This commit is contained in:
macbook_max
2026-10-15 20:36:59 +08:00
commit 6bf949949c
7923 changed files with 1369582 additions and 0 deletions
@@ -0,0 +1,795 @@
from __future__ import annotations
import asyncio
import logging
import os
import socket
import ssl as ssl_module
import traceback
import urllib.parse
from collections.abc import AsyncIterator, Generator, Sequence
from types import TracebackType
from typing import Any, Callable, Literal, cast
from ..client import ClientProtocol, backoff, process_exception
from ..datastructures import Headers, HeadersLike
from ..exceptions import (
InvalidProxyMessage,
InvalidProxyStatus,
InvalidStatus,
ProxyError,
SecurityError,
)
from ..extensions.base import ClientExtensionFactory
from ..extensions.permessage_deflate import enable_client_permessage_deflate
from ..headers import validate_subprotocols
from ..http11 import USER_AGENT, Response
from ..protocol import CONNECTING, Event
from ..proxy import Proxy, get_proxy, parse_proxy, prepare_connect_request
from ..streams import StreamReader
from ..typing import LoggerLike, Origin, PathLike, Subprotocol
from ..uri import WebSocketURI, parse_uri
from .connection import Connection
__all__ = ["connect", "unix_connect", "ClientConnection"]
MAX_REDIRECTS = int(os.environ.get("WEBSOCKETS_MAX_REDIRECTS", "10"))
class ClientConnection(Connection):
"""
:mod:`asyncio` implementation of a WebSocket client connection.
:class:`ClientConnection` provides :meth:`recv` and :meth:`send` coroutines
for receiving and sending messages.
It supports asynchronous iteration to receive messages::
async for message in websocket:
await process(message)
The iterator exits normally when the connection is closed with code
1000 (OK) or 1001 (going away) or without a close code. It raises a
:exc:`~websockets.exceptions.ConnectionClosedError` when the connection is
closed with any other code.
The ``ping_interval``, ``ping_timeout``, ``close_timeout``, ``max_queue``,
and ``write_limit`` arguments have the same meaning as in :func:`connect`.
Args:
protocol: Sans-I/O connection.
"""
def __init__(
self,
protocol: ClientProtocol,
*,
ping_interval: float | None = 20,
ping_timeout: float | None = 20,
close_timeout: float | None = 10,
max_queue: int | None | tuple[int | None, int | None] = 16,
write_limit: int | tuple[int, int | None] = 2**15,
) -> None:
self.protocol: ClientProtocol
super().__init__(
protocol,
ping_interval=ping_interval,
ping_timeout=ping_timeout,
close_timeout=close_timeout,
max_queue=max_queue,
write_limit=write_limit,
)
self.response_rcvd: asyncio.Future[None] = self.loop.create_future()
async def handshake(
self,
additional_headers: HeadersLike | None = None,
user_agent_header: str | None = USER_AGENT,
) -> None:
"""
Perform the opening handshake.
"""
self.request = self.protocol.connect()
if additional_headers is not None:
self.request.headers.update(additional_headers)
if user_agent_header is not None:
self.request.headers.setdefault("User-Agent", user_agent_header)
async with self.send_context(expected_state=CONNECTING):
self.protocol.send_request(self.request)
await asyncio.wait(
[self.response_rcvd, self.connection_lost_waiter],
return_when=asyncio.FIRST_COMPLETED,
)
# self.protocol.handshake_exc is set when the connection is lost before
# receiving a response, when the response cannot be parsed, or when the
# response fails the handshake.
if self.protocol.handshake_exc is not None:
raise self.protocol.handshake_exc
def process_event(self, event: Event) -> None:
"""
Process one incoming event.
"""
# First event - handshake response.
if self.response is None:
assert isinstance(event, Response)
self.response = event
self.response_rcvd.set_result(None)
# Later events - frames.
else:
super().process_event(event)
# This is spelled in lower case because it's exposed as a callable in the API.
class connect:
"""
Connect to the WebSocket server at ``uri``.
:func:`connect` should be treated as an asynchronous context manager
yielding a :class:`ClientConnection`, which can then receive and send
messages::
from websockets.asyncio.client import connect
async with connect(...) as websocket:
...
