aboutsummaryrefslogtreecommitdiff
path: root/urllib3
diff options
context:
space:
mode:
Diffstat (limited to 'urllib3')
-rw-r--r--urllib3/__init__.py26
-rw-r--r--urllib3/_collections.py6
-rw-r--r--urllib3/connection.py51
-rw-r--r--urllib3/connectionpool.py182
-rw-r--r--urllib3/contrib/ntlmpool.py6
-rw-r--r--urllib3/contrib/pyopenssl.py202
-rw-r--r--urllib3/exceptions.py37
-rw-r--r--urllib3/fields.py8
-rw-r--r--urllib3/filepost.py6
-rw-r--r--urllib3/packages/ordered_dict.py1
-rw-r--r--urllib3/poolmanager.py41
-rw-r--r--urllib3/request.py8
-rw-r--r--urllib3/response.py31
-rw-r--r--urllib3/util/__init__.py9
-rw-r--r--urllib3/util/connection.py56
-rw-r--r--urllib3/util/request.py13
-rw-r--r--urllib3/util/retry.py279
-rw-r--r--urllib3/util/ssl_.py5
-rw-r--r--urllib3/util/timeout.py67
-rw-r--r--urllib3/util/url.py9
20 files changed, 640 insertions, 403 deletions
diff --git a/urllib3/__init__.py b/urllib3/__init__.py
index c80d5da..56f5bf4 100644
--- a/urllib3/__init__.py
+++ b/urllib3/__init__.py
@@ -1,16 +1,10 @@
-# urllib3/__init__.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
-
"""
urllib3 - Thread-safe connection pooling and re-using.
"""
__author__ = 'Andrey Petrov (andrey.petrov@shazow.net)'
__license__ = 'MIT'
-__version__ = '1.8.3'
+__version__ = '1.9'
from .connectionpool import (
@@ -23,7 +17,10 @@ from . import exceptions
from .filepost import encode_multipart_formdata
from .poolmanager import PoolManager, ProxyManager, proxy_from_url
from .response import HTTPResponse
-from .util import make_headers, get_host, Timeout
+from .util.request import make_headers
+from .util.url import get_host
+from .util.timeout import Timeout
+from .util.retry import Retry
# Set default logging handler to avoid "No handler found" warnings.
@@ -51,8 +48,19 @@ def add_stderr_logger(level=logging.DEBUG):
handler.setFormatter(logging.Formatter('%(asctime)s %(levelname)s %(message)s'))
logger.addHandler(handler)
logger.setLevel(level)
- logger.debug('Added an stderr logging handler to logger: %s' % __name__)
+ logger.debug('Added a stderr logging handler to logger: %s' % __name__)
return handler
# ... Clean up.
del NullHandler
+
+
+# Set security warning to only go off once by default.
+import warnings
+warnings.simplefilter('module', exceptions.InsecureRequestWarning)
+
+def disable_warnings(category=exceptions.HTTPWarning):
+ """
+ Helper for quickly disabling all urllib3 warnings.
+ """
+ warnings.simplefilter('ignore', category)
diff --git a/urllib3/_collections.py b/urllib3/_collections.py
index ccf0d5f..d77ebb8 100644
--- a/urllib3/_collections.py
+++ b/urllib3/_collections.py
@@ -1,9 +1,3 @@
-# urllib3/_collections.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
-
from collections import Mapping, MutableMapping
try:
from threading import RLock
diff --git a/urllib3/connection.py b/urllib3/connection.py
index fbb63ed..0d578d7 100644
--- a/urllib3/connection.py
+++ b/urllib3/connection.py
@@ -1,9 +1,3 @@
-# urllib3/connection.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 sys
import socket
from socket import timeout as SocketTimeout
@@ -35,13 +29,16 @@ from .exceptions import (
)
from .packages.ssl_match_hostname import match_hostname
from .packages import six
-from .util import (
- assert_fingerprint,
+
+from .util.ssl_ import (
resolve_cert_reqs,
resolve_ssl_version,
ssl_wrap_socket,
+ assert_fingerprint,
)
+from .util import connection
+
port_by_scheme = {
'http': 80,
@@ -82,15 +79,23 @@ class HTTPConnection(_HTTPConnection, object):
#: ``[(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1)]``
default_socket_options = [(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1)]
+ #: Whether this connection verifies the host's certificate.
+ is_verified = False
+
def __init__(self, *args, **kw):
if six.PY3: # Python 3
kw.pop('strict', None)
- if sys.version_info < (2, 7): # Python 2.6 and older
- kw.pop('source_address', None)
# Pre-set source_address in case we have an older Python like 2.6.
self.source_address = kw.get('source_address')
+ if sys.version_info < (2, 7): # Python 2.6
+ # _HTTPConnection on Python 2.6 will balk at this keyword arg, but
+ # not newer versions. We can still use it when creating a
+ # connection though, so we pop it *after* we have saved it as
+ # self.source_address.
+ kw.pop('source_address', None)
+
#: The socket options provided by the user. If no options are
#: provided, we use the default options.
self.socket_options = kw.pop('socket_options', self.default_socket_options)
@@ -103,22 +108,22 @@ class HTTPConnection(_HTTPConnection, object):
:return: New socket connection.
"""
- extra_args = []
- if self.source_address: # Python 2.7+
- extra_args.append(self.source_address)
+ extra_kw = {}
+ if self.source_address:
+ extra_kw['source_address'] = self.source_address
+
+ if self.socket_options:
+ extra_kw['socket_options'] = self.socket_options
try:
- conn = socket.create_connection(
- (self.host, self.port), self.timeout, *extra_args)
+ conn = connection.create_connection(
+ (self.host, self.port), self.timeout, **extra_kw)
except SocketTimeout:
raise ConnectTimeoutError(
self, "Connection to %s timed out. (connect timeout=%s)" %
(self.host, self.timeout))
- # Set options on the socket.
- self._set_options_on(conn)
-
return conn
def _prepare_conn(self, conn):
@@ -132,14 +137,6 @@ class HTTPConnection(_HTTPConnection, object):
# Mark this connection as not reusable
self.auto_open = 0
- def _set_options_on(self, conn):
- # Disable all socket options if the user passes ``socket_options=None``
- if self.socket_options is None:
- return
-
- for opt in self.socket_options:
- conn.setsockopt(*opt)
-
def connect(self):
conn = self._new_conn()
self._prepare_conn(conn)
@@ -225,6 +222,8 @@ class VerifiedHTTPSConnection(HTTPSConnection):
match_hostname(self.sock.getpeercert(),
self.assert_hostname or hostname)
+ self.is_verified = resolved_cert_reqs == ssl.CERT_REQUIRED
+
if ssl:
