#
# Copyright 2009 Facebook
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
"""HTTP utility code shared by clients and servers.
This module also defines the `HTTPServerRequest` class which is exposed
via `tornado.web.RequestHandler.request`.
"""
from __future__ import absolute_import, division, print_function
import calendar
import collections
import copy
import dataclasses
import datetime
import email.utils
import numbers
import re
import time
import unicodedata
import warnings
from tornado.escape import native_str, parse_qs_bytes, utf8
from tornado.log import gen_log
from tornado.util import ObjectDict, PY3, unicode_type
if PY3:
import http.cookies as Cookie
from http.client import responses
from urllib.parse import urlencode, urlparse, urlunparse, parse_qsl
else:
import Cookie
from httplib import responses
from urllib import urlencode
from urlparse import urlparse, urlunparse, parse_qsl
# responses is unused in this file, but we re-export it to other files.
# Reference it so pyflakes doesn't complain.
responses
try:
from ssl import SSLError
except ImportError:
# ssl is unavailable on app engine.
class _SSLError(Exception):
pass
# Hack around a mypy limitation. We can't simply put "type: ignore"
# on the class definition itself; must go through an assignment.
SSLError = _SSLError # type: ignore
try:
import typing # noqa: F401
except ImportError:
pass
class _ABNF:
"""Class that holds a subset of ABNF rules from RFC 9110 and friends.
Class attributes are re.Pattern objects, with the same name as in the RFC
(with hyphens changed to underscores). Currently contains only the subset
we use (which is why this class is not public). Unfortunately the fields
cannot be alphabetized as they are in the RFCs because of dependencies.
"""
# RFC 5234 (ABNF)
VCHAR = re.compile(r"[\x21-\x7E]")
# RFC 9110 (HTTP Semantics)
obs_text = re.compile(r"[\x80-\xFF]")
tchar = re.compile(r"[!#$%&'*+\-.^_`|~0-9A-Za-z]")
token = re.compile(rf"{tchar.pattern}+")
field_name = token
# RFC 9112 (HTTP/1.1)
HTTP_version = re.compile(r"HTTP/[0-9]\.[0-9]")
reason_phrase = re.compile(rf"(?:[\t ]|{VCHAR.pattern}|{obs_text.pattern})+")
status_code = re.compile(r"[0-9]{3}")
status_line = re.compile(
rf"({HTTP_version.pattern}) ({status_code.pattern}) ({reason_phrase.pattern})?"
)
# RFC 7230 section 3.5: a recipient MAY recognize a single LF as a line
# terminator and ignore any preceding CR.
_CRLF_RE = re.compile(r'\r?\n')
class _NormalizedHeaderCache(dict):
"""Dynamic cached mapping of header names to Http-Header-Case.
Implemented as a dict subclass so that cache hits are as fast as a
normal dict lookup, without the overhead of a python function
call.
>>> normalized_headers = _NormalizedHeaderCache(10)
>>> normalized_headers["coNtent-TYPE"]
'Content-Type'
"""
def __init__(self, size):
super(_NormalizedHeaderCache, self).__init__()
self.size = size
self.queue = collections.deque()
def __missing__(self, key):
normalized = "-".join([w.capitalize() for w in key.split("-")])
self[key] = normalized
self.queue.append(key)
if len(self.queue) > self.size:
# Limit the size of the cache. LRU would be better, but this
# simpler approach should be fine. In Python 2.7+ we could
# use OrderedDict (or in 3.2+, @functools.lru_cache).
old_key = self.queue.popleft()
del self[old_key]
return normalized
_normalized_headers = _NormalizedHeaderCache(1000)
[docs]class HTTPServerRequest(object):
"""A single HTTP request.
All attributes are type `str` unless otherwise noted.
.. attribute:: method
HTTP request method, e.g. "GET" or "POST"
.. attribute:: uri
The requested uri.
