What can I parse abiword with?

It seems I cannot parse abiword documents with ONSGMLS.

Attributes id and xid are given type ID, but they are all numeric,
including the first character. When I add the numeric characters
to NAMESTRT in the sgml declaration, ONSGMLS pukes. If I change
the (local) awml.dtd so the type is CDATA, of course ONSGMLS will
parse it, but then it does not check id and xid for uniqueness.

The reason I want to do this, is I want to filter essentially plain
text documents through a perl script to produce an abiword document.
(So that I can compose in a text-mode editor such as JOE, as the
abiword GUI is practically unusable, and I have to build abiword
templates by hand anyway. The abiword format is, of course, very
brittle, and abiword offers no diagnostics when it rejects a document.
It would be helpful to parse the candidate documents first to eliminate
the obvious errors.)

What will parse abiword documents correctly and throw meaningful
error messages?

-- 
Lars Eighner
http://www.larseighner.com/index.html
8800 N IH35 APT 1191 AUSTIN TX 78753-5266
-----------------------------------------------
To unsubscribe from this list, send a message to
abiword-user-request@abisource.com with the word
unsubscribe in the message body.