# Make a copy for testing.
diff --git a/urllib3/connectionpool.py b/urllib3/connectionpool.py
index ab205fa..9317fdc 100644
--- a/urllib3/connectionpool.py
+++ b/urllib3/connectionpool.py
@@ -1,12 +1,7 @@
-# urllib3/connectionpool.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 sys
import errno
import logging
+import sys
+import warnings
from socket import error as SocketError, timeout as SocketTimeout
import socket
@@ -20,15 +15,16 @@ except ImportError:
from .exceptions import (
ClosedPoolError,
- ConnectionError,
+ ProtocolError,
EmptyPoolError,
HostChangedError,
- LocationParseError,
+ LocationValueError,
MaxRetryError,
+ ProxyError,
+ ReadTimeoutError,
SSLError,
TimeoutError,
- ReadTimeoutError,
- ProxyError,
+ InsecureRequestWarning,
)
from .packages.ssl_match_hostname import CertificateError
from .packages import six
@@ -40,11 +36,11 @@ from .connection import (
)
from .request import RequestMethods
from .response import HTTPResponse
-from .util import (
- get_host,
- is_connection_dropped,
- Timeout,
-)
+
+from .util.connection import is_connection_dropped
+from .util.retry import Retry
+from .util.timeout import Timeout
+from .util.url import get_host
xrange = six.moves.xrange
@@ -65,13 +61,11 @@ class ConnectionPool(object):
QueueCls = LifoQueue
def __init__(self, host, port=None):
- if host is None:
- raise LocationParseError(host)
+ if not host:
+ raise LocationValueError("No host specified.")
# httplib doesn't like it when we include brackets in ipv6 addresses
- host = host.strip('[]')
-
- self.host = host
+ self.host = host.strip('[]')
self.port = port
def __str__(self):
@@ -126,6 +120,9 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
Headers to include with all requests, unless other headers are given
explicitly.
+ :param retries:
+ Retry configuration to use by default with requests in this pool.
+
:param _proxy:
Parsed proxy URL, should not be used directly, instead, see
:class:`urllib3.connectionpool.ProxyManager`"
@@ -144,18 +141,22 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
def __init__(self, host, port=None, strict=False,
timeout=Timeout.DEFAULT_TIMEOUT, maxsize=1, block=False,
- headers=None, _proxy=None, _proxy_headers=None, **conn_kw):
+ headers=None, retries=None,
+ _proxy=None, _proxy_headers=None,
+ **conn_kw):
ConnectionPool.__init__(self, host, port)
RequestMethods.__init__(self, headers)
self.strict = strict
- # This is for backwards compatibility and can be removed once a timeout
- # can only be set to a Timeout object
if not isinstance(timeout, Timeout):
timeout = Timeout.from_float(timeout)
+ if retries is None:
+ retries = Retry.DEFAULT
+
self.timeout = timeout
+ self.retries = retries
self.pool = self.QueueCls(maxsize)
self.block = block
@@ -259,6 +260,12 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
if conn:
conn.close()
+ def _validate_conn(self, conn):
+ """
+ Called right before a request is made, after the socket is created.
+ """
+ pass
+
def _get_timeout(self, timeout):
""" Helper that always returns a :class:`urllib3.util.Timeout` """
if timeout is _Default:
@@ -290,9 +297,12 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
self.num_requests += 1
timeout_obj = self._get_timeout(timeout)
-
timeout_obj.start_connect()
conn.timeout = timeout_obj.connect_timeout
+
+ # Trigger any extra validation we need to do.
+ self._validate_conn(conn)
+
# conn.request() calls httplib.*.request, not the method in
# urllib3.request. It also calls makefile (recv) on the socket.
conn.request(method, url, **httplib_request_kw)
@@ -301,7 +311,7 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
read_timeout = timeout_obj.read_timeout
# App Engine doesn't have a sock attr
- if hasattr(conn, 'sock'):
+ if getattr(conn, 'sock', None):
# In Python 3 socket.py will catch EAGAIN and return None when you
# try and read into the file pointer created by http.client, which
# instead raises a BadStatusLine exception. Instead of catching
@@ -309,8 +319,7 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
# timeouts, check for a zero timeout before making the request.
if read_timeout == 0:
raise ReadTimeoutError(
- self, url,
- "Read timed out. (read timeout=%s)" % read_timeout)
+ self, url, "Read timed out. (read timeout=%s)" % read_timeout)
if read_timeout is Timeout.DEFAULT_TIMEOUT:
conn.sock.settimeout(socket.getdefaulttimeout())
else: # None or a value
@@ -332,7 +341,8 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
# http://bugs.python.org/issue10272
if 'timed out' in str(e) or \
'did not complete (read)' in str(e): # Python 2.6
- raise ReadTimeoutError(self, url, "Read timed out.")
+ raise ReadTimeoutError(
+ self, url, "Read timed out. (read timeout=%s)" % read_timeout)
raise
@@ -341,8 +351,7 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
# have to specifically catch it and throw the timeout error
if e.errno in _blocking_errnos:
raise ReadTimeoutError(
- self, url,
- "Read timed out. (read timeout=%s)" % read_timeout)
+ self, url, "Read timed out. (read timeout=%s)" % read_timeout)
raise
@@ -388,7 +397,7 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
return (scheme, host, port) == (self.scheme, self.host, self.port)
- def urlopen(self, method, url, body=None, headers=None, retries=3,
+ def urlopen(self, method, url, body=None, headers=None, retries=None,
redirect=True, assert_same_host=True, timeout=_Default,
pool_timeout=None, release_conn=None, **response_kw):
"""
@@ -422,9 +431,20 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
these headers completely replace any pool-specific headers.
:param retries:
- Number of retries to allow before raising a MaxRetryError exception.
- If `False`, then retries are disabled and any exception is raised
- immediately.
+ Configure the number of retries to allow before raising a
+ :class:`~urllib3.exceptions.MaxRetryError` exception.
+
+ Pass ``None`` to retry until you receive a response. Pass a
+ :class:`~urllib3.util.retry.Retry` object for fine-grained control
+ over different types of retries.
+ Pass an integer number to retry connection errors that many times,
+ but no other types of errors. Pass zero to never retry.
+
+ If ``False``, then retries are disabled and any exception is raised
+ immediately. Also, instead of raising a MaxRetryError on redirects,
+ the redirect response will be returned.
+
+ :type retries: :class:`~urllib3.util.retry.Retry`, False, or an int.
:param redirect:
If True, automatically handle redirects (status codes 301, 302,
@@ -463,15 +483,15 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
if headers is None:
headers = self.headers
- if retries < 0 and retries is not False:
- raise MaxRetryError(self, url)
+ if not isinstance(retries, Retry):
+ retries = Retry.from_int(retries, redirect=redirect, default=self.retries)
if release_conn is None:
release_conn = response_kw.get('preload_content', True)
# Check host
if assert_same_host and not self.is_same_host(url):
- raise HostChangedError(self, url, retries - 1)
+ raise HostChangedError(self, url, retries)
conn = None
@@ -487,10 +507,10 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
err = None
try:
- # Request a connection from the queue
+ # Request a connection from the queue.
conn = self._get_conn(timeout=pool_timeout)
- # Make the request on the httplib connection object
+ # Make the request on the httplib connection object.
httplib_response = self._make_request(conn, method, url,
timeout=timeout,
body=body, headers=headers)
@@ -529,21 +549,15 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
conn.close()
conn = None
- if not retries:
- if isinstance(e, TimeoutError):
- # TimeoutError is exempt from MaxRetryError-wrapping.
