[jdom-interest] XSLT and JDOM

[Dennis Sosnoski]

philip.nelson at omniresources.com wrote:

> > My suggestion for JDOM is that if it is to support efficient
> > XSLT implementation
>
> I think its not ever very efficient to do a transformation by first parsing
> into a tree structure like JDOM so I don't know that this would ever rate
> very highly as a goal of the api.

Perhaps I'm misunderstanding something, but I thought this was how most (if not
all) XSLT processors worked. There are some optimizations possible to apply
transformations and generate the output document in parallel to the processing
of the input document, but these just reduce latency - they still construct a
tree representation of the source document as it is parsed. In theory it should
be possible to avoid this in many cases, but I don't know that anyone has
successfully implemented such an approach yet.

I've seen JDOM suggested as an alternative to DOM representations for XSLT on
other mailing lists in the past, so I'd assumed this was one of the use cases
being considered. It's certainly the most widespread current use of XML document
representations.

  - Dennis