The connection is closed automatically when exiting the context.
:func:`connect` can also be treated as an infinite asynchronous iterator
to reconnect automatically on errors::
async for websocket in connect(...):
try:
...
except websockets.exceptions.ConnectionClosed:
continue
If the connection fails with a transient error, it is retried with
exponential backoff. If it fails with a fatal error, the exception is
raised, breaking out of the loop.
The connection is closed automatically after each iteration of the loop.
:func:`connect` can be awaited directly::
websocket = await connect(...)
In that case, you're responsible for closing the connection with
:meth:`ClientConnection.close` when no longer needed.
Args:
uri: URI of the WebSocket server.
origin: Value of the ``Origin`` header, for servers that require it.
extensions: List of supported extensions, in order in which they
should be negotiated and run.
subprotocols: List of supported subprotocols, in order of decreasing
preference.
compression: The "permessage-deflate" extension is enabled by default.
Set ``compression`` to :obj:`None` to disable it. See the
:doc:`compression guide <../../topics/compression>` for details.
additional_headers: Arbitrary HTTP headers to add to the handshake
request.
user_agent_header: Value of the ``User-Agent`` request header.
It defaults to ``"Python/x.y.z websockets/X.Y"``.
Setting it to :obj:`None` removes the header.
proxy: If a proxy is configured, it is used by default. Set ``proxy``
to :obj:`None` to disable the proxy or to the address of a proxy
to override the system configuration. See the :doc:`proxy docs
<../../topics/proxies>` for details.
process_exception: When reconnecting automatically, tell whether an
error is transient or fatal. The default behavior is defined by
:func:`~websockets.client.process_exception`. Refer to its
documentation for details.
open_timeout: Timeout for opening the connection in seconds.
:obj:`None` disables the timeout.
ping_interval: Interval between keepalive pings in seconds.
:obj:`None` disables keepalive.
ping_timeout: Timeout for keepalive pings in seconds.
:obj:`None` disables timeouts.
close_timeout: Timeout for closing the connection in seconds.
:obj:`None` disables the timeout.
reconnect_delays: Delays in seconds between reconnection attempts.
Default is exponential backoff with 5s jitter, capped at 60s.
max_size: Maximum size of incoming messages in bytes.
:obj:`None` disables the limit. You may pass a ``(max_message_size,
max_fragment_size)`` tuple to set different limits for messages and
fragments when you expect long messages sent in short fragments.
max_queue: High-water mark of the buffer where frames are received.
It defaults to 16 frames. The low-water mark defaults to ``max_queue
// 4``. You may pass a ``(high, low)`` tuple to set the high-water
and low-water marks. If you want to disable flow control entirely,
you may set it to ``None``, although that's a bad idea.
write_limit: High-water mark of write buffer in bytes. It is passed to
:meth:`~asyncio.WriteTransport.set_write_buffer_limits`. It defaults
to 32 KiB. You may pass a ``(high, low)`` tuple to set the
high-water and low-water marks.
logger: Logger for this client.
It defaults to ``logging.getLogger("websockets.client")``.
See the :doc:`logging guide <../../topics/logging>` for details.
create_connection: Factory for the :class:`ClientConnection` managing
the connection. Set it to a wrapper or a subclass to customize
connection handling.
Any other keyword arguments are passed to the event loop's
:meth:`~asyncio.loop.create_connection` method.
For example:
* You can set ``sock`` to provide a preexisting TCP socket. You may call
:func:`socket.create_connection` (not to be confused with the event loop's
:meth:`~asyncio.loop.create_connection` method) to create a suitable
client socket and customize it.
* You can set ``ssl`` to a :class:`~ssl.SSLContext` to enforce TLS settings.
When connecting to a ``wss://`` URI, if ``ssl`` isn't provided, a TLS
context is created with :func:`~ssl.create_default_context`.
* You can set ``server_hostname`` to override the host name from ``uri`` in
the TLS handshake.
* You can configure ``ssl_handshake_timeout`` and ``ssl_shutdown_timeout``.