- # FIXME: ... Not sure why. Add a reason here.
- raise
-
- # Wrap unexpected exceptions with the most appropriate
- # module-level exception and re-raise.
- if isinstance(e, SocketError) and self.proxy:
- raise ProxyError('Cannot connect to proxy.', e)
-
- if retries is False:
- raise ConnectionError('Connection failed.', e)
+ stacktrace = sys.exc_info()[2]
+ if isinstance(e, SocketError) and self.proxy:
+ e = ProxyError('Cannot connect to proxy.', e)
+ elif isinstance(e, (SocketError, HTTPException)):
+ e = ProtocolError('Connection aborted.', e)
- raise MaxRetryError(self, url, e)
+ retries = retries.increment(method, url, error=e,
+ _pool=self, _stacktrace=stacktrace)
+ retries.sleep()
# Keep track of the error for the retry warning.
err = e
@@ -557,23 +571,43 @@ class HTTPConnectionPool(ConnectionPool, RequestMethods):
if not conn:
# Try again
- log.warning("Retrying (%d attempts remain) after connection "
+ log.warning("Retrying (%r) after connection "
"broken by '%r': %s" % (retries, err, url))
- return self.urlopen(method, url, body, headers, retries - 1,
+ return self.urlopen(method, url, body, headers, retries,
redirect, assert_same_host,
timeout=timeout, pool_timeout=pool_timeout,
release_conn=release_conn, **response_kw)
# Handle redirect?
redirect_location = redirect and response.get_redirect_location()
- if redirect_location and retries is not False:
+ if redirect_location:
if response.status == 303:
method = 'GET'
+
+ try:
+ retries = retries.increment(method, url, response=response, _pool=self)
+ except MaxRetryError:
+ if retries.raise_on_redirect:
+ raise
+ return response
+
log.info("Redirecting %s -> %s" % (url, redirect_location))
return self.urlopen(method, redirect_location, body, headers,
- retries - 1, redirect, assert_same_host,
- timeout=timeout, pool_timeout=pool_timeout,
- release_conn=release_conn, **response_kw)
+ retries=retries, redirect=redirect,
+ assert_same_host=assert_same_host,
+ timeout=timeout, pool_timeout=pool_timeout,
+ release_conn=release_conn, **response_kw)
+
+ # Check if we should retry the HTTP response.
+ if retries.is_forced_retry(method, status_code=response.status):
+ retries = retries.increment(method, url, response=response, _pool=self)
+ retries.sleep()
+ log.info("Forced retry: %s" % url)
+ return self.urlopen(method, url, body, headers,
+ retries=retries, redirect=redirect,
+ assert_same_host=assert_same_host,
+ timeout=timeout, pool_timeout=pool_timeout,
+ release_conn=release_conn, **response_kw)
return response
@@ -600,8 +634,8 @@ class HTTPSConnectionPool(HTTPConnectionPool):
ConnectionCls = HTTPSConnection
def __init__(self, host, port=None,
- strict=False, timeout=None, maxsize=1,
- block=False, headers=None,
+ strict=False, timeout=Timeout.DEFAULT_TIMEOUT, maxsize=1,
+ block=False, headers=None, retries=None,
_proxy=None, _proxy_headers=None,
key_file=None, cert_file=None, cert_reqs=None,
ca_certs=None, ssl_version=None,
@@ -609,7 +643,7 @@ class HTTPSConnectionPool(HTTPConnectionPool):
**conn_kw):
HTTPConnectionPool.__init__(self, host, port, strict, timeout, maxsize,
- block, headers, _proxy, _proxy_headers,
+ block, headers, retries, _proxy, _proxy_headers,
**conn_kw)
self.key_file = key_file
self.cert_file = cert_file
@@ -640,7 +674,7 @@ class HTTPSConnectionPool(HTTPConnectionPool):
set_tunnel = conn.set_tunnel
except AttributeError: # Platform-specific: Python 2.6
set_tunnel = conn._set_tunnel
-
+
if sys.version_info <= (2, 6, 4) and not self.proxy_headers: # Python 2.6.4 and older
set_tunnel(self.host, self.port)
else:
@@ -677,6 +711,24 @@ class HTTPSConnectionPool(HTTPConnectionPool):
return self._prepare_conn(conn)
+ def _validate_conn(self, conn):
+ """
+ Called right before a request is made, after the socket is created.
+ """
+ super(HTTPSConnectionPool, self)._validate_conn(conn)
+
+ # Force connect early to allow us to validate the connection.
+ if not conn.sock:
+ conn.connect()
+
+ if not conn.is_verified:
+ warnings.warn((
+ 'Unverified HTTPS request is being made. '
+ 'Adding certificate verification is strongly advised. See: '
+ 'https://urllib3.readthedocs.org/en/latest/security.html '
+ '(This warning will only appear once by default.)'),
+ InsecureRequestWarning)
+
def connection_from_url(url, **kw):
"""
@@ -693,7 +745,7 @@ def connection_from_url(url, **kw):
:class:`.ConnectionPool`. Useful for specifying things like
timeout, maxsize, headers, etc.
- Example: ::
+ Example::
>>> conn = connection_from_url('http://google.com/')
>>> r = conn.request('GET', '/')
diff --git a/urllib3/contrib/ntlmpool.py b/urllib3/contrib/ntlmpool.py
index b8cd933..c6b266f 100644
--- a/urllib3/contrib/ntlmpool.py
+++ b/urllib3/contrib/ntlmpool.py
@@ -1,9 +1,3 @@
-# urllib3/contrib/ntlmpool.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
-
"""
NTLM authenticating pool, contributed by erikcederstran
diff --git a/urllib3/contrib/pyopenssl.py b/urllib3/contrib/pyopenssl.py
index 21a12c6..7a9ea2e 100644
--- a/urllib3/contrib/pyopenssl.py
+++ b/urllib3/contrib/pyopenssl.py
@@ -54,7 +54,6 @@ from pyasn1.type import univ, constraint
from socket import _fileobject, timeout
import ssl
import select
-from cStringIO import StringIO
from .. import connection
from .. import util
@@ -155,196 +154,37 @@ def get_subj_alt_name(peer_cert):
return dns_name
-class fileobject(_fileobject):
-
- def _wait_for_sock(self):
- rd, wd, ed = select.select([self._sock], [], [],
- self._sock.gettimeout())
- if not rd:
- raise timeout()
-
-
- def read(self, size=-1):
- # Use max, disallow tiny reads in a loop as they are very inefficient.
- # We never leave read() with any leftover data from a new recv() call
- # in our internal buffer.
- rbufsize = max(self._rbufsize, self.default_bufsize)
- # Our use of StringIO rather than lists of string objects returned by
- # recv() minimizes memory usage and fragmentation that occurs when
- # rbufsize is large compared to the typical return value of recv().