.. attribute:: path
The path portion of `uri`
.. attribute:: query
The query portion of `uri`
.. attribute:: version
HTTP version specified in request, e.g. "HTTP/1.1"
.. attribute:: headers
`.HTTPHeaders` dictionary-like object for request headers. Acts like
a case-insensitive dictionary with additional methods for repeated
headers.
.. attribute:: body
Request body, if present, as a byte string.
.. attribute:: remote_ip
Client's IP address as a string. If ``HTTPServer.xheaders`` is set,
will pass along the real IP address provided by a load balancer
in the ``X-Real-Ip`` or ``X-Forwarded-For`` header.
.. versionchanged:: 3.1
The list format of ``X-Forwarded-For`` is now supported.
.. attribute:: protocol
The protocol used, either "http" or "https". If ``HTTPServer.xheaders``
is set, will pass along the protocol used by a load balancer if
reported via an ``X-Scheme`` header.
.. attribute:: host
The requested hostname, usually taken from the ``Host`` header.
.. attribute:: arguments
GET/POST arguments are available in the arguments property, which
maps arguments names to lists of values (to support multiple values
for individual names). Names are of type `str`, while arguments
are byte strings. Note that this is different from
`.RequestHandler.get_argument`, which returns argument values as
unicode strings.
.. attribute:: query_arguments
Same format as ``arguments``, but contains only arguments extracted
from the query string.
.. versionadded:: 3.2
.. attribute:: body_arguments
Same format as ``arguments``, but contains only arguments extracted
from the request body.
.. versionadded:: 3.2
.. attribute:: files
File uploads are available in the files property, which maps file
names to lists of `.HTTPFile`.
.. attribute:: connection
An HTTP request is attached to a single HTTP connection, which can
be accessed through the "connection" attribute. Since connections
are typically kept open in HTTP/1.1, multiple requests can be handled
sequentially on a single connection.
.. versionchanged:: 4.0
Moved from ``tornado.httpserver.HTTPRequest``.
"""
def __init__(self, method=None, uri=None, version="HTTP/1.0", headers=None,
body=None, host=None, files=None, connection=None,
start_line=None, server_connection=None):
if start_line is not None:
method, uri, version = start_line
self.method = method
self.uri = uri
self.version = version
self.headers = headers or HTTPHeaders()
self.body = body or b""
# set remote IP and protocol
context = getattr(connection, 'context', None)
self.remote_ip = getattr(context, 'remote_ip', None)
self.protocol = getattr(context, 'protocol', "http")
self.host = host or self.headers.get("Host") or "127.0.0.1"
self.host_name = split_host_and_port(self.host.lower())[0]
self.files = files or {}
self.connection = connection
self.server_connection = server_connection
self._start_time = time.time()
self._finish_time = None
self.path, sep, self.query = uri.partition('?')
self.arguments = parse_qs_bytes(self.query, keep_blank_values=True)
self.query_arguments = copy.deepcopy(self.arguments)
self.body_arguments = {}
[docs] def supports_http_1_1(self):
"""Returns True if this request supports HTTP/1.1 semantics.
.. deprecated:: 4.0
Applications are less likely to need this information with
the introduction of `.HTTPConnection`. If you still need
it, access the ``version`` attribute directly. This method
will be removed in Tornado 6.0.
"""
warnings.warn("supports_http_1_1() is deprecated, use request.version instead",
DeprecationWarning)
return self.version == "HTTP/1.1"
@property
def cookies(self):
"""A dictionary of Cookie.Morsel objects."""
if not hasattr(self, "_cookies"):
self._cookies = Cookie.SimpleCookie()
if "Cookie" in self.headers:
try:
parsed = parse_cookie(self.headers["Cookie"])
except Exception:
pass
else:
for k, v in parsed.items():
try:
self._cookies[k] = v
except Exception:
# SimpleCookie imposes some restrictions on keys;
# parse_cookie does not. Discard any cookies
# with disallowed keys.
pass
return self._cookies
[docs] def write(self, chunk, callback=None):
"""Writes the given chunk to the response stream.
