Ran all the testcases I could think of.
The good news is that for the testcases themselves, we pass regression
testing; same number of Pass-cases/Pass-checks.
The bad news is when testing the new code, the "scan" target is
malfunctioning, complaining that
[java] ERROR: 'file:///home/keshlam/git/xalantest/FailScanner.xsl:
line 25: Variable 'redirectFilename' is multiply defined in the
same scope.'
That's an incorrect error message. So something extremely subtle is
different about the ant versus maven Xalan/Serializer builds.
(Is a regression only seen when analyzing regression tests a regression?
Unfortunately, I think so...)
I'll try to provoke this under debugger for analysis. But probably not
for a few days, so if anyone's brave enough to try poking at it please
do so!
(I also discovered that I left a few debugging printouts in the
Mavenized cup/lexx code. I'll fix that now. Advantages of having an
eyecatcher unique to the individual; if something says GONK I know it's
my fault.)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@xalan.apache.org
For additional commands, e-mail: dev-h...@xalan.apache.org