You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
313 lines
10 KiB
Python
313 lines
10 KiB
Python
# urllib3/response.py
|
|
# Copyright 2008-2013 Andrey Petrov and contributors (see CONTRIBUTORS.txt)
|
|
#
|
|
# This module is part of urllib3 and is released under
|
|
# the MIT License: http://www.opensource.org/licenses/mit-license.php
|
|
|
|
|
|
import logging
|
|
import zlib
|
|
import io
|
|
|
|
from .exceptions import DecodeError
|
|
from .packages.six import string_types as basestring, binary_type
|
|
from .util import is_fp_closed
|
|
|
|
|
|
log = logging.getLogger(__name__)
|
|
|
|
|
|
class DeflateDecoder(object):
|
|
|
|
def __init__(self):
|
|
self._first_try = True
|
|
self._data = binary_type()
|
|
self._obj = zlib.decompressobj()
|
|
|
|
def __getattr__(self, name):
|
|
return getattr(self._obj, name)
|
|
|
|
def decompress(self, data):
|
|
if not self._first_try:
|
|
return self._obj.decompress(data)
|
|
|
|
self._data += data
|
|
try:
|
|
return self._obj.decompress(data)
|
|
except zlib.error:
|
|
self._first_try = False
|
|
self._obj = zlib.decompressobj(-zlib.MAX_WBITS)
|
|
try:
|
|
return self.decompress(self._data)
|
|
finally:
|
|
self._data = None
|
|
|
|
|
|
def _get_decoder(mode):
|
|
if mode == 'gzip':
|
|
return zlib.decompressobj(16 + zlib.MAX_WBITS)
|
|
|
|
return DeflateDecoder()
|
|
|
|
|
|
class HTTPResponse(io.IOBase):
|
|
"""
|
|
HTTP Response container.
|
|
|
|
Backwards-compatible to httplib's HTTPResponse but the response ``body`` is
|
|
loaded and decoded on-demand when the ``data`` property is accessed.
|
|
|
|
Extra parameters for behaviour not present in httplib.HTTPResponse:
|
|
|
|
:param preload_content:
|
|
If True, the response's body will be preloaded during construction.
|
|
|
|
:param decode_content:
|
|
If True, attempts to decode specific content-encoding's based on headers
|
|
(like 'gzip' and 'deflate') will be skipped and raw data will be used
|
|
instead.
|
|
|
|
:param original_response:
|
|
When this HTTPResponse wrapper is generated from an httplib.HTTPResponse
|
|
object, it's convenient to include the original for debug purposes. It's
|
|
otherwise unused.
|
|
"""
|
|
|
|
CONTENT_DECODERS = ['gzip', 'deflate']
|
|
REDIRECT_STATUSES = [301, 302, 303, 307, 308]
|
|
|
|
def __init__(self, body='', headers=None, status=0, version=0, reason=None,
|
|
strict=0, preload_content=True, decode_content=True,
|
|
original_response=None, pool=None, connection=None):
|
|
self.headers = headers or {}
|
|
self.status = status
|
|
self.version = version
|
|
self.reason = reason
|
|
self.strict = strict
|
|
self.decode_content = decode_content
|
|
|
|
self._decoder = None
|
|
self._body = body if body and isinstance(body, basestring) else None
|
|
self._fp = None
|
|
self._original_response = original_response
|
|
self._fp_bytes_read = 0
|
|
|
|
self._pool = pool
|
|
self._connection = connection
|
|
|
|
if hasattr(body, 'read'):
|
|
self._fp = body
|
|
|
|
if preload_content and not self._body:
|
|
self._body = self.read(decode_content=decode_content)
|
|
|
|
def get_redirect_location(self):
|
|
"""
|
|
Should we redirect and where to?
|
|
|
|
:returns: Truthy redirect location string if we got a redirect status
|
|
code and valid location. ``None`` if redirect status and no
|
|
location. ``False`` if not a redirect status code.
|
|
"""
|
|
if self.status in self.REDIRECT_STATUSES:
|
|
return self.headers.get('location')
|
|
|
|
return False
|
|
|
|
def release_conn(self):
|
|
if not self._pool or not self._connection:
|
|
return
|
|
|
|
self._pool._put_conn(self._connection)
|
|
self._connection = None
|
|
|
|
@property
|
|
def data(self):
|
|
# For backwords-compat with earlier urllib3 0.4 and earlier.
|
|
if self._body:
|
|
return self._body
|
|
|
|
if self._fp:
|
|
return self.read(cache_content=True)
|
|
|
|
def tell(self):
|
|
"""
|
|
Obtain the number of bytes pulled over the wire so far. May differ from
|
|
the amount of content returned by :meth:``HTTPResponse.read`` if bytes
|
|
are encoded on the wire (e.g, compressed).
|
|
"""
|
|
return self._fp_bytes_read
|
|
|
|
def read(self, amt=None, decode_content=None, cache_content=False):
|
|
"""
|
|
Similar to :meth:`httplib.HTTPResponse.read`, but with two additional
|
|
parameters: ``decode_content`` and ``cache_content``.
|
|
|
|
:param amt:
|
|
How much of the content to read. If specified, caching is skipped
|
|
because it doesn't make sense to cache partial content as the full
|
|
response.
|
|
|
|
:param decode_content:
|
|
If True, will attempt to decode the body based on the
|
|
'content-encoding' header.