.. deprecated:: 4.0
Use ``request.connection`` and the `.HTTPConnection` methods
to write the response. This method will be removed in Tornado 6.0.
"""
warnings.warn("req.write deprecated, use req.connection.write and write_headers instead",
DeprecationWarning)
assert isinstance(chunk, bytes)
assert self.version.startswith("HTTP/1."), \
"deprecated interface only supported in HTTP/1.x"
self.connection.write(chunk, callback=callback)
[docs] def finish(self):
"""Finishes this HTTP request on the open connection.
.. deprecated:: 4.0
Use ``request.connection`` and the `.HTTPConnection` methods
to write the response. This method will be removed in Tornado 6.0.
"""
warnings.warn("req.finish deprecated, use req.connection.finish instead",
DeprecationWarning)
self.connection.finish()
self._finish_time = time.time()
[docs] def full_url(self):
"""Reconstructs the full URL for this request."""
return self.protocol + "://" + self.host + self.uri
[docs] def request_time(self):
"""Returns the amount of time it took for this request to execute."""
if self._finish_time is None:
return time.time() - self._start_time
else:
return self._finish_time - self._start_time
[docs] def get_ssl_certificate(self, binary_form=False):
"""Returns the client's SSL certificate, if any.
To use client certificates, the HTTPServer's
`ssl.SSLContext.verify_mode` field must be set, e.g.::
ssl_ctx = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH)
ssl_ctx.load_cert_chain("foo.crt", "foo.key")
ssl_ctx.load_verify_locations("cacerts.pem")
ssl_ctx.verify_mode = ssl.CERT_REQUIRED
server = HTTPServer(app, ssl_options=ssl_ctx)
By default, the return value is a dictionary (or None, if no
client certificate is present). If ``binary_form`` is true, a
DER-encoded form of the certificate is returned instead. See
SSLSocket.getpeercert() in the standard library for more
details.
http://docs.python.org/library/ssl.html#sslsocket-objects
"""
try:
return self.connection.stream.socket.getpeercert(
binary_form=binary_form)
except SSLError:
return None
def _parse_body(self):
parse_body_arguments(
self.headers.get("Content-Type", ""), self.body,
self.body_arguments, self.files,
self.headers)
for k, v in self.body_arguments.items():
self.arguments.setdefault(k, []).extend(v)
def __repr__(self):
attrs = ("protocol", "host", "method", "uri", "version", "remote_ip")
args = ", ".join(["%s=%r" % (n, getattr(self, n)) for n in attrs])
return "%s(%s)" % (self.__class__.__name__, args)
[docs]class HTTPOutputError(Exception):
"""Exception class for errors in HTTP output.
.. versionadded:: 4.0
"""
pass
[docs]class HTTPServerConnectionDelegate(object):
"""Implement this interface to handle requests from `.HTTPServer`.
.. versionadded:: 4.0
"""
[docs] def start_request(self, server_conn, request_conn):
"""This method is called by the server when a new request has started.
:arg server_conn: is an opaque object representing the long-lived
(e.g. tcp-level) connection.
:arg request_conn: is a `.HTTPConnection` object for a single
request/response exchange.
This method should return a `.HTTPMessageDelegate`.
"""
raise NotImplementedError()
[docs] def on_close(self, server_conn):
"""This method is called when a connection has been closed.
:arg server_conn: is a server connection that has previously been
passed to ``start_request``.
"""
pass
[docs]class HTTPMessageDelegate(object):
"""Implement this interface to handle an HTTP request or response.
.. versionadded:: 4.0
"""
[docs] def data_received(self, chunk):
"""Called when a chunk of data has been received.
May return a `.Future` for flow control.
"""
pass
[docs] def finish(self):
"""Called after the last chunk of data has been received."""
pass
[docs] def on_connection_close(self):
"""Called if the connection is closed without finishing the request.
