XML Issues

Date: Fri, 30 Jun 2000 14:53:31 +0100
From: Jeff Dalton 
Subject: Re: I-Face: first focus - XML Parser
To: Austin Tate , rhr@aiai.ed.ac.uk
Cc: jussi@aiai.ed.ac.uk, johnl@aiai.ed.ac.uk

> As just discussed Robert, I think this will be a nice early target for the
> I-Face part of the I-Face/I-PE work.  WE need to choose an XML parser we
> can use in our work.  It should be openly usable for any purpose.  WE hope
> we do not have to write our own.

There are a number of Java XML parsers, and some of them, at least,
have licenses that are considered non-restrictive.

There's also the question of whether they validate and whether they'll
be able to understand XML schemas.  (And then there's the DOM etc, and
other things.  The XML world is a large world.)

I don't think it's necessary to use a validating parser at run-time,
though it might of course be desirable.  We could have a separate
validating parser to use when we want to validate any files that are
prepared in advance of use (and debug our DTDs).  That would cover
quite a few cases.

It's still early days for XML schemas, so we may not be able to
use them initially, but I think they're more suitable than DTDs for
describing data.

For some work, we may have to use a specific XML parser, because it's
used by some other things we want to use.  I think our current work
with KAoS may be an example.

Anyway, here are two good starting points:



See also the IMB "developerWorks" page, http://www.ibm.com/developer/xml/
[Note added 11 Aug 00]

The Apache parser, Xerces-J (J for the Java version), looks interesting:


Another parser I'd put on the list is James Clark's XP:


(James Clark is a big name in the XML world, by the way.)

IBM's XML parser (which we're implicitly using in the KAoS work)
is also worth checking out:


Note that:

  Updated: Fri Jun 30 15:32:18 2000
  XML4J 3.0.1 is based on the Apache Xerces XML Parser Version 1.0.3

-- jeff