[pmwiki-users] Anyone using DITA and Pmwiki together?

[John Rankin]

On Friday, 10 February 2006 12:35 PM, Bellave Jayaram <bellavejayaram at cox.net> wrote:
>John,
>	Thanks. I want to work at the DITA schema level and not the DTD
>level. Also, I would like to use Java or Python to write the mapping rules
>and writer code. Any hints on where I should start? Of course, I will take a
>look at what exists in DITA-OT that I could leverage but on the Pmwiki side,
>I am clueless. I don't think there is any model (UML, OWL) for Pmwiki, or is
>there?

There are a number of things you might like to think about. I speak with
a pmwiki user's perspective.

We know that pmwiki can generate xml output that validates against at
least 2 different DTDs -- xhtml and the wikibook print xml. It is an
open question whether it can generate xml that will validate against
a possibly more stringent xml schema. If however, the DITA DTD and
DITA schema are just 2 different expressions of the same structural
rules, there should be no problem.

PmWiki is almost markup agnostic. That is, it can take a wide range
of input markups and transform them to a wide range of output markups.
By substituting a local version of scripts/stdmarkup.php an 
administrator can change both the input markup recognised and the
output markup produced. This is a very powerful capability.

So to produce DITA XML from wiki pages, there are at least 3
possible approaches:
- train pmwiki to produce DITA XML -- this is the approach we
  took for Publish PDF, we trained pmwiki to produce print XML

- let pmwiki continue to produce xhtml and use xslt to transform
  the xhtml to DITA XML -- there may well be an existing xslt 
  script that does this

- take the raw wiki markup and write a script to transform it
  to DITA XML -- this looks to me like a non-trivial exercise
  of no obvious benefit if either of the above approaches can
  be made to work

You might want to study how some of the cookbook recipes
have added to or changed pmwiki's markup interpretation. In
essence, you tell pmwiki what input to match and what output
to produce; it takes care of the rest.

Whether wiki markup will be sufficient to represent the kinds
of DITA documents you wish to produce will depend on your
specific requirements. It is likely that you will find the
wiki markups map to a DITA subset. For more information see
http://www.pmwiki.org/wiki/PmWiki/MarkupMasterIndex

Hope this helps.
>
>	I was hoping not to have to introduce new Pmwiki markups. This is
>kind of still in the brainstorming stage so if anyone would like to
>collaborate on this, please contact me - <ID> bellavejayaram <Domain> cox
>dot net.
>
>Jayaram
>


-- 
JR
--
John Rankin