Mailman 3 July 2008 - Twisted

[Twisted-Python] trial & ThreadedResolver._cleanup
by Justin Mazzola Paluska 14 Jul '08

14 Jul '08

Hi, I’m writing a bunch of unit tests using Twisted’s Trial. I’ve been trying to chase down some issues with my code leaving the reactor unclean. I think I’ve gotten most of them out of the way, except for one: /usr/lib/python2.4/site-packages/twisted/trial/reporter.py:142: twisted.trial.reporter.BrokenTestCaseWarning: REACTOR UNCLEAN! traceback(s) follow: File "/usr/lib/python2.4/site-packages/twisted/trial/util.py", line 72, in _dispatch getattr(self, "do_%s" % attr)() File "/usr/lib/python2.4/site-packages/twisted/trial/util.py", line 99, in do_cleanPending raise PendingTimedCallsError(s) twisted.trial.util.PendingTimedCallsError: pendingTimedCalls still pending (consider setting twisted.internet.base.DelayedCall.debug = True): <DelayedCall -1219829012 [59.9999289513s] called=0 cancelled=0 ThreadedResolver._cleanup(u'localhost', <Deferred at 0xB74AE46CL>)> When I run my tests individually (one invocation of trial per test), I don’t get any errors. It’s only when I run an entire suite that I get these errors. It looks like this is a reported issue (http://twistedmatrix.com/trac/ticket/2579), but with no fix. Does anyone know a workaround for this issue? —Justin

1 0

[Twisted-Python] stdio.StandardIO and control keys
by mayamatakeshi 13 Jul '08

13 Jul '08

Hello, I've just started learning twisted. I'm trying to build an utility to connect to a tcp port (telnet-like) and exchange messages with an application. I'm using twisted.internet.stdio.StandardIO passing a Protocol to its constructor (I'm tinkering with the dataforward.py example from the twisted book). I'm adding preprocessing to this utility so that I can check what the user has input before sending it to the peer. So far it is working fine, as I just have to override the method Protocol.dataReceived. But I'm trying to add history of previously input strings and for this, I need to check if the user is pressing control keys (basically 'up' and 'down' keys, but alt, shift and friends would be nice too). I'm not sure if this is a twisted related issue, but is it possible to somehow capture these key presses ? (they are not notified to dataReceived) Regards, takeshi

1 0

[Twisted-Python] trial and python 2.5.2?
by Mikhail Terekhov 13 Jul '08

13 Jul '08

Hi, I'm trying to run trial from twisted-8.1.0 on SuSE-11.0 with python 2.5.2 and get the following error (with super and old style classes IMHO). What would be the fix or workaround for this? BTW on SuSE-10.3 with python 2.5.1 everything is OK. Thank you Mike mike@linux-lp:~/python/twisted> tar jxf Twisted-8.1.0.tar.bz2 mike@linux-lp:~/python/twisted> cd Twisted-8.1.0 mike@linux-lp:~/python/twisted/Twisted-8.1.0> ./bin/trial twisted /home/mike/python/twisted/Twisted-8.1.0/twisted/python/reflect.py:377: DeprecationWarning: twisted.flow is unmaintained. return __import__(importName) /home/mike/python/twisted/Twisted-8.1.0/twisted/scripts/test/test_mktap.py:16: DeprecationWarning: mktap and related support modules are deprecated as of Twisted 8.0. Use Twisted Application Plugins with the 'twistd' command directly, as described in 'Writing a Twisted Application Plugin for twistd' chapter of the Developer Guide. from twisted.scripts.mktap import run, getid /home/mike/python/twisted/Twisted-8.1.0/twisted/enterprise/util.py:19: DeprecationWarning: twisted.enterprise.util is deprecated since Twisted 8.0.0. category=DeprecationWarning) /home/mike/python/twisted/Twisted-8.1.0/twisted/test/test_persisted.py:161: DeprecationWarning: twisted.persisted.marmalade is deprecated from twisted.persisted import marmalade /home/mike/python/twisted/Twisted-8.1.0/twisted/words/im/__init__.py:8: UserWarning: twisted.im will be undergoing a rewrite at some point in the future. warnings.warn("twisted.im will be undergoing a rewrite at some point in the future.") Traceback (most recent call last): File "./bin/trial", line 24, in <module> run() File "/home/mike/python/twisted/Twisted-8.1.0/twisted/scripts/trial.py", line 361, in run test_result = trialRunner.run(suite) File "/home/mike/python/twisted/Twisted-8.1.0/twisted/trial/runner.py", line 780, in run suite = TrialSuite([test]) File "/home/mike/python/twisted/Twisted-8.1.0/twisted/trial/runner.py", line 263, in __init__ super(TrialSuite, self).__init__([suite]) TypeError: super() argument 1 must be type, not classobj Exception exceptions.TypeError: TypeError("'NoneType' object is not callable",) in <bound method Popen3.__del__ of <popen2.Popen3 instance at 0x933a78c>> ignored mike@linux-lp:~/python/twisted/Twisted-8.1.0> mike@linux-lp:~/python/twisted/Twisted-8.1.0> python --version Python 2.5.2 mike@linux-lp:~/python/twisted/Twisted-8.1.0>

2 2

[Twisted-Python] Weekly Bug Summary
by exarkun＠twistedmatrix.com 12 Jul '08

12 Jul '08

1 0

[Twisted-Python] domish.Element comparison not working
by Gabriel Rossetti 11 Jul '08

