You are viewing a plain text version of this content. The canonical link for it is here.
Posted to j-dev@xerces.apache.org by Andy Clark <an...@apache.org> on 2001/08/23 09:34:28 UTC

[Xerces2] Regressions from Merge

There is a number of regressions in the Xerces2 codebase
from the merge. They weren't there before the merge -- I
just checked.

I'm using the OASIS test suite dated 15 Mar 2001. It's
quite old but it enables me to see if any changes I make
cause serious problems. The following is the results:

  Before merge: 26 provisional tests failed
  After merge:  83 provisional tests failed

I'm sure it's probably just some simple thing that
happened during the merge that made the regressions
appear. I'll be looking into it but would appreciate
help.

I was hoping to spin another beta release of Xerces2
tomorrow that fixes a bunch of minor problems I've
discovered as well as add more documentation to the
package. Depending on how well I do on tackling this
regression, I still may be able to do a release. I'll
keep the list up to date on my progress.

-- 
Andy Clark * IBM, TRL - Japan * andyc@apache.org

---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-j-dev-help@xml.apache.org


Re: [Xerces2] Regressions from Merge -- NEVERMIND

Posted by Andy Clark <an...@apache.org>.
I figured out the problem. In my overzealousness in making
changes to the XMLDTDValidator, I introduced a regression
*before* checking the code in. Because I was swapping
between various versions of the source file without
"touching" the file, my Ant build was not picking up the
swapped in file. Therefore, when I thought I was testing
the code from the merge I was in reality checking the
code that I wrote! Oops!

Anyway, I fixed it and we're back to where we were before.
Except now the standard Xerces2 components actually accept
the features and properties that they are documented to
accept. :)

Some problems that I found (and fixed) were: 1) the 
namespace binder was paying no heed to the namespaces
feature and would *always* bind the namespaces even when
it was off. This wasn't a problem previously because this
component was used internally by the "universal" validator
which would determine whether it needed to call out to the
namespace binder. 2) The DTD validator didn't have the
dynamic validation feature implemented. So I fixed that.

-- 
Andy Clark * IBM, TRL - Japan * andyc@apache.org

---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-j-dev-help@xml.apache.org


Re: [Xerces2] Regressions from Merge

Posted by Andy Clark <an...@apache.org>.
Strange. I just tried a "build clean all" again and now it has 
the same results as previous runs -- no regressions. I don't
understand it, really. Strange.

Therefore, disregard my previous message. Sorry 'bout that.

-- 
Andy Clark * IBM, TRL - Japan * andyc@apache.org

---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-j-dev-help@xml.apache.org