* You can set ``host`` and ``port`` to connect to a different host and port
from those found in ``uri``. This only changes the destination of the TCP
connection. The host name from ``uri`` is still used in the TLS handshake
for secure connections and in the ``Host`` header.
When using a proxy:
* Prefix keyword arguments with ``proxy_`` for configuring TLS between the
client and an HTTPS proxy: ``proxy_ssl``, ``proxy_server_hostname``,
``proxy_ssl_handshake_timeout``, and ``proxy_ssl_shutdown_timeout``.
* Use the standard keyword arguments for configuring TLS between the proxy
and the WebSocket server: ``ssl``, ``server_hostname``,
``ssl_handshake_timeout``, and ``ssl_shutdown_timeout``.
* Other keyword arguments are used only for connecting to the proxy.
Raises:
InvalidURI: If ``uri`` isn't a valid WebSocket URI.
InvalidProxy: If ``proxy`` isn't a valid proxy.
OSError: If the TCP connection fails.
InvalidHandshake: If the opening handshake fails.
TimeoutError: If the opening handshake times out.
"""
def __init__(
self,
uri: str,
*,
# WebSocket
origin: Origin | None = None,
extensions: Sequence[ClientExtensionFactory] | None = None,
subprotocols: Sequence[Subprotocol] | None = None,
compression: str | None = "deflate",
# HTTP
additional_headers: HeadersLike | None = None,
user_agent_header: str | None = USER_AGENT,
proxy: str | Literal[True] | None = True,
process_exception: Callable[[Exception], Exception | None] = process_exception,
# Timeouts
open_timeout: float | None = 10,
ping_interval: float | None = 20,
ping_timeout: float | None = 20,
close_timeout: float | None = 10,
reconnect_delays: Callable[[], Generator[float]] = backoff,
# Limits
max_size: int | None | tuple[int | None, int | None] = 2**20,
max_queue: int | None | tuple[int | None, int | None] = 16,
write_limit: int | tuple[int, int | None] = 2**15,
# Logging
logger: LoggerLike | None = None,
# Escape hatches for advanced customization
create_connection: type[ClientConnection] | None = None,
# Other keyword arguments are passed to loop.create_connection
**kwargs: Any,
) -> None:
self.uri = uri
self.ws_uri = parse_uri(uri)
if not self.ws_uri.secure and kwargs.get("ssl") is not None:
raise ValueError("ssl argument is incompatible with a ws:// URI")
if subprotocols is not None:
validate_subprotocols(subprotocols)
if compression == "deflate":
extensions = enable_client_permessage_deflate(extensions)
elif compression is not None:
raise ValueError(f"unsupported compression: {compression}")
if logger is None:
logger = logging.getLogger("websockets.client")
if create_connection is None:
create_connection = ClientConnection
self.additional_headers = additional_headers
self.user_agent_header = user_agent_header
self.proxy = proxy
self.process_exception = process_exception
self.open_timeout = open_timeout
self.reconnect_delays = reconnect_delays
self.logger = logger
self.create_connection_kwargs = kwargs
def factory(uri: WebSocketURI) -> ClientConnection:
# This is a protocol in the Sans-I/O implementation of websockets.
protocol = ClientProtocol(
uri,
origin=origin,
extensions=extensions,
subprotocols=subprotocols,
max_size=max_size,
logger=logger,
)
# This is a connection in websockets and a protocol in asyncio.
# create_connection defaults to ClientConnection.
connection = create_connection(
protocol,
ping_interval=ping_interval,
ping_timeout=ping_timeout,
close_timeout=close_timeout,
max_queue=max_queue,
write_limit=write_limit,
)
return connection
self.factory = factory
async def open_tcp_connection(self) -> ClientConnection:
"""Open a TCP or Unix connection to the server, possibly through a proxy."""
