2022-09-09 21:11:13 +08:00
|
|
|
<!--
|
2023-01-02 20:51:48 +08:00
|
|
|
Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
|
|
|
SPDX-License-Identifier: curl
|
|
|
|
-->
|
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
# WebSocket in curl
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-11-18 16:39:26 +08:00
|
|
|
## URL
|
|
|
|
|
|
|
|
WebSocket communication with libcurl is done by setting up a transfer to a URL
|
2023-02-10 16:32:57 +08:00
|
|
|
using the `ws://` or `wss://` URL schemes. The latter one being the secure
|
2022-11-18 16:39:26 +08:00
|
|
|
version done over HTTPS.
|
|
|
|
|
|
|
|
When using `wss://` to do WebSocket over HTTPS, the standard TLS and HTTPS
|
|
|
|
options will be acknowledged for the CA, verification of server certificate
|
|
|
|
etc.
|
|
|
|
|
|
|
|
WebSocket communication is done by upgrading a connection from either HTTP or
|
|
|
|
HTTPS. When given a WebSocket URL to work with, libcurl considers it a
|
|
|
|
transfer failure if the upgrade procedure fails. This means that a plain HTTP
|
|
|
|
200 response code is considered an error for this work.
|
|
|
|
|
2022-09-09 21:11:13 +08:00
|
|
|
## API
|
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
The WebSocket API is described in the individual man pages for the new API.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
WebSocket with libcurl can be done two ways.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
1. Get the WebSocket frames from the server sent to the write callback. You
|
2022-09-09 21:11:14 +08:00
|
|
|
can then respond with `curl_ws_send()` from within the callback (or outside
|
|
|
|
of it).
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
2. Set `CURLOPT_CONNECT_ONLY` to 2L (new for WebSocket), which makes libcurl
|
2022-10-26 14:59:35 +08:00
|
|
|
do an HTTP GET + `Upgrade:` request plus response in the
|
2022-09-09 21:11:14 +08:00
|
|
|
`curl_easy_perform()` call before it returns and then you can use
|
2022-09-13 15:17:53 +08:00
|
|
|
`curl_ws_recv()` and `curl_ws_send()` to receive and send WebSocket frames
|
2022-09-09 21:11:14 +08:00
|
|
|
from and to the server.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
|
|
|
The new options to `curl_easy_setopt()`:
|
|
|
|
|
2022-09-09 21:11:14 +08:00
|
|
|
`CURLOPT_WS_OPTIONS` - to control specific behavior. `CURLWS_RAW_MODE` makes
|
2022-09-13 15:17:53 +08:00
|
|
|
libcurl provide all WebSocket traffic raw in the callback.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
|
|
|
The new function calls:
|
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
`curl_ws_recv()` - receive a WebSocket frame
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
`curl_ws_send()` - send a WebSocket frame
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
`curl_ws_meta()` - return WebSocket metadata within a write callback
|
2022-09-09 21:11:13 +08:00
|
|
|
|
|
|
|
## Max frame size
|
|
|
|
|
|
|
|
The current implementation only supports frame sizes up to a max (64K right
|
|
|
|
now). This is because the API delivers full frames and it then cannot manage
|
|
|
|
the full 2^63 bytes size.
|
|
|
|
|
|
|
|
If we decide we need to support (much) larger frames than 64K, we need to
|
|
|
|
adjust the API accordingly to be able to deliver partial frames in both
|
|
|
|
directions.
|
|
|
|
|
|
|
|
## Errors
|
|
|
|
|
|
|
|
If the given WebSocket URL (using `ws://` or `wss://`) fails to get upgraded
|
|
|
|
via a 101 response code and instead gets another response code back from the
|
|
|
|
HTTP server - the transfer will return `CURLE_HTTP_RETURNED_ERROR` for that
|
|
|
|
transfer. Note then that even 2xx response codes are then considered error
|
|
|
|
since it failed to provide a WebSocket transfer.
