xmld20 - an XML Schema for d20 gaming systems - Part II
Part II
The time has come, I've decided, to move to XSLT 2.0.
Why? Lately, coding has become more and more complex, with
calculations for ability save DCs, multiple attacks with weapons, and
inheritance of properties. These newer features are becoming more and
more difficult to eke out of XSLT 1.0. I can do it, sure, and have
done so, and it is a bit surprising how used to workarounds you can
get, but now that the project is coming along nicely, with both new
and old stat blocks, I think it's time for a move to a more powerful
processing language.
Also, looking on the Saxon website I found out that they've got .NET
executables, which means that XSLT 2.0 is much more accessible to the
masses; no more having to fight with Java classpaths and such to use
this library.
So far, Saxon hasn't failed me. I've only been doing a handful of
tests, including running old XSLT 1.0 code, which works fine. Testing
out the new features has been going flawlessly as well.
So now what? Now I move the library, stat block and character sheet
generator code to version two, and recode them all. The biggest
addition will be the xsl:function element, so I
can stop making template calls that simulate string-returning
functions, and can now fetch node-sets as return values. This will
clean up a LOT of code that's in the current library and stylesheets.
What else? Little things like having a lower-case() and upper-case() function will cut down on a lot of
unreadability; string-join() and the @separator attribute will definitely improve the
readability of lists, and cut down on a lot of cruft that currently
has to force this readability in.
Things might get a little quiet here while the existing codebase
gets ported over, but I don't expect it to take too long, and we
can return to fighting with the more interesting aspects of monster
and character development.
©2002-2008 Wayne Pearson