cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Colin Paul Adams <>
Subject Re: RELAXNG grammar for sitemap
Date Fri, 25 Oct 2002 07:18:25 GMT
>>>>> "David" == David Crossley <> writes:

    David> I am not sure which tool you refer to. I had no problems at

rng which uses msv, by the author of msv.

    David> I presume that the draft sitemap DTD is using a workaround
    David> for the namespace lack, by having explicit element names
    David> which contain the "map:" prefix, e.g. <!ELEMENT
    David> map:components ...  This workaround is then reflected in
    David> the generated RNG.  Is that what you are referring to with
    David> your next statement?

That's correct.

    David> I gather that you are suggesting to maintain a canonical
    David> RNG grammar, which we use for system validation. Then from

I am suggesting just that. In my opinion the DTD grammar has no other
value than a starting point for some other grammar.

    David> that canonical RNG we can generate a WXS schema that can be
    David> used by any XML editors that cannot use RNG. That sounds
    David> like the way to go.

I agree.

    David> We should compare the two RNG grammars that are generated
    David> from sitemap-v04.dtd to decide which one to base any
    David> further work upon.

I have already done that. The one you generated is preferable in
several small ways (apart from the fact that it is legal! -
The rng tool creates defines of the form map:name, which is
illegal. The tool you used generates names of the form, which
is ok)

    David> Or perhaps we should really write it
    David> from scratch, just using the generated ones for guidance.

What I shall be doing today (I'm offline during the day, BTW), if I
have time after looking into the sitemap namespaces issue, is to
take the rng grammar you generated (or generate a fresh one from a new
DTD depending on what I find out about sitemap namespaces), and start
changing it to overcome the DTD limitations (first change is to
specify that map:components can be in any order, but at most one -
using optionals nested in interleave - and go on in the same vain,
iterating towards a better grammar).

I think we might one to eventually have a schematron schema AS WELL as
an RNG and WXS grammar. The full validation would then consist of
checking against both schemas, but I'm running too far ahead for now.

    David> Did you see the demonstration that i provided in the
    David> abovementioned thread? It provides an Ant task for
    David> build-time validation. Also Pete Royal followed up with
    David> mention about validation of Configuration objects via
    David> Avalon soon.  

I haven't had time to try it yet, but I downloaded the patches.
Colin Paul Adams
Preston Lancashire

To unsubscribe, e-mail:
For additional commands, email:

View raw message