- buf = self._rbuf
- buf.seek(0, 2) # seek end
- if size < 0:
- # Read until EOF
- self._rbuf = StringIO() # reset _rbuf. we consume it via buf.
- while True:
- try:
- data = self._sock.recv(rbufsize)
- except OpenSSL.SSL.WantReadError:
- self._wait_for_sock()
- continue
- if not data:
- break
- buf.write(data)
- return buf.getvalue()
- else:
- # Read until size bytes or EOF seen, whichever comes first
- buf_len = buf.tell()
- if buf_len >= size:
- # Already have size bytes in our buffer? Extract and return.
- buf.seek(0)
- rv = buf.read(size)
- self._rbuf = StringIO()
- self._rbuf.write(buf.read())
- return rv
-
- self._rbuf = StringIO() # reset _rbuf. we consume it via buf.
- while True:
- left = size - buf_len
- # recv() will malloc the amount of memory given as its
- # parameter even though it often returns much less data
- # than that. The returned data string is short lived
- # as we copy it into a StringIO and free it. This avoids
- # fragmentation issues on many platforms.
- try:
- data = self._sock.recv(left)
- except OpenSSL.SSL.WantReadError:
- self._wait_for_sock()
- continue
- if not data:
- break
- n = len(data)
- if n == size and not buf_len:
- # Shortcut. Avoid buffer data copies when:
- # - We have no data in our buffer.
- # AND
- # - Our call to recv returned exactly the
- # number of bytes we were asked to read.
- return data
- if n == left:
- buf.write(data)
- del data # explicit free
- break
- assert n <= left, "recv(%d) returned %d bytes" % (left, n)
- buf.write(data)
- buf_len += n
- del data # explicit free
- #assert buf_len == buf.tell()
- return buf.getvalue()
-
- def readline(self, size=-1):
- buf = self._rbuf
- buf.seek(0, 2) # seek end
- if buf.tell() > 0:
- # check if we already have it in our buffer
- buf.seek(0)
- bline = buf.readline(size)
- if bline.endswith('\n') or len(bline) == size:
- self._rbuf = StringIO()
- self._rbuf.write(buf.read())
- return bline
- del bline
- if size < 0:
- # Read until \n or EOF, whichever comes first
- if self._rbufsize <= 1:
- # Speed up unbuffered case
- buf.seek(0)
- buffers = [buf.read()]
- self._rbuf = StringIO() # reset _rbuf. we consume it via buf.
- data = None
- recv = self._sock.recv
- while True:
- try:
- while data != "\n":
- data = recv(1)
- if not data:
- break
- buffers.append(data)
- except OpenSSL.SSL.WantReadError:
- self._wait_for_sock()
- continue
- break
- return "".join(buffers)
-
- buf.seek(0, 2) # seek end
- self._rbuf = StringIO() # reset _rbuf. we consume it via buf.
- while True:
- try:
- data = self._sock.recv(self._rbufsize)
- except OpenSSL.SSL.WantReadError:
- self._wait_for_sock()
- continue
- if not data:
- break
- nl = data.find('\n')
- if nl >= 0:
- nl += 1
- buf.write(data[:nl])
- self._rbuf.write(data[nl:])
- del data
- break
- buf.write(data)
- return buf.getvalue()
- else:
- # Read until size bytes or \n or EOF seen, whichever comes first
- buf.seek(0, 2) # seek end
- buf_len = buf.tell()
- if buf_len >= size:
- buf.seek(0)
- rv = buf.read(size)
- self._rbuf = StringIO()
- self._rbuf.write(buf.read())
- return rv
- self._rbuf = StringIO() # reset _rbuf. we consume it via buf.
- while True:
- try:
- data = self._sock.recv(self._rbufsize)
- except OpenSSL.SSL.WantReadError:
- self._wait_for_sock()
- continue
- if not data:
- break
- left = size - buf_len
- # did we just receive a newline?
- nl = data.find('\n', 0, left)
- if nl >= 0:
- nl += 1
- # save the excess data to _rbuf
- self._rbuf.write(data[nl:])
- if buf_len:
- buf.write(data[:nl])
- break
- else:
- # Shortcut. Avoid data copy through buf when returning
- # a substring of our first recv().
- return data[:nl]
- n = len(data)
- if n == size and not buf_len:
- # Shortcut. Avoid data copy through buf when
- # returning exactly all of our first recv().
- return data
- if n >= left:
- buf.write(data[:left])
- self._rbuf.write(data[left:])
- break
- buf.write(data)
- buf_len += n
- #assert buf_len == buf.tell()
- return buf.getvalue()
-
-
class WrappedSocket(object):
'''API-compatibility wrapper for Python OpenSSL's Connection-class.'''
- def __init__(self, connection, socket):
+ def __init__(self, connection, socket, suppress_ragged_eofs=True):
self.connection = connection
self.socket = socket
+ self.suppress_ragged_eofs = suppress_ragged_eofs
def fileno(self):
return self.socket.fileno()
def makefile(self, mode, bufsize=-1):
- return fileobject(self.connection, mode, bufsize)
+ return _fileobject(self, mode, bufsize)
+
+ def recv(self, *args, **kwargs):
+ try:
+ data = self.connection.recv(*args, **kwargs)
+ except OpenSSL.SSL.SysCallError as e:
+ if self.suppress_ragged_eofs and e.args == (-1, 'Unexpected EOF'):
+ return b''
+ else:
+ raise
+ except OpenSSL.SSL.WantReadError:
+ rd, wd, ed = select.select(
+ [self.socket], [], [], self.socket.gettimeout())
+ if not rd:
+ raise timeout()
+ else:
+ return self.recv(*args, **kwargs)
+ else:
+ return data
def settimeout(self, timeout):
return self.socket.settimeout(timeout)
diff --git a/urllib3/exceptions.py b/urllib3/exceptions.py
index b4df831..fff8bfa 100644
--- a/urllib3/exceptions.py
+++ b/urllib3/exceptions.py
@@ -1,9 +1,3 @@
-# urllib3/exceptions.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
-
## Base Exceptions
@@ -11,6 +5,11 @@ class HTTPError(Exception):
"Base exception used by this module."
pass
+class HTTPWarning(Warning):
+ "Base warning used by this module."
+ pass
+
+
class PoolError(HTTPError):
"Base exception for errors caused within a pool."
@@ -44,16 +43,20 @@ class ProxyError(HTTPError):
pass
-class ConnectionError(HTTPError):
- "Raised when a normal connection fails."
+class DecodeError(HTTPError):
+ "Raised when automatic decoding based on Content-Type fails."
pass
-class DecodeError(HTTPError):
- "Raised when automatic decoding based on Content-Type fails."
+class ProtocolError(HTTPError):
+ "Raised when something unexpected happens mid-request/response."
pass
+#: Renamed to ProtocolError but aliased for backwards compatibility.