11 Jul '08

Hello everyone, I need to compare two domish.Element objects in a unit test using trial. I can't compare the string xml (using .toXml()) because sometimes the attributes are not in the same order, and since it's a string it uses normal string comparison and you see the problem, but I can't use direct comparison either, so what should I use? If you look at the example below you will see what I mean : >>> from twisted.words.xish import domish >>> root = domish.Element(('', 'test')) >>> s = domish.Element(('', 'feed'), attribs={'name':'myname'}) >>> s.addContent("toto") 'toto' >>> root.addChild(s) <twisted.words.xish.domish.Element object at 0x848390c> >>> root.toXml() u"<test><feed name='myname'>toto</feed></test>" >>> >>> root2 = domish.Element(('', 'test')) >>> s = domish.Element(('', 'feed'), attribs={'name':'myname'}) >>> s.addContent("toto") 'toto' >>> root2.addChild(s) <twisted.words.xish.domish.Element object at 0x84a342c> >>> root2.toXml() u"<test><feed name='myname'>toto</feed></test>" >>> >>> root == root2 False As you see, root and root2 are equal in the sense that they contain the same xml structure, same identifiers and the same values, but the "==" comparison says they are false. Am i doing something wrong or is the comparison just not implemented for those objects? Thank you, Gabriel

2 3

[Twisted-Python] twisted and functional programming idioms
by Rodney D Price 10 Jul '08

10 Jul '08

I'm just getting started with Twisted, and trying to find my way around. I've done quite a bit of functional programming (Haskell), where recursion is a very common idiom. I keep finding myself heading in that direction as I try to understand how to use Twisted. For example, if I wanted to write a loop in Haskell that woke up once a second and did something, I might write main = heartbeat heartbeat = do sleep 1000 doSomething heartbeat This code first sleeps for one second, then does something, then calls itself. (The Haskell compiler knows how to keep the stack from overflowing here, in case anyone is wondering.) So if I wanted to do something similar in Twisted, my first inclination is to use callLater, def heartbeat(reactor): doSomething() reactor.callLater(1, heartbeat, reactor) reactor.run() But then my head starts hurting, because I keep thinking that I have to set up all the callbacks before I start the reactor. This code sets callbacks at runtime. Will that be a problem? -Rod P.S. I know there's probably a function somewhere in Twisted to do this already, but I'm not asking how to do this particular thing. I'm really asking if thinking in terms of recursion is a useful Twisted idiom.

3 2

[Twisted-Python] AMP compliance tests
by Jonathan Lange 10 Jul '08

10 Jul '08

Hello all, Where can I find a set of automated tests that check that a particular AMP implementation is complete? I seem to recall someone on #twisted mentioning that they were working on such a suite. thanks, jml

2 2

[Twisted-Python] OSX Leopard version still at 8.0.1?
by Rodney D Price 09 Jul '08

09 Jul '08

Is there a problem with 8.1.0 on Leopard (OSX 10.5)? I ran trial twisted on 8.1.0 and got some errors, but being new at Twisted, I'm not sure whether the errors were real or I just did something wrong. I'm primarily interested in twisted.word XMPP support. -Rod

3 2

[Twisted-Python] Bloody Twisted VFS
by Andy Gayton 09 Jul '08

09 Jul '08

Hey guys, I'm really keen to get twisted.vfs to a stable point that is acceptably releasable. There's 6 tickets to knock off to get it to this state: This ticket will give an easily runnable vfs server, which'll be useful for experimenting with vfs as it is worked on: * http://twistedmatrix.com/trac/ticket/2821 This ticket ports the backend vfs interface to be firstly async, and also, not quite so lame: * http://twistedmatrix.com/trac/ticket/2815 Then we can port the 3 currently supported protocols over to use the new backend interface: * http://twistedmatrix.com/trac/ticket/2816 * http://twistedmatrix.com/trac/ticket/2817 * http://twistedmatrix.com/trac/ticket/2818 And finally once everything is ported, there can be a good clean up: * http://twistedmatrix.com/trac/ticket/2819 Exarkun also has some good suggestions for #2815. I've gone ahead and fixed the ones mentioned - but this is defining an interface - which isn't something to lock into lightly. As he says: 'What these interfaces should actually say is still an open question, I think. This would probably benefit from some real-time conversation with various interested parties involved. You've probably already noticed IFTPShell which is one take on this interface (it's not exactly the same shape, but many of the ideas apply here).' is very reasonable. Just before the first .au sprint I put a bit of time into this - and conch's IConchUser interface seemed to be a better base for vfs than IFTPShell - but this is something that could use more discussion. I've quit my job in LA that's sucked up all my life for the last 2 or so years and I'm now on a road trip around the US. I've hauled up in vermont for the week to catch up on a bunch of stuff - top of the list is to try and get vfs up to scratch finally. I'm planning to be in the Massachusetts area next week - again for around a week. If anyone is available to meet up and hash this out, that'd be awesome. It feels like if we can hit #2815 on the head, the rest will fall into place. Therve has been awesome at chugging through previous vfs stuff - porting protocols to use the agreed upon backend interface should be straight forward. cheers, Andy.

5 23

[Twisted-Python] deferToThread in unittest (trial)?
by Gabriel Rossetti 09 Jul '08

09 Jul '08

Hello, I have a quick question, are there any known side-effects/problems/limitations with using deferToThread in unit-tests (trial)? I'm actually testing a class that uses a class that has a call to deferToThread, the problem is that the callbacks are never called, even if the function passed to deferToThread does nothing (so it should return). Thanks, Gabriel

1 0