If ``headers_received`` is called, either ``finish`` or
``on_connection_close`` will be called, but not both.
"""
pass
[docs]class HTTPConnection(object):
"""Applications use this interface to write their responses.
.. versionadded:: 4.0
"""
[docs] def write(self, chunk, callback=None):
"""Writes a chunk of body data.
The callback will be run when the write is complete. If no callback
is given, returns a Future.
.. deprecated:: 5.1
The ``callback`` argument is deprecated and will be removed
in Tornado 6.0.
"""
raise NotImplementedError()
[docs] def finish(self):
"""Indicates that the last body data has been written.
"""
raise NotImplementedError()
[docs]def url_concat(url, args):
"""Concatenate url and arguments regardless of whether
url has existing query parameters.
``args`` may be either a dictionary or a list of key-value pairs
(the latter allows for multiple values with the same key.
>>> url_concat("http://example.com/foo", dict(c="d"))
'http://example.com/foo?c=d'
>>> url_concat("http://example.com/foo?a=b", dict(c="d"))
'http://example.com/foo?a=b&c=d'
>>> url_concat("http://example.com/foo?a=b", [("c", "d"), ("c", "d2")])
'http://example.com/foo?a=b&c=d&c=d2'
"""
if args is None:
return url
parsed_url = urlparse(url)
if isinstance(args, dict):
parsed_query = parse_qsl(parsed_url.query, keep_blank_values=True)
parsed_query.extend(args.items())
elif isinstance(args, list) or isinstance(args, tuple):
parsed_query = parse_qsl(parsed_url.query, keep_blank_values=True)
parsed_query.extend(args)
else:
err = "'args' parameter should be dict, list or tuple. Not {0}".format(
type(args))
raise TypeError(err)
final_query = urlencode(parsed_query)
url = urlunparse((
parsed_url[0],
parsed_url[1],
parsed_url[2],
parsed_url[3],
final_query,
parsed_url[5]))
return url
[docs]class HTTPFile(ObjectDict):
"""Represents a file uploaded via a form.
For backwards compatibility, its instance attributes are also
accessible as dictionary keys.
* ``filename``
* ``body``
* ``content_type``
"""
pass
def _parse_request_range(range_header):
"""Parses a Range header.
Returns either ``None`` or tuple ``(start, end)``.
Note that while the HTTP headers use inclusive byte positions,
this method returns indexes suitable for use in slices.
>>> start, end = _parse_request_range("bytes=1-2")
>>> start, end
(1, 3)
>>> [0, 1, 2, 3, 4][start:end]
[1, 2]
>>> _parse_request_range("bytes=6-")
(6, None)
>>> _parse_request_range("bytes=-6")
(-6, None)
>>> _parse_request_range("bytes=-0")
(None, 0)
>>> _parse_request_range("bytes=")
(None, None)
>>> _parse_request_range("foo=42")
>>> _parse_request_range("bytes=1-2,6-10")
Note: only supports one range (ex, ``bytes=1-2,6-10`` is not allowed).
See [0] for the details of the range header.
[0]: http://greenbytes.de/tech/webdav/draft-ietf-httpbis-p5-range-latest.html#byte.ranges
"""
unit, _, value = range_header.partition("=")
unit, value = unit.strip(), value.strip()
if unit != "bytes":
return None
start_b, _, end_b = value.partition("-")
try:
start = _int_or_none(start_b)
end = _int_or_none(end_b)
except ValueError:
return None
if end is not None:
if start is None:
if end != 0:
start = -end
end = None
else:
end += 1
return (start, end)
def _get_content_range(start, end, total):
"""Returns a suitable Content-Range header:
>>> print(_get_content_range(None, 1, 4))
bytes 0-0/4
>>> print(_get_content_range(1, 3, 4))
bytes 1-2/4
>>> print(_get_content_range(None, None, 4))
bytes 0-3/4
"""
start = start or 0
end = (end or total) - 1
return "bytes %s-%s/%s" % (start, end, total)
def _int_or_none(val):
val = val.strip()
if val == "":
return None
return int(val)
[docs]@dataclasses.dataclass
class ParseMultipartConfig:
"""This class configures the parsing of ``multipart/form-data`` request bodies.
