-Ranges", "none"):
raise HTTPRangeRequestUnsupported("range request is not supported")
self._check_zip()
@property
def mode(self) -> str:
"""Opening mode, which is always rb."""
return "rb"
@property
def name(self) -> str:
"""Path to the underlying file."""
return self._file.name
def seekable(self) -> bool:
"""Return whether random access is supported, which is True."""
return True
def close(self) -> None:
"""Close the file."""
self._file.close()
@property
def closed(self) -> bool:
"""Whether the file is closed."""
return self._file.closed
def read(self, size: int = -1) -> bytes:
"""Read up to size bytes from the object and return them.
As a convenience, if size is unspecified or -1,
all bytes until EOF are returned. Fewer than
size bytes may be returned if EOF is reached.
"""
download_size = max(size, self._chunk_size)
start, length = self.tell(), self._length
stop = length if size < 0 else min(start + download_size, length)
start = max(0, stop - download_size)
self._download(start, stop - 1)
return self._file.read(size)
def readable(self) -> bool:
"""Return whether the file is readable, which is True."""
return True
def seek(self, offset: int, whence: int = 0) -> int:
"""Change stream position and return the new absolute position.
Seek to offset relative position indicated by whence:
* 0: Start of stream (the default). pos should be >= 0;
* 1: Current position - pos may be negative;
* 2: End of stream - pos usually negative.
"""
return self._file.seek(offset, whence)
def tell(self) -> int:
"""Return the current position."""
return self._file.tell()
def truncate(self, size: Optional[int] = None) -> int:
"""Resize the stream to the given size in bytes.
If size is unspecified resize to the current position.
The current stream position isn't changed.
Return the new file size.
"""
return self._file.truncate(size)
def writable(self) -> bool:
"""Return False."""
return False
def __enter__(self) -> "LazyZipOverHTTP":
self._file.__enter__()
return self
def __exit__(self, *exc: Any) -> Optional[bool]:
return self._file.__exit__(*exc)
@contextmanager
def _stay(self) -> Iterator[None]:
"""Return a context manager keeping the position.
At the end of the block, seek back to original position.
"""
pos = self.tell()
try:
yield
finally:
self.seek(pos)
def _check_zip(self) -> None:
"""Check and download until the file is a valid ZIP."""
end = self._length - 1
for start in reversed(range(0, end, self._chunk_size)):
self._download(start, end)
with self._stay():
try:
# For read-only ZIP files, ZipFile only needs
# methods read, seek, seekable and tell.
ZipFile(self) # type: ignore
except BadZipfile:
pass
else:
break
def _stream_response(
self, start: int, end: int, base_headers: Dict[str, str] = HEADERS
) -> Response:
"""Return HTTP response to a range request from start to end."""
headers = base_headers.copy()
headers["Range"] = f"bytes={start}-{end}"
# TODO: Get range requests to be correctly cached
headers["Cache-Control"] = "no-cache"
return self._session.get(self._url, headers=headers, stream=True)
def _merge(
self, start: int, end: int, left: int, right: int
) -> Iterator[Tuple[int, int]]:
"""Return an iterator of intervals to be fetched.
Args:
start (int): Start of needed interval
end (int): End of needed interval
left (int): Index of first overlapping downloaded data
right (int): Index after last overlapping downloaded data
"""
lslice, rslice = self._left[left:right], self._right[left:right]
i = start = min([start] + lslice[:1])
end = max([end] + rslice[-1:])
for j, k in zip(lslice, rslice):
if j > i:
yield i, j - 1
i = k + 1
if i <= end:
yield i, end
self._left[left:right], self._right[left:right] = [start], [end]
def _download(self, start: int, end: int) -> None:
"""Download bytes from start to end inclusively."""
with self._stay():
left = bisect_left(self._right, start)
right = bisect_right(self._left, end)
for start, end in self._merge(start, end, left, right):
response = self._stream_response(start, end)
response.raise_for_status()
self.seek(start)
for chunk in response_chunks(response, self._chunk_size):
self._file.write(chunk)
|