Mailman 3 November 2002 - Twisted

[Twisted-Python] Continuing problem with woven templateFile
by Alex Nov. 18, 2002

Nov. 18, 2002

When using templateFile for either a View or a self-templating Widget, I am unable to include a directory name. So, templateFile='tmpl-framework.html' works fine, but templateFile='tmpl/framework.html' generates a traceback that says the file doesn't exist -- even though both are in the same proper location on the server. Earlier, putting in a path would result in a different error; Twisted.web would complain that it couldn't create '.tmpl/framework.pxp', rather than creating 'tmpl/.framework.pxp'. I'm not sure if that problem has been fixed or if it's simply been displaced by the one above. Either way, it's odd that template files should not be accessible anywhere other than a site's base document path. -- Alex Levy WWW: http://mesozoic.geecs.org Tel: 617.835.0778 "Never let your sense of morals prevent you from doing what is right." -- Salvor Hardin, Isaac Asimov's _Foundation_

2 1

[Twisted-Python] New multicast support in Twisted -- please comment
by Itamar Shtull-Trauring Nov. 17, 2002

Nov. 17, 2002

I've just checked in multicast support into CVS. reactor.listenMulticast/reactor.connectMulticast, otherwise it's just like the new UDP stuff, except the transport has a new set of methods documented in twisted.internet.interfaces.IMulticastTransport. As with the UDP API, I've never actually used multicast. Any comments from people who actually have used this in the past or can test this or whatever would be very welcome. -- Itamar Shtull-Trauring http://itamarst.org/ Available for Python, Twisted, Zope and Java consulting ***> http://VoteNoWar.org -- vote/donate/volunteer <***

1 0

[Twisted-Python] MultiServiceTestCase.testDeferredStopService failure.
by Andrew Bennetts Nov. 16, 2002

Nov. 16, 2002

In test_internet, MultiServiceTestCase.testDeferredStopService is failing for me with Python 2.1. The problem appears to be that the test assumes a particular order for the callback results, but the object is internally holding them in a dict so that order isn't guaranteed. So is the test or the code buggy? Can someone familiar with how MultiService should work look at this? Thanks. -Andrew.

2 1

[Twisted-Python] stupid case sensitive web clients
by Hamish Lawson Nov. 16, 2002

Nov. 16, 2002

[This time with a subject line!] Jeff Epler wrote: >Bob Ippolito wrote: > > > I have a stupid device that's talking to a Twisted web server that > > demands HTTP headers to be capitalized Like-This, however Twisted only > > capitalizes the first character.. >[...] > > Any chance someone could patch that up? > >Doubtful. > >http://www.w3.org/Protocols/rfc2616/rfc2616-sec4.html#sec4.2 > "Field names are case-insensitive" Nevertheless the headers are almost always presented in the HTTP spec and other reference materials in the style 'Content-Type'. I would guess therefore that this style will be the one most widely adopted by those broken clients that expect a particular capitalisation. Hamish Lawson

3 2

[Twisted-Python] tap-relative filenames?
by Paul Moore Nov. 15, 2002

Nov. 15, 2002

I'm not even sure I know how to ask this question properly, so please bear with me... I'm writing an application in Twisted (a news server, serving mbox files as read-only groups, FWIW). I'm including a simple interface to dynamically add new files (groups). No problem, but I'd rather keep things as relocatable as possible. Ultimately, I'll have a tap file, which I'll use to start the application using the twistd daemon. So what I'd like to do is store filenames relative to the directory which contains the tap file. OK, so I can cd to that directory, and start twistd from there. But in the more general case, I may start the app from an arbitrary directory and give the full path to the tap file. So what I'm saying is, can I find the name of the tap file from within the application code? I imagine this is non-trivial, because the code won't always even be started from a tap file (my test rig does reactor.listenTCP/reactor.run manually) but when it isn't then names won't be getting persisted, so I don't care (I guess...) Can anybody help? Does anyone even follow the question??? :-) Thanks, Paul -- This signature intentionally left blank

4 4

[Twisted-Python] (no subject)
by Hamish Lawson Nov. 15, 2002

Nov. 15, 2002

Jeff Epler wrote: >Bob Ippolito wrote: > > > I have a stupid device that's talking to a Twisted web server that > > demands HTTP headers to be capitalized Like-This, however Twisted only > > capitalizes the first character.. >[...] > > Any chance someone could patch that up? > >Doubtful. > >http://www.w3.org/Protocols/rfc2616/rfc2616-sec4.html#sec4.2 > "Field names are case-insensitive" Nevertheless the headers are almost always presented in the HTTP spec and other reference materials in the style 'Content-Type'. I would guess therefore that this style will be the one most widely adopted by those broken clients that expect a particular capitalisation. Hamish Lawson