+ConnectionError = ProtocolError
+
+
## Leaf Exceptions
class MaxRetryError(RequestError):
@@ -64,7 +67,7 @@ class MaxRetryError(RequestError):
message = "Max retries exceeded with url: %s" % url
if reason:
- message += " (Caused by %s: %s)" % (type(reason), reason)
+ message += " (Caused by %r)" % reason
else:
message += " (Caused by redirect)"
@@ -116,7 +119,12 @@ class ClosedPoolError(PoolError):
pass
-class LocationParseError(ValueError, HTTPError):
+class LocationValueError(ValueError, HTTPError):
+ "Raised when there is something wrong with a given URL input."
+ pass
+
+
+class LocationParseError(LocationValueError):
"Raised when get_host or similar fails to parse the URL input."
def __init__(self, location):
@@ -124,3 +132,8 @@ class LocationParseError(ValueError, HTTPError):
HTTPError.__init__(self, message)
self.location = location
+
+
+class InsecureRequestWarning(HTTPWarning):
+ "Warned when making an unverified HTTPS request."
+ pass
diff --git a/urllib3/fields.py b/urllib3/fields.py
index dceafb4..c853f8d 100644
--- a/urllib3/fields.py
+++ b/urllib3/fields.py
@@ -1,9 +1,3 @@
-# urllib3/fields.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 email.utils
import mimetypes
@@ -81,7 +75,7 @@ class RequestField(object):
Supports constructing :class:`~urllib3.fields.RequestField` from
parameter of key/value strings AND key/filetuple. A filetuple is a
(filename, data, MIME type) tuple where the MIME type is optional.
- For example: ::
+ For example::
'foo': 'bar',
'fakefile': ('foofile.txt', 'contents of foofile'),
diff --git a/urllib3/filepost.py b/urllib3/filepost.py
index c3db30c..0fbf488 100644
--- a/urllib3/filepost.py
+++ b/urllib3/filepost.py
@@ -1,9 +1,3 @@
-# urllib3/filepost.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 codecs
from uuid import uuid4
diff --git a/urllib3/packages/ordered_dict.py b/urllib3/packages/ordered_dict.py
index 7f8ee15..4479363 100644
--- a/urllib3/packages/ordered_dict.py
+++ b/urllib3/packages/ordered_dict.py
@@ -2,7 +2,6 @@
# Passes Python2.7's test suite and incorporates all the latest updates.
# Copyright 2009 Raymond Hettinger, released under the MIT License.
# http://code.activestate.com/recipes/576693/
-
try:
from thread import get_ident as _get_ident
except ImportError:
diff --git a/urllib3/poolmanager.py b/urllib3/poolmanager.py
index 3945f5d..515dc96 100644
--- a/urllib3/poolmanager.py
+++ b/urllib3/poolmanager.py
@@ -1,9 +1,3 @@
-# urllib3/poolmanager.py
-# Copyright 2008-2014 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
try: # Python 3
@@ -14,8 +8,10 @@ except ImportError:
from ._collections import RecentlyUsedContainer
from .connectionpool import HTTPConnectionPool, HTTPSConnectionPool
from .connectionpool import port_by_scheme
+from .exceptions import LocationValueError
from .request import RequestMethods
-from .util import parse_url
+from .util.url import parse_url
+from .util.retry import Retry
__all__ = ['PoolManager', 'ProxyManager', 'proxy_from_url']
@@ -49,7 +45,7 @@ class PoolManager(RequestMethods):
Additional parameters are used to create fresh
:class:`urllib3.connectionpool.ConnectionPool` instances.
- Example: ::
+ Example::
>>> manager = PoolManager(num_pools=2)
>>> r = manager.request('GET', 'http://google.com/')
@@ -102,10 +98,11 @@ class PoolManager(RequestMethods):
``urllib3.connectionpool.port_by_scheme``.
"""
- scheme = scheme or 'http'
+ if not host:
+ raise LocationValueError("No host specified.")
+ scheme = scheme or 'http'
port = port or port_by_scheme.get(scheme, 80)
-
pool_key = (scheme, host, port)
with self.pools.lock:
@@ -118,6 +115,7 @@ class PoolManager(RequestMethods):
# Make a fresh ConnectionPool of the desired type
pool = self._new_pool(scheme, host, port)
self.pools[pool_key] = pool
+
return pool
def connection_from_url(self, url):
@@ -165,9 +163,14 @@ class PoolManager(RequestMethods):
if response.status == 303:
method = 'GET'
- log.info("Redirecting %s -> %s" % (url, redirect_location))
- kw['retries'] = kw.get('retries', 3) - 1 # Persist retries countdown
+ retries = kw.get('retries')
+ if not isinstance(retries, Retry):
+ retries = Retry.from_int(retries, redirect=redirect)
+
+ kw['retries'] = retries.increment(method, redirect_location)
kw['redirect'] = redirect
+
+ log.info("Redirecting %s -> %s" % (url, redirect_location))
return self.urlopen(method, redirect_location, **kw)
@@ -208,12 +211,16 @@ class ProxyManager(PoolManager):
if not proxy.port:
port = port_by_scheme.get(proxy.scheme, 80)
proxy = proxy._replace(port=port)
+
+ assert proxy.scheme in ("http", "https"), \
+ 'Not supported proxy scheme %s' % proxy.scheme
+
self.proxy = proxy
self.proxy_headers = proxy_headers or {}
- assert self.proxy.scheme in ("http", "https"), \
- 'Not supported proxy scheme %s' % self.proxy.scheme
+
connection_pool_kw['_proxy'] = self.proxy
connection_pool_kw['_proxy_headers'] = self.proxy_headers
+
super(ProxyManager, self).__init__(
num_pools, headers, **connection_pool_kw)
@@ -248,10 +255,10 @@ class ProxyManager(PoolManager):
# For proxied HTTPS requests, httplib sets the necessary headers
# on the CONNECT to the proxy. For HTTP, we'll definitely
# need to set 'Host' at the very least.
- kw['headers'] = self._set_proxy_headers(url, kw.get('headers',
- self.headers))
+ headers = kw.get('headers', self.headers)
+ kw['headers'] = self._set_proxy_headers(url, headers)
- return super(ProxyManager, self).urlopen(method, url, redirect, **kw)
+ return super(ProxyManager, self).urlopen(method, url, redirect=redirect, **kw)
def proxy_from_url(url, **kw):
diff --git a/urllib3/request.py b/urllib3/request.py
index 7a46f1b..51fe238 100644
--- a/urllib3/request.py
+++ b/urllib3/request.py
@@ -1,9 +1,3 @@
-# urllib3/request.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
-
try:
from urllib.parse import urlencode
except ImportError:
@@ -105,7 +99,7 @@ class RequestMethods(object):
Supports an optional ``fields`` parameter of key/value strings AND
key/filetuple. A filetuple is a (filename, data, MIME type) tuple where
- the MIME type is optional. For example: ::
+ the MIME type is optional. For example::
fields = {
'foo': 'bar',
diff --git a/urllib3/response.py b/urllib3/response.py
index 13ffba4..7e0d47f 100644
--- a/urllib3/response.py
+++ b/urllib3/response.py
@@ -1,18 +1,13 @@
-# 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 zlib
import io
from socket import timeout as SocketTimeout
from ._collections import HTTPHeaderDict
-from .exceptions import DecodeError, ReadTimeoutError
+from .exceptions import ProtocolError, DecodeError, ReadTimeoutError
from .packages.six import string_types as basestring, binary_type
-from .util import is_fp_closed
+from .connection import HTTPException, BaseSSLError
+from .util.response import is_fp_closed
+
class DeflateDecoder(object):
@@ -88,11 +83,14 @@ class HTTPResponse(io.IOBase):
self.decode_content = decode_content
self._decoder = None
- self._body = body if body and isinstance(body, basestring) else None
+ self._body = None
self._fp = None
self._original_response = original_response
self._fp_bytes_read = 0
+ if body and isinstance(body, (basestring, binary_type)):
+ self._body = body
+
self._pool = pool
self._connection = connection
@@ -199,6 +197,19 @@ class HTTPResponse(io.IOBase):
# there is yet no clean way to get at it from this context.
raise ReadTimeoutError(self._pool, None, 'Read timed out.')