|
|
|
|
:param cache_content:
|
|
If True, will save the returned data such that the same result is
|
|
returned despite of the state of the underlying file object. This
|
|
is useful if you want the ``.data`` property to continue working
|
|
after having ``.read()`` the file object. (Overridden if ``amt`` is
|
|
set.)
|
|
"""
|
|
# Note: content-encoding value should be case-insensitive, per RFC 2616
|
|
# Section 3.5
|
|
content_encoding = self.headers.get('content-encoding', '').lower()
|
|
if self._decoder is None:
|
|
if content_encoding in self.CONTENT_DECODERS:
|
|
self._decoder = _get_decoder(content_encoding)
|
|
if decode_content is None:
|
|
decode_content = self.decode_content
|
|
|
|
if self._fp is None:
|
|
return
|
|
|
|
flush_decoder = False
|
|
|
|
try:
|
|
if amt is None:
|
|
# cStringIO doesn't like amt=None
|
|
data = self._fp.read()
|
|
flush_decoder = True
|
|
else:
|
|
cache_content = False
|
|
data = self._fp.read(amt)
|
|
if amt != 0 and not data: # Platform-specific: Buggy versions of Python.
|
|
# Close the connection when no data is returned
|
|
#
|
|
# This is redundant to what httplib/http.client _should_
|
|
# already do. However, versions of python released before
|
|
# December 15, 2012 (http://bugs.python.org/issue16298) do not
|
|
# properly close the connection in all cases. There is no harm
|
|
# in redundantly calling close.
|
|
self._fp.close()
|
|
flush_decoder = True
|
|
|
|
self._fp_bytes_read += len(data)
|
|
|
|
try:
|
|
if decode_content and self._decoder:
|
|
data = self._decoder.decompress(data)
|
|
except (IOError, zlib.error) as e:
|
|
raise DecodeError(
|
|
"Received response with content-encoding: %s, but "
|
|
"failed to decode it." % content_encoding,
|
|
e)
|
|
|
|
if flush_decoder and decode_content and self._decoder:
|
|
buf = self._decoder.decompress(binary_type())
|
|
data += buf + self._decoder.flush()
|
|
|
|
if cache_content:
|
|
self._body = data
|
|
|
|
return data
|
|
|
|
finally:
|
|
if self._original_response and self._original_response.isclosed():
|
|
self.release_conn()
|
|
|
|
def stream(self, amt=2**16, decode_content=None):
|
|
"""
|
|
A generator wrapper for the read() method. A call will block until
|
|
``amt`` bytes have been read from the connection or until the
|
|
connection is closed.
|
|
|
|
:param amt:
|
|
How much of the content to read. The generator will return up to
|
|
much data per iteration, but may return less. This is particularly
|
|
likely when using compressed data. However, the empty string will
|
|
never be returned.
|
|
|
|
:param decode_content:
|
|
If True, will attempt to decode the body based on the
|
|
'content-encoding' header.
|
|
"""
|
|
while not is_fp_closed(self._fp):
|
|
data = self.read(amt=amt, decode_content=decode_content)
|
|
|
|
if data:
|
|
yield data
|
|
|
|
|
|
@classmethod
|
|
def from_httplib(ResponseCls, r, **response_kw):
|
|
"""
|
|
Given an :class:`httplib.HTTPResponse` instance ``r``, return a
|
|
corresponding :class:`urllib3.response.HTTPResponse` object.
|
|
|
|
Remaining parameters are passed to the HTTPResponse constructor, along
|
|
with ``original_response=r``.
|
|
"""
|
|
|
|
# Normalize headers between different versions of Python
|
|
headers = {}
|
|
for k, v in r.getheaders():
|
|
# Python 3: Header keys are returned capitalised
|
|
k = k.lower()
|
|
|
|
has_value = headers.get(k)
|
|
if has_value: # Python 3: Repeating header keys are unmerged.
|
|
v = ', '.join([has_value, v])
|
|
|
|
headers[k] = v
|
|
|
|
# HTTPResponse objects in Python 3 don't have a .strict attribute
|
|
strict = getattr(r, 'strict', 0)
|
|
return ResponseCls(body=r,
|
|
headers=headers,
|
|
status=r.status,
|
|
version=r.version,
|
|
reason=r.reason,
|
|
strict=strict,
|
|
original_response=r,
|
|
**response_kw)
|
|
|
|
# Backwards-compatibility methods for httplib.HTTPResponse
|
|
def getheaders(self):
|
|
return self.headers
|
|
|
|
def getheader(self, name, default=None):
|
|
return self.headers.get(name, default)
|
|
|
|
# Overrides from io.IOBase
|
|
def close(self):
|
|
if not self.closed:
|
|
self._fp.close()
|
|
|
|
@property
|
|
def closed(self):
|
|
if self._fp is None:
|
|
return True
|
|
elif hasattr(self._fp, 'closed'):
|
|
return self._fp.closed
|
|
elif hasattr(self._fp, 'isclosed'): # Python 2
|
|
return self._fp.isclosed()
|
|
else:
|
|
return True
|
|
|
|
def fileno(self):
|
|
if self._fp is None:
|
|
raise IOError("HTTPResponse has no file to get a fileno from")
|
|
elif hasattr(self._fp, "fileno"):
|
|
return self._fp.fileno()
|
|
else:
|
|
raise IOError("The file-like object this HTTPResponse is wrapped "
|
|
"around has no file descriptor")
|
|
|
|
def flush(self):
|
|
if self._fp is not None and hasattr(self._fp, 'flush'):
|
|
return self._fp.flush()
|
|
|
|
def readable(self):
|
|
return True
|