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 jo...@us.ibm.com on 2002/01/24 23:08:48 UTC

For what it's worth....

I've confirmed that someone is still sending 
 (CR) characters into the
SAX handler's characters() call under some circumstances. I have not yet
isolated a stand-alone illustration of the problem for easy diagnosis;
attempting to do so now.

In at least some of these calls, _only_ the CR; is being sent in... as if
someone had misunderstood newline normalization and was converting
everything to CR rather than to LF, or as if normalization was not being
done at all and the input contained only CRs.

Unplugging Xerces2 and plugging in Xerces1 reliably makes the problem
vanish, so it's either a bug introduced or a behavior changed which is
exposing an old bug in our code.

When I find out more, I'll report more. Just wanted to make sure folks knew
there's still a concern in this area.


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