loop = asyncio.get_running_loop()
kwargs = self.create_connection_kwargs.copy()
proxy = self.proxy
if kwargs.get("unix", False):
proxy = None
if kwargs.get("sock") is not None:
proxy = None
if proxy is True:
proxy = get_proxy(self.ws_uri)
def factory() -> ClientConnection:
return self.factory(self.ws_uri)
if self.ws_uri.secure:
kwargs.setdefault("ssl", True)
if kwargs.get("ssl") is None:
raise ValueError("ssl=None is incompatible with a wss:// URI")
kwargs.setdefault("server_hostname", self.ws_uri.host)
if kwargs.pop("unix", False):
_, connection = await loop.create_unix_connection(factory, **kwargs)
return connection
if proxy is not None:
proxy_parsed = parse_proxy(proxy)
if proxy_parsed.scheme[:5] == "socks":
# Connect to the server through the SOCKS proxy.
sock = await connect_socks_proxy(
proxy_parsed,
self.ws_uri,
# python_socks uses local_addr like asyncio.
local_addr=kwargs.pop("local_addr", None),
)
# Initialize WebSocket connection via the proxy.
_, connection = await loop.create_connection(
factory,
sock=sock,
**kwargs,
)
return connection
elif proxy_parsed.scheme[:4] == "http":
# Split keyword arguments between the proxy and the server.
all_kwargs, proxy_kwargs, kwargs = kwargs, {}, {}
for key, value in all_kwargs.items():
if key.startswith("ssl") or key == "server_hostname":
kwargs[key] = value
elif key.startswith("proxy_"):
proxy_kwargs[key[6:]] = value
else:
proxy_kwargs[key] = value
# Validate the proxy_ssl argument.
if proxy_parsed.scheme == "https":
proxy_kwargs.setdefault("ssl", True)
if proxy_kwargs.get("ssl") is None:
raise ValueError(
"proxy_ssl=None is incompatible with an https:// proxy"
)
else:
if proxy_kwargs.get("ssl") is not None:
raise ValueError(
"proxy_ssl argument is incompatible with an http:// proxy"
)
# Connect to the server through the HTTP proxy.
transport = await connect_http_proxy(
proxy_parsed,
self.ws_uri,
user_agent_header=self.user_agent_header,
**proxy_kwargs,
)
# Initialize WebSocket connection via the proxy.
connection = factory()
transport.set_protocol(connection)
ssl = kwargs.pop("ssl", None)
if ssl is True:
ssl = ssl_module.create_default_context()
if ssl is not None:
new_transport = await loop.start_tls(
transport, connection, ssl, **kwargs
)
assert new_transport is not None # help mypy
transport = new_transport
connection.connection_made(transport)
return connection
else:
raise AssertionError("parse_proxy returned unsupported proxy")
else: # proxy is None
# Connect to the server directly.
if kwargs.get("sock") is None:
kwargs.setdefault("host", self.ws_uri.host)
kwargs.setdefault("port", self.ws_uri.port)
# Initialize WebSocket connection.
_, connection = await loop.create_connection(factory, **kwargs)
return connection
def process_redirect(self, exc: Exception) -> Exception | str:
"""
Determine whether a connection error is a redirect that can be followed.
Return the new URI if it's a valid redirect. Else, return an exception.
"""
if not (
isinstance(exc, InvalidStatus)
and exc.response.status_code
in [
300, # Multiple Choices
301, # Moved Permanently
302, # Found
303, # See Other
307, # Temporary Redirect
308, # Permanent Redirect
]
and "Location" in exc.response.headers
):
return exc
old_ws_uri = self.ws_uri
new_uri = urllib.parse.urljoin(self.uri, exc.response.headers["Location"])
new_ws_uri = parse_uri(new_uri)
# If connect() received a socket, it is closed and cannot be reused.
if self.create_connection_kwargs.get("sock") is not None:
return ValueError(
f"cannot follow redirect to {new_uri} with a preexisting socket"
)
# TLS downgrade is forbidden.
if old_ws_uri.secure and not new_ws_uri.secure:
return SecurityError(f"cannot follow redirect to non-secure URI {new_uri}")
# Apply restrictions to cross-origin redirects.
if (
old_ws_uri.secure != new_ws_uri.secure
or old_ws_uri.host != new_ws_uri.host
or old_ws_uri.port != new_ws_uri.port
):
