DOCBOOK: about catalogs...

[D dot Thomas at imb dot uq dot edu dot au (Danny Thomas)]

while this is not actually a DocBook question, the Ch2 of the DocBook book
has the best description of catalogs I've seen, but I just don't have the
understanding to make full sense of it.

I want to validate html pages using nsgmls augmented by a range of
policy-based checks, eg to report tabs used within PRE markup

I have two catalog issues

1) can I use a DOCTYPE line in the catalog to set the system identifier
   used when the html page doesn't declare a doctype ?
   I'm a little confused because Ch2 talks about this applying to
   an SGML document. I'm unclear whether this just means an SGML file
   such as a dtd, or whether it also applies to documents created
   in an SGML defined language, such as HTML

2) as well as validating against the html pages declared doctype, I want
   to also validate against other doctypes, eg I might want to see how
   many warnings get thrown up when tested against the xhtml dtd.
   I'm also parsing the file so can determine the doctype, so one
   approach after determining the doctype is to simply generate a
   catalog mapping the public identifier to the dtd file I want it
   checked against.
   eg  PUBLIC "-//W3C//DTD HTML 4.01//EN"  "path/to/xhtml-dtd"
   I'm wondering if there's a cleaner way to over-ride the doctype?
   eg  PUBLIC "*"  "path/to/xhtml-dtd"
   but the closest to this seems delegate which does support partial
   public id matching, but only to select another catalog


thanks for you indulgence,
Danny Thomas