Add send_timeout, close_timeout, and send_x timeout parameter

[frederikaalund]

What do these changes do?

See the title and the discussion in #2310.

Are there changes in behavior for the user?

No. The old timeout parameter will take precedence over the new close_timeout parameter. The default close_timeout of 10.0 seconds is still in place.

Related issue number

#2310

Checklist

• I think the code is well written
• Unit tests for the changes exist
• Documentation reflects the changes
• If you provide code modification, please add yourself to CONTRIBUTORS.txt
  • The format is <Name> <Surname>.
  • Please keep alphabetical order, the file is sorted by names.
• Add a new news fragment into the CHANGES folder
  • name it <issue_id>.<type> for example (588.bug)
  • if you don't have an issue_id change it to the pr id after creating the pr
  • ensure type is one of the following:
    • .feature: Signifying a new feature.
    • .bugfix: Signifying a bug fix.
    • .doc: Signifying a documentation improvement.
    • .removal: Signifying a deprecation or removal of public API.
    • .misc: A ticket has been closed, but it is not of interest to users.
  • Make sure to use full sentences with correct case and punctuation, for example: "Fix issue with non-ascii contents in doctest text files."

[frederikaalund]

I tried writing a functional test for the send_x calls. However, I failed to do so. Seemingly, the send_x calls return immediately despite my best efforts to stall them. I know from experience that the send_x can stall in practice (#2309) but I'm unable to simulate these conditions in a test setting.

Here is my attempt (that doesn't work):

async def test_send_bytes_timeout(loop, test_client):
    closed = loop.create_future()
    raised = False

    async def handler(request):
        ws = web.WebSocketResponse(send_timeout=0.1)
        await ws.prepare(request)

        try:
            # The send_bytes call returns immediately.
            # I also tried with 1 GB of data to defeat buffering. That didn't work.
            await ws.send_bytes(b'0')
            # Calling drain doesn't help either.
            await ws.drain()
        except asyncio.TimeoutError:
            nonlocal raised
            raised = True

        closed.set_result(1)
        await ws.close()
        return ws

    app = web.Application()
    app.router.add_get('/', handler)
    client = await test_client(app)
    async with client.ws_connect('/') as ws:
        await closed
    assert raised

[codecov-io]

Codecov Report

Merging #2537 into master will increase coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #2537      +/-   ##
==========================================
+ Coverage   97.56%   97.56%   +<.01%     
==========================================
  Files          40       40              
  Lines        8085     8088       +3     
  Branches     1413     1413              
==========================================
+ Hits         7888     7891       +3     
  Misses         74       74              
  Partials      123      123

Impacted Files	Coverage Δ
examples/web_ws.py	95.47% <0%> (+0.04%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 81594ec...9147863. Read the comment docs.

[asvetlov]

Would you write tests for expired timeouts?

Should we use a timeout for ping and pong?
Pretty sure yes.

Also please add send_timeout etc to client_ws.py -- there is no reason to have no such functionality in client web socket as well.

[asvetlov]

Please deprecate timeout.
Something like

if timeout is not None:
    warnings.warn("timeout parameter is deprecated, use close_timeout instead", DeprecationWarning)
    close_timeout = timeout

[asvetlov]

Please describe added params below.

[asvetlov]

@frederikaalund ping

[frederikaalund]

Sorry about the delay. I'll try to find time to implement the changes this week. Thanks for the detailed review.

[asvetlov]

No problem!