# Cross-origin redirects on Unix sockets don't quite make sense.
if self.create_connection_kwargs.get("unix", False):
return ValueError(
f"cannot follow cross-origin redirect to {new_uri} "
f"with a Unix socket"
)
# Cross-origin redirects when host and port are overridden are ill-defined.
if (
self.create_connection_kwargs.get("host") is not None
or self.create_connection_kwargs.get("port") is not None
):
return ValueError(
f"cannot follow cross-origin redirect to {new_uri} "
f"with an explicit host or port"
)
# Strip credentials to avoid leaking them to a different origin.
if self.additional_headers is not None:
self.additional_headers = Headers(
(
(key, value)
for key, value in Headers(self.additional_headers).raw_items()
if key.lower()
not in ["authorization", "cookie", "proxy-authorization"]
)
)
return new_uri
async def connect(self) -> ClientConnection:
try:
async with asyncio.timeout(self.open_timeout):
for _ in range(MAX_REDIRECTS):
connection = await self.open_tcp_connection()
try:
await connection.handshake(
self.additional_headers,
self.user_agent_header,
)
except asyncio.CancelledError:
connection.transport.abort()
raise
except Exception as exc:
# Always close the connection even though keep-alive is
# the default in HTTP/1.1 because create_connection ties
# opening the network connection with initializing the
# protocol. In the current design of connect(), there is
# no easy way to reuse the network connection that works
# in every case nor to reinitialize the protocol.
connection.transport.abort()
exc_or_uri = self.process_redirect(exc)
if isinstance(exc_or_uri, Exception):
# Response isn't a valid redirect; raise the exception.
if exc_or_uri is exc:
raise
else:
raise exc_or_uri from exc
else:
# Response is a valid redirect; follow it.
self.uri = exc_or_uri
self.ws_uri = parse_uri(exc_or_uri)
continue
else:
connection.start_keepalive()
return connection
else:
raise SecurityError(f"more than {MAX_REDIRECTS} redirects")
except TimeoutError as exc:
# Re-raise exception with an informative error message.
raise TimeoutError("timed out during opening handshake") from exc
# ... = await connect(...)
def __await__(self) -> Generator[Any, None, ClientConnection]:
# Create a suitable iterator by calling __await__ on a coroutine.
return self.connect().__await__()
# async with connect(...) as ...: ...
async def __aenter__(self) -> ClientConnection:
if hasattr(self, "connection"):
raise RuntimeError("connect() isn't reentrant")
self.connection = await self
return await self.connection.__aenter__()
async def __aexit__(
self,
exc_type: type[BaseException] | None,
exc_value: BaseException | None,
traceback: TracebackType | None,
) -> None:
try:
return await self.connection.__aexit__(exc_type, exc_value, traceback)
finally:
del self.connection
# async for ... in connect(...):
async def __aiter__(self) -> AsyncIterator[ClientConnection]:
delays: Generator[float] | None = None
while True:
try:
async with self as connection:
yield connection
except Exception as exc:
# Determine whether the exception is retryable or fatal.
# The API of process_exception is "return an exception or None";
# "raise an exception" is also supported because it's a frequent
# mistake. It isn't documented in order to keep the API simple.
try:
new_exc = self.process_exception(exc)
except Exception as raised_exc:
new_exc = raised_exc
# The connection failed with a fatal error.
# Raise the exception and exit the loop.
if new_exc is exc:
raise
if new_exc is not None:
raise new_exc from exc
# The connection failed with a retryable error.
# Start or continue backoff and reconnect.
if delays is None:
delays = self.reconnect_delays()
delay = next(delays)
self.logger.info(
"connect failed; reconnecting in %.1f seconds: %s",
delay,
traceback.format_exception_only(exc)[0].strip(),
)
await asyncio.sleep(delay)
else:
# The connection succeeded. Reset backoff.
delays = None
def unix_connect(
path: PathLike | None = None,
uri: str | None = None,
**kwargs: Any,
) -> connect:
"""
Connect to a WebSocket server listening on a Unix socket.
This function accepts the same keyword arguments as :func:`connect`.
It's only available on Unix.
It's mainly useful for debugging servers listening on Unix sockets.
