Xerces and ignorable whitespace

[Morten Jorgensen <mo...@ireland.sun.com>]

Guys,

I am trying to make XSLTC's DOM builder a bit faster. I was hoping
that I could remove the whitespace text-node handling, as the SAX2
specs requires a validating parser to pass all ignorable whitespace
nodes through the ignorableWhitespace() SAX entry point. I tested
this by monitoring all text nodes passed to the characters() SAX
entry point, and I recevied loads of text nodes containing blanks
and line-feeds only.

How is this handled in Xalan? Do you rely on the parser to tell
you which text-nodes are WS and which are not? Or maybe I've
misunderstood thay mean by whitespace text in SAX2?

Morten