Its primary purpose is to place limits on the size and complexity of request messages
to avoid potential denial-of-service attacks.
.. versionadded:: 6.5.5
"""
enabled: bool = True
"""Set this to false to disable the parsing of ``multipart/form-data`` requests entirely.
This may be desirable for applications that do not need to handle this format, since
multipart request have a history of DoS vulnerabilities in Tornado. Multipart requests
are used primarily for ``<input type="file">`` in HTML forms, or in APIs that mimic this
format. File uploads that use the HTTP ``PUT`` method generally do not use the multipart
format.
"""
max_parts: int = 100
"""The maximum number of parts accepted in a multipart request.
Each ``<input>`` element in an HTML form corresponds to at least one "part".
"""
max_part_header_size: int = 10 * 1024
"""The maximum size of the headers for each part of a multipart request.
The header for a part contains the name of the form field and optionally the filename
and content type of the uploaded file.
"""
[docs]@dataclasses.dataclass
class ParseBodyConfig:
"""This class configures the parsing of request bodies.
.. versionadded:: 6.5.5
"""
multipart: ParseMultipartConfig = dataclasses.field(
default_factory=ParseMultipartConfig
)
"""Configuration for ``multipart/form-data`` request bodies."""
_DEFAULT_PARSE_BODY_CONFIG = ParseBodyConfig()
[docs]def set_parse_body_config(config):
r"""Sets the **global** default configuration for parsing request bodies.
This global setting is provided as a stopgap for applications that need to raise the limits
introduced in Tornado 6.5.5, or who wish to disable the parsing of multipart/form-data bodies
entirely. Non-global configuration for this functionality will be introduced in a future
release.
>>> content_type = "multipart/form-data; boundary=foo"
>>> multipart_body = b"--foo--\r\n"
>>> parse_body_arguments(content_type, multipart_body, {}, {})
>>> multipart_config = ParseMultipartConfig(enabled=False)
>>> config = ParseBodyConfig(multipart=multipart_config)
>>> set_parse_body_config(config)
>>> set_parse_body_config(ParseBodyConfig()) # reset to defaults
.. versionadded:: 6.5.5
"""
global _DEFAULT_PARSE_BODY_CONFIG
_DEFAULT_PARSE_BODY_CONFIG = config
[docs]def parse_body_arguments(content_type, body, arguments, files, headers=None, *, config = None):
"""Parses a form request body.
Supports ``application/x-www-form-urlencoded`` and
``multipart/form-data``. The ``content_type`` parameter should be
a string and ``body`` should be a byte string. The ``arguments``
and ``files`` parameters are dictionaries that will be updated
with the parsed contents.
"""
if config is None:
config = _DEFAULT_PARSE_BODY_CONFIG
if headers and 'Content-Encoding' in headers:
gen_log.warning("Unsupported Content-Encoding: %s",
headers['Content-Encoding'])
return
if content_type.startswith("application/x-www-form-urlencoded"):
try:
uri_arguments = parse_qs_bytes(native_str(body), keep_blank_values=True)
except Exception as e:
gen_log.warning('Invalid x-www-form-urlencoded body: %s', e)
uri_arguments = {}
for name, values in uri_arguments.items():
if values:
arguments.setdefault(name, []).extend(values)
elif content_type.startswith("multipart/form-data"):
try:
fields = content_type.split(";")
if fields[0].strip() != "multipart/form-data":
# This catches "Content-Type: multipart/form-dataxyz"
raise HTTPInputError("Invalid content type")
for field in fields:
k, sep, v = field.strip().partition("=")
if k == "boundary" and v:
parse_multipart_form_data(
utf8(v), body, arguments, files, config=config.multipart
)
break
else:
raise ValueError("multipart boundary not found")
except Exception as e:
gen_log.warning("Invalid multipart/form-data: %s", e)
RequestStartLine = collections.namedtuple(
'RequestStartLine', ['method', 'path', 'version'])
[docs]def parse_request_start_line(line):
"""Returns a (method, path, version) tuple for an HTTP 1.x request line.
