-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Friday 14 June 2002 13:56, Joseph Kesselman/Watson/IBM wrote:
> I'd recommend that you seriously consider calling an extension function
> instead. The downside is that extensions are generally not very portable
> from one implementation of XSLT to another. But they're often _FAR_ faster
> for this sort of string manipulation, and considerably easier to understand
> and to write.
Actually, I do already. The code I gave is actually part of my fallback code,
which I use when running on a processor that does not support my extension
(currently written for Xalan). The best of both worlds (portability and, 90%
of the time, speed)!
What I actually have is a template that tests if it is running on Xalan: if
so, then use the extension function, and if not, fall back to the recursive
templates. The extension function is written in Javascript, and is very
straightforward. I couldn't get <xsl:fallback/> to work for this, so it uses
the output from system-property('xsl:vendor').
- --
Peter Davis
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
iD8DBQE9Cl2wNSZCJx7tYycRAie2AJ4/NCqIo6AqN4JB8KHm3X3g9lDsPACg2wc3
AQY3HLT7bqGkqBp+XyutSqc=
=XHza
-----END PGP SIGNATURE-----
