[
Lists Home |
Date Index |
Thread Index
]
From: "Bill de hÓra" <dehora@eircom.net> on LISP
# <irked>
# Yes it is great; you'd do well to take a look at it. It subsumes XML,
# Xpath and XSLT, and in a saner industry, it should have been the
# language you're using for your job; people have been using Lisp for the
# kind of work XSLT+XPATH makes a pita for a long time.
# </irked>
(Flame-redardant: I have used LISP on and off for 17+ years, and worked supporting
LISP based operating systems/hardware for TI for a while, and co-developed a simple
custom LISP for processing SGML in about '89. So I am not antagonistic to LISP.)
But one of the major delights of LISP is that it is so easy to construct little
language parsers and then actually program in that little language.
Anyone using LISP would probably be having exactly the same discussion:
where should XPath stop and let its host language take over?
I think Jeni's concerns are pretty good. XPath does not need to do everything,
because that has a cost in complexity.
Another approach is to build in a profiling system into XPath 2, on the assumption
that everyone may need to customize it for particular uses anyway: XSLT would
like a medium amount, XML Schemas key/unique would like a streamable subset,
Schematron is happy with whatever comes along (as long as that daffy "optimization"
is defeatable).
Cheers
Rick Jelliffe
|