+ except BaseSSLError as e:
+ # FIXME: Is there a better way to differentiate between SSLErrors?
+ if not 'read operation timed out' in str(e): # Defensive:
+ # This shouldn't happen but just in case we're missing an edge
+ # case, let's avoid swallowing SSL errors.
+ raise
+
+ raise ReadTimeoutError(self._pool, None, 'Read timed out.')
+
+ except HTTPException as e:
+ # This includes IncompleteRead.
+ raise ProtocolError('Connection broken: %r' % e, e)
+
self._fp_bytes_read += len(data)
try:
diff --git a/urllib3/util/__init__.py b/urllib3/util/__init__.py
index a40185e..8becc81 100644
--- a/urllib3/util/__init__.py
+++ b/urllib3/util/__init__.py
@@ -1,9 +1,4 @@
-# urllib3/util/__init__.py
-# Copyright 2008-2014 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
-
+# For backwards compatibility, provide imports that used to be here.
from .connection import is_connection_dropped
from .request import make_headers
from .response import is_fp_closed
@@ -19,6 +14,8 @@ from .timeout import (
current_time,
Timeout,
)
+
+from .retry import Retry
from .url import (
get_host,
parse_url,
diff --git a/urllib3/util/connection.py b/urllib3/util/connection.py
index c67ef04..062ee9d 100644
--- a/urllib3/util/connection.py
+++ b/urllib3/util/connection.py
@@ -1,4 +1,4 @@
-from socket import error as SocketError
+import socket
try:
from select import poll, POLLIN
except ImportError: # `poll` doesn't exist on OSX and other platforms
@@ -31,7 +31,7 @@ def is_connection_dropped(conn): # Platform-specific
try:
return select([sock], [], [], 0.0)[0]
- except SocketError:
+ except socket.error:
return True
# This version is better on platforms that support it.
@@ -41,3 +41,55 @@ def is_connection_dropped(conn): # Platform-specific
if fno == sock.fileno():
# Either data is buffered (bad), or the connection is dropped.
return True
+
+
+# This function is copied from socket.py in the Python 2.7 standard
+# library test suite. Added to its signature is only `socket_options`.
+def create_connection(address, timeout=socket._GLOBAL_DEFAULT_TIMEOUT,
+ source_address=None, socket_options=None):
+ """Connect to *address* and return the socket object.
+
+ Convenience function. Connect to *address* (a 2-tuple ``(host,
+ port)``) and return the socket object. Passing the optional
+ *timeout* parameter will set the timeout on the socket instance
+ before attempting to connect. If no *timeout* is supplied, the
+ global default timeout setting returned by :func:`getdefaulttimeout`
+ is used. If *source_address* is set it must be a tuple of (host, port)
+ for the socket to bind as a source address before making the connection.
+ An host of '' or port 0 tells the OS to use the default.
+ """
+
+ host, port = address
+ err = None
+ for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM):
+ af, socktype, proto, canonname, sa = res
+ sock = None
+ try:
+ sock = socket.socket(af, socktype, proto)
+ if timeout is not socket._GLOBAL_DEFAULT_TIMEOUT:
+ sock.settimeout(timeout)
+ if source_address:
+ sock.bind(source_address)
+ # If provided, set socket level options before connecting.
+ # This is the only addition urllib3 makes to this function.
+ _set_socket_options(sock, socket_options)
+ sock.connect(sa)
+ return sock
+
+ except socket.error as _:
+ err = _
+ if sock is not None:
+ sock.close()
+
+ if err is not None:
+ raise err
+ else:
+ raise socket.error("getaddrinfo returns an empty list")
+
+
+def _set_socket_options(sock, options):
+ if options is None:
+ return
+
+ for opt in options:
+ sock.setsockopt(*opt)
diff --git a/urllib3/util/request.py b/urllib3/util/request.py
index bfd7a98..bc64f6b 100644
--- a/urllib3/util/request.py
+++ b/urllib3/util/request.py
@@ -1,7 +1,6 @@
from base64 import b64encode
-from ..packages import six
-
+from ..packages.six import b
ACCEPT_ENCODING = 'gzip,deflate'
@@ -29,13 +28,13 @@ def make_headers(keep_alive=None, accept_encoding=None, user_agent=None,
auth header.
:param proxy_basic_auth:
- Colon-separated username:password string for
- 'proxy-authorization: basic ...' auth header.
+ Colon-separated username:password string for 'proxy-authorization: basic ...'
+ auth header.
:param disable_cache:
If ``True``, adds 'cache-control: no-cache' header.
- Example: ::
+ Example::
>>> make_headers(keep_alive=True, user_agent="Batman/1.0")
{'connection': 'keep-alive', 'user-agent': 'Batman/1.0'}
@@ -60,11 +59,11 @@ def make_headers(keep_alive=None, accept_encoding=None, user_agent=None,
if basic_auth:
headers['authorization'] = 'Basic ' + \
- b64encode(six.b(basic_auth)).decode('utf-8')
+ b64encode(b(basic_auth)).decode('utf-8')
if proxy_basic_auth:
headers['proxy-authorization'] = 'Basic ' + \
- b64encode(six.b(proxy_basic_auth)).decode('utf-8')
+ b64encode(b(proxy_basic_auth)).decode('utf-8')
if disable_cache:
headers['cache-control'] = 'no-cache'
diff --git a/urllib3/util/retry.py b/urllib3/util/retry.py
new file mode 100644
index 0000000..9013197
--- /dev/null
+++ b/urllib3/util/retry.py
@@ -0,0 +1,279 @@
+import time
+import logging
+
+from ..exceptions import (
+ ProtocolError,
+ ConnectTimeoutError,
+ ReadTimeoutError,
+ MaxRetryError,
+)
+from ..packages import six
+
+
+log = logging.getLogger(__name__)
+
+
+class Retry(object):
+ """ Retry configuration.
+
+ Each retry attempt will create a new Retry object with updated values, so
+ they can be safely reused.