|
|
|
|
|
|
|
|
## Test suite
|
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
I looked for an existing small WebSocket server implementation with maximum
|
2022-09-09 21:11:13 +08:00
|
|
|
flexibility to dissect and cram into the test suite but I ended up deciding
|
2022-09-13 15:17:53 +08:00
|
|
|
that extending the existing test suite server sws to deal with WebSocket
|
2022-09-09 21:11:13 +08:00
|
|
|
might be the better way.
|
|
|
|
|
|
|
|
- This server is already integrated and working in the test suite
|
|
|
|
|
|
|
|
- We want maximum control and ability to generate broken protocol and negative
|
|
|
|
tests as well. A dumber and simpler TCP server could then be easier to
|
2022-09-13 15:17:53 +08:00
|
|
|
massage into this than a "proper" WebSocket server.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
## Command line tool WebSocket
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
The plan is to make curl do WebSocket similar to telnet/nc. That part of the
|
2022-09-09 21:11:13 +08:00
|
|
|
work has not been started.
|
|
|
|
|
|
|
|
Ideas:
|
|
|
|
|
|
|
|
- Read stdin and send off as messages. Consider newline as end of fragment.
|
|
|
|
(default to text? offer option to set binary)
|
|
|
|
- Respond to PINGs automatically
|
|
|
|
- Issue PINGs at some default interval (option to switch off/change interval?)
|
|
|
|
- Allow `-d` to specify (initial) data to send (should the format allow for
|
|
|
|
multiple separate frames?)
|
|
|
|
- Exit after N messages received, where N can be zero.
|
|
|
|
|
|
|
|
## Future work
|
|
|
|
|
|
|
|
- Verify the Sec-WebSocket-Accept response. It requires a sha-1 function.
|
2022-09-13 15:17:53 +08:00
|
|
|
- Verify Sec-WebSocket-Extensions and Sec-WebSocket-Protocol in the response
|
|
|
|
- Make WebSocket work with hyper
|
2022-09-09 21:11:13 +08:00
|
|
|
- Consider a `curl_ws_poll()`
|
2022-09-13 15:17:53 +08:00
|
|
|
- Make sure WebSocket code paths are fuzzed
|
2022-09-09 21:11:13 +08:00
|
|
|
- Add client-side PING interval
|
|
|
|
- Provide option to disable PING-PONG automation
|
|
|
|
- Support compression (`CURLWS_COMPRESS`)
|
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
## Why not libWebSocket
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-11-08 00:23:31 +08:00
|
|
|
[libWebSocket](https://libWebSockets.org/) is said to be a solid, fast and
|
2022-09-13 15:17:53 +08:00
|
|
|
efficient WebSocket library with a vast amount of users. My plan was
|
2022-09-21 05:30:19 +08:00
|
|
|
originally to build upon it to skip having to implement the low level parts of
|
2022-09-13 15:17:53 +08:00
|
|
|
WebSocket myself.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
Here are the reasons why I have decided to move forward with WebSocket in
|
|
|
|
curl **without using libWebSocket**:
|
2022-09-09 21:11:13 +08:00
|
|
|
|
2022-09-19 23:30:30 +08:00
|
|
|
- doxygen generated docs only makes them hard to navigate. No tutorial, no
|
|
|
|
clearly written explanatory pages for specific functions.
|
2022-09-09 21:11:13 +08:00
|
|
|
|
|
|
|
- seems (too) tightly integrated with a specific TLS library, while we want to
|
2022-09-13 15:17:53 +08:00
|
|
|
support WebSocket with whatever TLS library libcurl was already made to
|
2022-09-09 21:11:13 +08:00
|
|
|
work with.
|
|
|
|
|
|
|
|
- seems (too) tightly integrated with event libraries
|
|
|
|
|
|
|
|
- the references to threads and thread-pools in code and APIs indicate too
|
|
|
|
much logic for our purposes
|
|
|
|
|
|
|
|
- "bloated" - it is a *huge* library that is actually more lines of code than
|
|
|
|
libcurl itself
|
|
|
|
|
2022-09-13 15:17:53 +08:00
|
|
|
- WebSocket is a fairly simple protocol on the network/framing layer so
|
2022-09-09 21:11:13 +08:00
|
|
|
making a homegrown handling of it should be fine
|