Although VTD sounds interesting, I immediately hit this alarming comment:
We propose a "non-extractive" tokenization approach that maintains the source document intact in memory.
and go
- turns out their "huge" XML file for
testing was (wait for it....)
"po_huge.xml" ----- 9,907,759 bytes
Yes you can accomplish miracles of speed if you make a few tiny asumptions.....
But if I had to run lots of XPath expressions on a medium size XML document, I would look into it.
Bill