+
+ Retries can be defined as a default for a pool::
+
+ retries = Retry(connect=5, read=2, redirect=5)
+ http = PoolManager(retries=retries)
+ response = http.request('GET', 'http://example.com/')
+
+ Or per-request (which overrides the default for the pool)::
+
+ response = http.request('GET', 'http://example.com/', retries=Retry(10))
+
+ Retries can be disabled by passing ``False``::
+
+ response = http.request('GET', 'http://example.com/', retries=False)
+
+ Errors will be wrapped in :class:`~urllib3.exceptions.MaxRetryError` unless
+ retries are disabled, in which case the causing exception will be raised.
+
+
+ :param int total:
+ Total number of retries to allow. Takes precedence over other counts.
+
+ Set to ``None`` to remove this constraint and fall back on other
+ counts. It's a good idea to set this to some sensibly-high value to
+ account for unexpected edge cases and avoid infinite retry loops.
+
+ Set to ``0`` to fail on the first retry.
+
+ Set to ``False`` to disable and imply ``raise_on_redirect=False``.
+
+ :param int connect:
+ How many connection-related errors to retry on.
+
+ These are errors raised before the request is sent to the remote server,
+ which we assume has not triggered the server to process the request.
+
+ Set to ``0`` to fail on the first retry of this type.
+
+ :param int read:
+ How many times to retry on read errors.
+
+ These errors are raised after the request was sent to the server, so the
+ request may have side-effects.
+
+ Set to ``0`` to fail on the first retry of this type.
+
+ :param int redirect:
+ How many redirects to perform. Limit this to avoid infinite redirect
+ loops.
+
+ A redirect is a HTTP response with a status code 301, 302, 303, 307 or
+ 308.
+
+ Set to ``0`` to fail on the first retry of this type.
+
+ Set to ``False`` to disable and imply ``raise_on_redirect=False``.
+
+ :param iterable method_whitelist:
+ Set of uppercased HTTP method verbs that we should retry on.
+
+ By default, we only retry on methods which are considered to be
+ indempotent (multiple requests with the same parameters end with the
+ same state). See :attr:`Retry.DEFAULT_METHOD_WHITELIST`.
+
+ :param iterable status_forcelist:
+ A set of HTTP status codes that we should force a retry on.
+
+ By default, this is disabled with ``None``.
+
+ :param float backoff_factor:
+ A backoff factor to apply between attempts. urllib3 will sleep for::
+
+ {backoff factor} * (2 ^ ({number of total retries} - 1))
+
+ seconds. If the backoff_factor is 0.1, then :func:`.sleep` will sleep
+ for [0.1s, 0.2s, 0.4s, ...] between retries. It will never be longer
+ than :attr:`Retry.MAX_BACKOFF`.
+
+ By default, backoff is disabled (set to 0).
+
+ :param bool raise_on_redirect: Whether, if the number of redirects is
+ exhausted, to raise a MaxRetryError, or to return a response with a
+ response code in the 3xx range.
+ """
+
+ DEFAULT_METHOD_WHITELIST = frozenset([
+ 'HEAD', 'GET', 'PUT', 'DELETE', 'OPTIONS', 'TRACE'])
+
+ #: Maximum backoff time.
+ BACKOFF_MAX = 120
+
+ def __init__(self, total=10, connect=None, read=None, redirect=None,
+ method_whitelist=DEFAULT_METHOD_WHITELIST, status_forcelist=None,
+ backoff_factor=0, raise_on_redirect=True, _observed_errors=0):
+
+ self.total = total
+ self.connect = connect
+ self.read = read
+
+ if redirect is False or total is False:
+ redirect = 0
+ raise_on_redirect = False
+
+ self.redirect = redirect
+ self.status_forcelist = status_forcelist or set()
+ self.method_whitelist = method_whitelist
+ self.backoff_factor = backoff_factor
+ self.raise_on_redirect = raise_on_redirect
+ self._observed_errors = _observed_errors # TODO: use .history instead?
+
+ def new(self, **kw):
+ params = dict(
+ total=self.total,
+ connect=self.connect, read=self.read, redirect=self.redirect,
+ method_whitelist=self.method_whitelist,
+ status_forcelist=self.status_forcelist,
+ backoff_factor=self.backoff_factor,
+ raise_on_redirect=self.raise_on_redirect,
+ _observed_errors=self._observed_errors,
+ )
+ params.update(kw)
+ return type(self)(**params)
+
+ @classmethod
+ def from_int(cls, retries, redirect=True, default=None):
+ """ Backwards-compatibility for the old retries format."""
+ if retries is None:
+ retries = default if default is not None else cls.DEFAULT
+
+ if isinstance(retries, Retry):
+ return retries
+
+ redirect = bool(redirect) and None
+ new_retries = cls(retries, redirect=redirect)
+ log.debug("Converted retries value: %r -> %r" % (retries, new_retries))
+ return new_retries
+
+ def get_backoff_time(self):
+ """ Formula for computing the current backoff
+
+ :rtype: float
+ """
+ if self._observed_errors <= 1:
+ return 0
+
+ backoff_value = self.backoff_factor * (2 ** (self._observed_errors - 1))
+ return min(self.BACKOFF_MAX, backoff_value)
+
+ def sleep(self):
+ """ Sleep between retry attempts using an exponential backoff.
+
+ By default, the backoff factor is 0 and this method will return
+ immediately.
+ """
+ backoff = self.get_backoff_time()
+ if backoff <= 0:
+ return
+ time.sleep(backoff)
+
+ def _is_connection_error(self, err):
+ """ Errors when we're fairly sure that the server did not receive the
+ request, so it should be safe to retry.
+ """
+ return isinstance(err, ConnectTimeoutError)
+
+ def _is_read_error(self, err):
+ """ Errors that occur after the request has been started, so we can't
+ assume that the server did not process any of it.
+ """
+ return isinstance(err, (ReadTimeoutError, ProtocolError))
+
+ def is_forced_retry(self, method, status_code):
+ """ Is this method/response retryable? (Based on method/codes whitelists)
+ """
+ if self.method_whitelist and method.upper() not in self.method_whitelist:
+ return False
+
+ return self.status_forcelist and status_code in self.status_forcelist
+
+ def is_exhausted(self):
+ """ Are we out of retries?
+ """
+ retry_counts = (self.total, self.connect, self.read, self.redirect)
+ retry_counts = list(filter(None, retry_counts))
+ if not retry_counts:
+ return False
+
+ return min(retry_counts) < 0
+
+ def increment(self, method=None, url=None, response=None, error=None, _pool=None, _stacktrace=None):
+ """ Return a new Retry object with incremented retry counters.
+
+ :param response: A response object, or None, if the server did not
+ return a response.
+ :type response: :class:`~urllib3.response.HTTPResponse`
+ :param Exception error: An error encountered during the request, or
+ None if the response was received successfully.
+
+ :return: A new ``Retry`` object.
+ """
+ if self.total is False and error:
+ # Disabled, indicate to re-raise the error.
+ raise six.reraise(type(error), error, _stacktrace)
+
+ total = self.total
+ if total is not None:
+ total -= 1
+
+ _observed_errors = self._observed_errors
+ connect = self.connect
+ read = self.read
+ redirect = self.redirect
+
+ if error and self._is_connection_error(error):
+ # Connect retry?
