At 11:12 AM 6/7/2005 -0700, Heikki Toivonen wrote:
Phillip J. Eby wrote:
> I've also investigated the run()/stop() issue, and as far as I can tell
> this should be a non-issue for tests that run using the Twisted default
> reactor (which is what Chandler uses). The reason that you aren't
That's strange, because I was testing with default reactor in a command
line test app. I just retested and the symptoms are that the (unit test)
program ends up in infinite loop where it gives timeout errors every now
and then, and does not respond to Ctrl+C. This is under Cygwin.
Can you send or point me to the test app in question? The default reactor
normally disables Ctrl-C (but the testreactor won't, as long as you never
call reactor.run() directly), but an infinite loop should only happen if
you don't have anything running that will stop the reactor.
> reactor.waitFor(deferred, timeout) -- iterate until the deferred
Since I haven't needed to use deferred's in my trivial test cases I'd
need to figure out how to call this.
Well, if all you want to do is have some code do something and then stop
the reactor, I could make two methods: reactor.waitFor(seconds) and
reactor.waitUntil(deferred,timeout). Then, you could just call waitFor()
and let your tests do reactor.stop() when done.
By the way, if you are doing assertions in code that is called by the
reactor, this may be a bit of a problem; the reactor expects to log all
such errors, and doesn't pass them on anywhere. So, you would definitely
need for all your assertions to be in your test method itself (i.e. *after*
the waitFor/waitUntil)
> Heikki, does that sound about like what you need for your tests?
From the API description it looks like it would do what I need, with the
caveat that I need to figure out how to deal with deferreds.
But I think it should go into twisted proper. I have a need for this in
M2Crypto unit tests, Grant could probably use this for zanshin unit
tests etc.
I was thinking we'd get it implemented and tested a bit more before
contributing it, but certainly contributing it would be a good idea.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "Dev" mailing list
http://lists.osafoundation.org/mailman/listinfo/dev