Args:
path: File system path to the Unix socket.
uri: URI of the WebSocket server. ``uri`` defaults to
``ws://localhost/`` or, when a ``ssl`` argument is provided, to
``wss://localhost/``.
"""
if uri is None:
if kwargs.get("ssl") is None:
uri = "ws://localhost/"
else:
uri = "wss://localhost/"
return connect(uri=uri, unix=True, path=path, **kwargs)
try:
from python_socks import ProxyType
from python_socks.async_.asyncio import Proxy as SocksProxy
except ImportError:
async def connect_socks_proxy(
proxy: Proxy,
ws_uri: WebSocketURI,
**kwargs: Any,
) -> socket.socket:
raise ImportError("connecting through a SOCKS proxy requires python-socks")
else:
SOCKS_PROXY_TYPES = {
"socks5h": ProxyType.SOCKS5,
"socks5": ProxyType.SOCKS5,
"socks4a": ProxyType.SOCKS4,
"socks4": ProxyType.SOCKS4,
}
SOCKS_PROXY_RDNS = {
"socks5h": True,
"socks5": False,
"socks4a": True,
"socks4": False,
}
async def connect_socks_proxy(
proxy: Proxy,
ws_uri: WebSocketURI,
**kwargs: Any,
) -> socket.socket:
"""Connect via a SOCKS proxy and return the socket."""
socks_proxy = SocksProxy(
SOCKS_PROXY_TYPES[proxy.scheme],
proxy.host,
proxy.port,
proxy.username,
proxy.password,
SOCKS_PROXY_RDNS[proxy.scheme],
)
# connect() is documented to raise OSError.
# socks_proxy.connect() doesn't raise TimeoutError; it gets canceled.
# Wrap other exceptions in ProxyError, a subclass of InvalidHandshake.
try:
return await socks_proxy.connect(ws_uri.host, ws_uri.port, **kwargs)
except OSError:
raise
except Exception as exc:
raise ProxyError("failed to connect to SOCKS proxy") from exc
class HTTPProxyConnection(asyncio.Protocol):
def __init__(
self,
ws_uri: WebSocketURI,
proxy: Proxy,
user_agent_header: str | None = None,
):
self.ws_uri = ws_uri
self.proxy = proxy
self.user_agent_header = user_agent_header
self.reader = StreamReader()
self.parser = Response.parse(
self.reader.read_line,
self.reader.read_exact,
self.reader.read_to_eof,
proxy=True,
)
loop = asyncio.get_running_loop()
self.response: asyncio.Future[Response] = loop.create_future()
def run_parser(self) -> None:
if self.response.done():
return
try:
next(self.parser)
except StopIteration as exc:
response = exc.value
if 200 <= response.status_code < 300:
self.response.set_result(response)
else:
self.response.set_exception(InvalidProxyStatus(response))
except Exception as exc:
proxy_exc = InvalidProxyMessage(
"did not receive a valid HTTP response from proxy"
)
proxy_exc.__cause__ = exc
self.response.set_exception(proxy_exc)
def connection_made(self, transport: asyncio.BaseTransport) -> None:
transport = cast(asyncio.Transport, transport)
self.transport = transport
self.transport.write(
prepare_connect_request(self.proxy, self.ws_uri, self.user_agent_header)
)
def data_received(self, data: bytes) -> None:
self.reader.feed_data(data)
self.run_parser()
def eof_received(self) -> None:
self.reader.feed_eof()
self.run_parser()
def connection_lost(self, exc: Exception | None) -> None:
self.reader.feed_eof()
self.run_parser()
async def connect_http_proxy(
proxy: Proxy,
ws_uri: WebSocketURI,
*,
user_agent_header: str | None = None,
**kwargs: Any,
) -> asyncio.Transport:
transport, protocol = await asyncio.get_running_loop().create_connection(
lambda: HTTPProxyConnection(ws_uri, proxy, user_agent_header),
proxy.host,
proxy.port,
**kwargs,
)
try:
# This raises exceptions if the connection to the proxy fails.
await protocol.response
except (asyncio.CancelledError, Exception):
transport.abort()
raise
return transport