The response is a `collections.namedtuple`.
>>> parse_request_start_line("GET /foo HTTP/1.1")
RequestStartLine(method='GET', path='/foo', version='HTTP/1.1')
"""
try:
method, path, version = line.split(" ")
except ValueError:
# https://tools.ietf.org/html/rfc7230#section-3.1.1
# invalid request-line SHOULD respond with a 400 (Bad Request)
raise HTTPInputError("Malformed HTTP request line")
if not _ABNF.HTTP_version.fullmatch(version):
raise HTTPInputError(
"Malformed HTTP version in HTTP Request-Line: %r" % version)
if not version.startswith("HTTP/1"):
# HTTP/2 and above doesn't use parse_request_start_line.
# This could be folded into the regex but we don't want to deviate
# from the ABNF in the RFCs.
raise HTTPInputError("Unexpected HTTP version %r" % version)
return RequestStartLine(method, path, version)
ResponseStartLine = collections.namedtuple(
'ResponseStartLine', ['version', 'code', 'reason'])
[docs]def parse_response_start_line(line):
"""Returns a (version, code, reason) tuple for an HTTP 1.x response line.
The response is a `collections.namedtuple`.
>>> parse_response_start_line("HTTP/1.1 200 OK")
ResponseStartLine(version='HTTP/1.1', code=200, reason='OK')
"""
match = _ABNF.status_line.fullmatch(line)
if not match:
raise HTTPInputError("Error parsing response start line")
r = ResponseStartLine(match.group(1), int(match.group(2)), match.group(3))
if not r.version.startswith("HTTP/1"):
# HTTP/2 and above doesn't use parse_response_start_line.
raise HTTPInputError("Unexpected HTTP version %r" % r.version)
return r
# _parseparam and _parse_header are copied and modified from python2.7's cgi.py
# The original 2.7 version of this code did not correctly support some
# combinations of semicolons and double quotes.
# It has also been modified to support valueless parameters as seen in
# websocket extension negotiations, and to support non-ascii values in
# RFC 2231/5987 format.
#
# _parseparam has been further modified with the logic from
# https://github.com/python/cpython/pull/136072/files
# to avoid quadratic behavior when parsing semicolons in quoted strings.
#
# TODO: See if we can switch to email.message.Message for this functionality.
# This is the suggested replacement for the cgi.py module now that cgi has
# been removed from recent versions of Python. We need to verify that
# the email module is consistent with our existing behavior (and all relevant
# RFCs for multipart/form-data) before making this change.
def _parseparam(s):
start = 0
while s.find(";", start) == start:
start += 1
end = s.find(";", start)
ind, diff = start, 0
while end > 0:
diff += s.count('"', ind, end) - s.count('\\"', ind, end)
if diff % 2 == 0:
break
end, ind = ind, s.find(";", end + 1)
if end < 0:
end = len(s)
f = s[start:end]
yield f.strip()
start = end
def _parse_header(line):
r"""Parse a Content-type like header.
Return the main content-type and a dictionary of options.
