[XML-SIG] Something's definitely wrong...

[ssteiner at idc]

I'm using the xmlrpc MultiCall class pretty heavily in an application and every time I've somehow caused an error in the xmlrpclib.py code, I get an exception trying to raise the exception.

Here's the most recent example: 

grouped = grouper(2, tuple(mc_result))
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1001, in __getitem__
raise Fault(item['faultCode'], item['faultString'])
Fault: <Fault 1: "<type 'exceptions.TypeError'>:'NoneType' object is unsubscriptable">

I could understand if I was making some sort of coding error, then getting an error, but the Fault object itself, whenever it is used, seems to be generating some sort of new error either through its use, or, as above, the code attempting to construct the Fault object fails.

The thing that's confusing about the above error is that there's a check, just before the line of code that fails, that explictly checks whether the type of "item" == type({}): so the code shouldn't even get called with "NoneType".

This all smells horribly like corrupt memory so I'm hoping someone on this list has seen symptoms like this and can point me in the right direction to start tracking this down.

This, btw, is a pretty simple, maybe 200 line pure Python app that is using the XML-RPC interface, SQLite, and zlib functions.

I'm wondering if an earlier, uncaught error is actually setting this up but it's only happened about 4 times, at completely unpredictable times, and doesn't seem to be repeatable.

Definitely in line for best bug report ever.

Anyone seen this specific symptom, regardless of cause?

S

Stephen F. Steiner
Integrated Development Corporation
ssteiner at integrateddevcorp.com
www.integrateddevcorp.com 
(603)433-1232




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/xml-sig/attachments/20100120/3d22721f/attachment.htm>