On Friday 20 October 2006 12:47, Henry S. Thompson wrote:
Further to the thread from August [1] on this subject, I've produced a
trivial proof-of-concept [2], in XSLT 1.0, which may be useful as a
stopgap until Erik completes his much more ambitious version [3].
Interesting.
I've written an XInclude implementation and is currently working on(in the
long term) on XSL-T 2.0. I haven't at all thought about implementing XInclude
in XSL-T, but I now see several advantages:
* Safe. It would use the XSL-T engine and therefore not be coded in relatively
low-level code(and the security/bug hazards it brings, be it Java or C++).
* Could be fast(er). Since the XSL-T "machinery" is used it could be that
existing caching etc is used.
* Should be a lot more simpler implementation.
It will be interesting to look at Erik's more ambitious project. If it
attempts to support for example the xpath10() scheme I can imagine there will
be a noticeble compile-time performance impact to transform it all into XSL-T
for a two-pass transform -- or other features that might make a "native"
XInclude implementation interesting.
Cheers,
Frans
--~------------------------------------------------------------------
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
To unsubscribe, go to: http://lists.mulberrytech.com/xsl-list/
or e-mail: <mailto:xsl-list-unsubscribe(_at_)lists(_dot_)mulberrytech(_dot_)com>
--~--