XSLT match with regex what's the best current solution

I am working on a suite of scripts that induce structure in
free text and eventually capture fine grained medical information.
I have been using AWK so far, but I am thinking about making
this a process largely of XML transformations. However, since I
must induce XML structure from semi-structured free text I need
some more parsing support. First, regular expressions. I know
there is EXSLT but are regex matches and replaces supported
in SAXON (I love SAXON, so I would prefer using it.)

Also, any ideas of additional parsing tools and their integration
into XSLT would be appreciated. Is there a way of running XSLT
in line-mode and have every line matched against regular
expressions? Well, I suppose so, with a simple sed script I could
first wrap each line into a <line>...</line> tag and then use regex
match on the text node of each <line> element.

Is SAXON easy to extend? I suppose there is some documentation
of SAXON that tells me how to write extensions in Java, right?
Any reason why it would be better to use something other than
SAXON if my platform is Java and I'm not interested in Web stuff
(in which case I would look into the Apache work.)

thanks for your ideas,
-Gunther

--
Gunther Schadow, M.D., Ph.D.                    gschadow@xxxxxxxxxxxxxxx
Medical Information Scientist      Regenstrief Institute for Health Care
Adjunct Assistant Professor        Indiana University School of Medicine
tel:1(317)630-7960                         http://aurora.regenstrief.org