>>> d = "form-data; foo=\"b\\\\a\\\"r\"; file*=utf-8''T%C3%A4st"
>>> ct, d = _parse_header(d)
>>> ct
'form-data'
>>> d['file'] == r'T\u00e4st'.encode('ascii').decode('unicode_escape')
True
>>> d['foo']
'b\\a"r'
"""
parts = _parseparam(';' + line)
key = next(parts)
# decode_params treats first argument special, but we already stripped key
params = [('Dummy', 'value')]
for p in parts:
i = p.find('=')
if i >= 0:
name = p[:i].strip().lower()
value = p[i + 1:].strip()
params.append((name, native_str(value)))
params = email.utils.decode_params(params)
params.pop(0) # get rid of the dummy again
pdict = {}
for name, value in params:
value = email.utils.collapse_rfc2231_value(value)
if len(value) >= 2 and value[0] == '"' and value[-1] == '"':
value = value[1:-1]
pdict[name] = value
return key, pdict
def _encode_header(key, pdict):
"""Inverse of _parse_header.
>>> _encode_header('permessage-deflate',
... {'client_max_window_bits': 15, 'client_no_context_takeover': None})
'permessage-deflate; client_max_window_bits=15; client_no_context_takeover'
"""
if not pdict:
return key
out = [key]
# Sort the parameters just to make it easy to test.
for k, v in sorted(pdict.items()):
if v is None:
out.append(k)
else:
# TODO: quote if necessary.
out.append('%s=%s' % (k, v))
return '; '.join(out)
[docs]def encode_username_password(username, password):
"""Encodes a username/password pair in the format used by HTTP auth.
The return value is a byte string in the form ``username:password``.
.. versionadded:: 5.1
"""
if isinstance(username, unicode_type):
username = unicodedata.normalize('NFC', username)
if isinstance(password, unicode_type):
password = unicodedata.normalize('NFC', password)
return utf8(username) + b":" + utf8(password)
def doctests():
import doctest
return doctest.DocTestSuite()
[docs]def split_host_and_port(netloc):
"""Returns ``(host, port)`` tuple from ``netloc``.
Returned ``port`` will be ``None`` if not present.
.. versionadded:: 4.1
"""
match = re.match(r'^(.+):(\d+)$', netloc)
if match:
host = match.group(1)
port = int(match.group(2))
else:
host = netloc
port = None
return (host, port)
[docs]def qs_to_qsl(qs):
"""Generator converting a result of ``parse_qs`` back to name-value pairs.
.. versionadded:: 5.0
"""
for k, vs in qs.items():
for v in vs:
yield (k, v)
_unquote_sub = re.compile(r"\\(?:([0-3][0-7][0-7])|(.))").sub
def _unquote_replace(m: re.Match) -> str:
if m[1]:
return chr(int(m[1], 8))
else:
return m[2]
def _unquote_cookie(str):
"""Handle double quotes and escaping in cookie values.
This method is copied verbatim from the Python 3.13 standard
library (http.cookies._unquote) so we don't have to depend on
non-public interfaces.
"""
# If there aren't any doublequotes,
# then there can't be any special characters. See RFC 2109.
if str is None or len(str) < 2:
return str
if str[0] != '"' or str[-1] != '"':
return str
# We have to assume that we must decode this string.
# Down to work.
# Remove the "s
str = str[1:-1]
# Check for special sequences. Examples:
# \012 --> \n
# \" --> "
#
return _unquote_sub(_unquote_replace, str)
[docs]def parse_cookie(cookie):
"""Parse a ``Cookie`` HTTP header into a dict of name/value pairs.
This function attempts to mimic browser cookie parsing behavior;
it specifically does not follow any of the cookie-related RFCs
(because browsers don't either).
The algorithm used is identical to that used by Django version 1.9.10.
.. versionadded:: 4.4.2
"""
cookiedict = {}
for chunk in cookie.split(str(';')):
if str('=') in chunk:
key, val = chunk.split(str('='), 1)
else:
# Assume an empty name per
# https://bugzilla.mozilla.org/show_bug.cgi?id=169091
key, val = str(''), chunk
key, val = key.strip(), val.strip()
if key or val:
# unquote using Python's algorithm.
cookiedict[key] = _unquote_cookie(val)
return cookiedict