1 0

Re: [Twisted-Python] WebDAV support
by Christopher Armstrong Nov. 15, 2002

Nov. 15, 2002

Hey Daniel, the link to MoulDAVia on your website is broken. How's it coming along? I'd like to check it out. -- Christopher Armstrong << radix(a)twistedmatrix.com >> http://twistedmatrix.com/users/radix.twistd/

2 1

[Twisted-Python] web.static patch waiting for commit (before 1.0.1)
by Paul Boehm Nov. 14, 2002

Nov. 14, 2002

the patch i submitted for pre-1.0 because to fix query_string's for index.rpy's (see list archive), hasn't been commited nor rejected so far. testcase is: http://www.twistedmatrix.com/, repeatedly adding ?bla=foo strings and wondering about the redirect that discards the query_string. anyone care to look at this?: diff -u -r1.74 static.py --- static.py 12 Nov 2002 13:00:06 -0000 1.74 +++ static.py 13 Nov 2002 19:31:41 -0000 @@ -267,10 +267,12 @@ # If someone is looking for children with a PathReferenceContext, # the request won't have a prepath, and we shouldn't do this kind # of mangling anyway because it has already been done. - if hasattr(request, 'postpath') and not request.postpath and request.uri[-1] != '/': + if hasattr(request, 'postpath') and not request.postpath and \ + string.split(request.uri,'?')[0][-1] != '/': return Redirect(request) if os.path.exists(childPath): - if hasattr(request, 'postpath') and not request.postpath and not self.getIndex(request): + if hasattr(request, 'postpath') and not request.postpath and \ + not self.getIndex(request): return widgets.WidgetPage(DirectoryListing(self.path)) ##

4 6

[Twisted-Python] missing woven patch
by Paul Boehm Nov. 13, 2002

Nov. 13, 2002

donovan, the patch you mailed me while twistedmatrix.com was lagging still hasn't made it's way into cvs. please apply: diff -u -r1.36 view.py --- view.py 31 Oct 2002 03:29:08 -0000 1.36 +++ view.py 13 Nov 2002 16:29:45 -0000 @@ -296,6 +296,11 @@ model = self.getTopOfModelStack() if not view or not isinstance(view, widgets.Widget): view = widgets.DefaultWidget(model) + + view.modelStack = self.modelStack[:] + view.viewStack = self.viewStack[:] + view.controllerStack = self.controllerStack[:] + if not controller: controller = input.DefaultHandler(model) controller.parent = self.controllerStack[0]

1 1

[Twisted-Python] pickle apparently doesn't go with curry ...
by Steve Waterbury Nov. 13, 2002

Nov. 13, 2002

Twisted Gurus: I'm using the "curry" recipe from the Python Cookbook (pretty much verbatim), and it appears to work fine in my code, but cPickle isn't happy about it: Traceback (most recent call last): File "/usr/local/bin/mktap", line 31, in ? run() File "/usr/local/lib/python2.2/site-packages/twisted/scripts/mktap.py", line 180, in run a.save() File "/usr/local/lib/python2.2/site-packages/twisted/internet/app.py", line 508, in save dumpFunc(self, f) File "/usr/local/lib/python2.2/site-packages/twisted/internet/app.py", line 490, in dumpFunc _dump(obj, file, 1) cPickle.PicklingError: Can't pickle <function curried_function at 0x83a2aa4>: it's not found as __main__.curried_function ------------------ Here is the recipe I'm using: def curry(*args, **create_time_kwds): """ Currying is a way to bind arguments with a function and wait for the rest of the arguments to show up later. You "curry in" the first few parameters to a function, giving you a function that takes subsequent parameters as input and calls the original with all of those parameters. Nick Perkins' latest version of curry-by-closure (from c.l.p, for the record) It has curry-from-left (for positional args) and call-time-dominates (for named args) semantics """ func = args[0] create_time_args = args[1:] def curried_function(*call_time_args, **call_time_kwds): args = create_time_args + call_time_args kwds = create_time_kwds.copy() kwds.update(call_time_kwds) return func(*args, **kwds) return curried_function ---------------------- Any ideas? Cheers, -- Steve.

8 14