|
From: | Benja Fallenstein |
Subject: | Re: [Gzz] Bug in test framework |
Date: | Wed, 23 Apr 2003 14:02:25 +0300 |
User-agent: | Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3) Gecko/20030327 Debian/1.3-4 |
Tuukka Hastrup wrote:
I don't know how to fix a bug we saw today in the test framework - running tests stopped at this:Java stack traces: --------------------------------------------------------------------------- org.nongnu.storm.impl.DirPool.testBlockTmpFile (Test that creating a blocktmpfile out of a block) Traceback (innermost last): File "test.py", line 195, in ? File "test.py", line 154, in main File "test.py", line 175, in run AttributeError: printStackTrace make: *** [test] Error 255
Whoops, fixed in Fenfire but not in Storm. Argh.
I fixed that, but don't understand how the test system should have handled this. Should it maybe handle Java exceptions and Python exceptions as either and not as both?
No, Python code can also throw Java exceptions, and those stacktraces are extremely hard to read because of all the Jython internals exposed to them. OTOH the Python stacktrace doesn't suffice, because if the exception happened in Java code, those bits aren't shown in the Python trace. Thus the system should (and now does) print out Python exceptions as Python exceptions and Java exceptions as both.
- Benja
[Prev in Thread] | Current Thread | [Next in Thread] |