+ if connect is False:
+ raise six.reraise(type(error), error, _stacktrace)
+ elif connect is not None:
+ connect -= 1
+ _observed_errors += 1
+
+ elif error and self._is_read_error(error):
+ # Read retry?
+ if read is False:
+ raise six.reraise(type(error), error, _stacktrace)
+ elif read is not None:
+ read -= 1
+ _observed_errors += 1
+
+ elif response and response.get_redirect_location():
+ # Redirect retry?
+ if redirect is not None:
+ redirect -= 1
+
+ else:
+ # FIXME: Nothing changed, scenario doesn't make sense.
+ _observed_errors += 1
+
+ new_retry = self.new(
+ total=total,
+ connect=connect, read=read, redirect=redirect,
+ _observed_errors=_observed_errors)
+
+ if new_retry.is_exhausted():
+ raise MaxRetryError(_pool, url, error)
+
+ log.debug("Incremented Retry for (url='%s'): %r" % (url, new_retry))
+
+ return new_retry
+
+
+ def __repr__(self):
+ return ('{cls.__name__}(total={self.total}, connect={self.connect}, '
+ 'read={self.read}, redirect={self.redirect})').format(
+ cls=type(self), self=self)
+
+
+# For backwards compatibility (equivalent to pre-v1.9):
+Retry.DEFAULT = Retry(3)
diff --git a/urllib3/util/ssl_.py b/urllib3/util/ssl_.py
index dee4b87..9cfe2d2 100644
--- a/urllib3/util/ssl_.py
+++ b/urllib3/util/ssl_.py
@@ -34,10 +34,9 @@ def assert_fingerprint(cert, fingerprint):
}
fingerprint = fingerprint.replace(':', '').lower()
+ digest_length, odd = divmod(len(fingerprint), 2)
- digest_length, rest = divmod(len(fingerprint), 2)
-
- if rest or digest_length not in hashfunc_map:
+ if odd or digest_length not in hashfunc_map:
raise SSLError('Fingerprint is of invalid length.')
# We need encode() here for py32; works on py2 and p33.
diff --git a/urllib3/util/timeout.py b/urllib3/util/timeout.py
index aaadc12..ea7027f 100644
--- a/urllib3/util/timeout.py
+++ b/urllib3/util/timeout.py
@@ -1,32 +1,49 @@
+# The default socket timeout, used by httplib to indicate that no timeout was
+# specified by the user
from socket import _GLOBAL_DEFAULT_TIMEOUT
import time
from ..exceptions import TimeoutStateError
+# A sentinel value to indicate that no timeout was specified by the user in
+# urllib3
+_Default = object()
def current_time():
"""
- Retrieve the current time, this function is mocked out in unit testing.
+ Retrieve the current time. This function is mocked out in unit testing.
"""
return time.time()
-_Default = object()
-# The default timeout to use for socket connections. This is the attribute used
-# by httplib to define the default timeout
+class Timeout(object):
+ """ Timeout configuration.
+ Timeouts can be defined as a default for a pool::
-class Timeout(object):
- """
- Utility object for storing timeout values.
+ timeout = Timeout(connect=2.0, read=7.0)
+ http = PoolManager(timeout=timeout)
+ response = http.request('GET', 'http://example.com/')
+
+ Or per-request (which overrides the default for the pool)::
+
+ response = http.request('GET', 'http://example.com/', timeout=Timeout(10))
- Example usage:
+ Timeouts can be disabled by setting all the parameters to ``None``::
+
+ no_timeout = Timeout(connect=None, read=None)
+ response = http.request('GET', 'http://example.com/, timeout=no_timeout)
+
+
+ :param total:
+ This combines the connect and read timeouts into one; the read timeout
+ will be set to the time leftover from the connect attempt. In the
+ event that both a connect timeout and a total are specified, or a read
+ timeout and a total are specified, the shorter timeout will be applied.
- .. code-block:: python
+ Defaults to None.
- timeout = urllib3.util.Timeout(connect=2.0, read=7.0)
- pool = HTTPConnectionPool('www.google.com', 80, timeout=timeout)
- pool.request(...) # Etc, etc
+ :type total: integer, float, or None
:param connect:
The maximum amount of time to wait for a connection attempt to a server
@@ -47,25 +64,15 @@ class Timeout(object):
:type read: integer, float, or None
- :param total:
- This combines the connect and read timeouts into one; the read timeout
- will be set to the time leftover from the connect attempt. In the
- event that both a connect timeout and a total are specified, or a read
- timeout and a total are specified, the shorter timeout will be applied.
-
- Defaults to None.
-
- :type total: integer, float, or None
-
.. note::
Many factors can affect the total amount of time for urllib3 to return
- an HTTP response. Specifically, Python's DNS resolver does not obey the
- timeout specified on the socket. Other factors that can affect total
- request time include high CPU load, high swap, the program running at a
- low priority level, or other behaviors. The observed running time for
- urllib3 to return a response may be greater than the value passed to
- `total`.
+ an HTTP response.
+
+ For example, Python's DNS resolver does not obey the timeout specified
+ on the socket. Other factors that can affect total request time include
+ high CPU load, high swap, the program running at a low priority level,
+ or other behaviors.
In addition, the read and total timeouts only measure the time between
read operations on the socket connecting the client and the server,
@@ -73,8 +80,8 @@ class Timeout(object):
response. For most requests, the timeout is raised because the server
has not sent the first byte in the specified time. This is not always
the case; if a server streams one byte every fifteen seconds, a timeout
- of 20 seconds will not ever trigger, even though the request will
- take several minutes to complete.
+ of 20 seconds will not trigger, even though the request will take
+ several minutes to complete.
If your goal is to cut off any request after a set amount of wall clock
time, consider having a second "watcher" thread to cut off a slow
diff --git a/urllib3/util/url.py b/urllib3/util/url.py
index 122108b..487d456 100644
--- a/urllib3/util/url.py
+++ b/urllib3/util/url.py
@@ -2,6 +2,7 @@ from collections import namedtuple
from ..exceptions import LocationParseError
+
url_attrs = ['scheme', 'auth', 'host', 'port', 'path', 'query', 'fragment']
@@ -47,7 +48,7 @@ def split_first(s, delims):
If not found, then the first part is the full input string.
- Example: ::
+ Example::
>>> split_first('foo/bar?baz', '?/=')
('foo', 'bar?baz', '/')
@@ -80,7 +81,7 @@ def parse_url(url):
Partly backwards-compatible with :mod:`urlparse`.
- Example: ::
+ Example::
>>> parse_url('http://google.com/mail/')
Url(scheme='http', host='google.com', port=None, path='/', ...)
@@ -95,6 +96,10 @@ def parse_url(url):
# Additionally, this implementations does silly things to be optimal
# on CPython.
+ if not url:
+ # Empty
+ return Url()
+
scheme = None
auth = None
host = None
generated by cgit v1.2.3 (git 2.45.0) at 2024-11-21 19:05:45 +0000