forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From David Crossley <cross...@apache.org>
Subject Re: Glossary Plugin FOR-755 - was RE: [jira] Updated: (FOR-639) define terminology for the various aspects of Dispatcher
Date Sun, 02 Apr 2006 04:19:59 GMT
Gav.... wrote:
>
> Ok, I have started this, created the basic Glossary output plugin,
> Now need to configure it.
> 
> I have questions (never done a plugin before so bear with me) :-
> 
> I'm referring to the attachments provided on FOR-755 here -
> 
> 1. Does glossary.xmap get renamed to and therefore replace output.xmap ?

As you discovered, it is an input plugin so
glossary.xmap becomes input.xmap
I have not looked at glossary.xmap. Perhaps it
needs some stuff removed. Look at another plugin
for guidance.

> 2.Where do the .dtd and .mod files need to go?
> 3a. I have put glossary2document.xsl in resources/stylesheets, is this
> correct?

> 3b. And will I need to rename it to glossary-to-document.xsl ? 

Yes obviously follow the naming convention.

> 3c. If yes to 3b, anything else to change?

Add to the locationmap.xml ... Again use another
plugin for guidance.

> 4. Glossary.xml is placed in the xdocs directory, this is the main file we
> Need to add our Terminology. Looking quickly, its an A-Z arrangement, can we
> and should we have a separate A-Z for Dispatcher Terminology, therefore a
> separate dispatcher-glossary.xml file, or will ALL Forrest terminology
> reside in the same Glossary.xml file?

No this doesn't happen in the plugin
(apart from a simple example). The real data files
go in whatever project that the user creates.
For example the forrest project will use it
in site-author/content/xdocs/**

I expect that is better to allow separate glossary
data files. They can be any filename because the input.xmap
should respond on the Content Aware Pipeline, i.e. the
source-type action responding on the document type declaration
or namespace.

The listLocations plugins uses a horrid workaround
to match on the filename. It needs to be fixed to
use the publicIdentifier in the XML instance document.

> I am doing this in Whiteboard at the moment, is this ok?

I don't know the answer to this. Needs a separate
discussion.

Wherever you do it is fine at the moment.

> I am not creating samples, I could rename Ross's original Glossary.xml to
> something else and use that as a sample ?

Use lowercase filenames.

No, that datafile is Ross' example. I reckon
create a really trivial example data file.
Say four entries, using very obvious terms, e.g.
HTML = HyperText Markup Language
CSS = Cascading Style Sheets
ASF = Apache Software Foundation
... perhaps use one of those with an example
link to another site.

> Anything else?
> 
> Lots of 'Q''s in there, sorry.
> 
> 1 hour later ...
> 
> Ant Test - Build Failed die to could not validate Glossary.xml, I guess
> that's because the .dtd and .mod files are in the wrong place.

As discussed in the Howto Build a Plugin,
make sure to use Forrest's Ant. Also this error
is probably because you don't have the DTD catalog.xcat
configured properly or using the wrong Public Identifier
for the XML file.

-David

> Gav...
> 
> > 
> > Failing that, i reckon just use an ordinary document
> > with concise definitions that link to examples in
> > the other documents.
> 
> I have started this also, in fact started in when your idea of
> Using the Glossary Plugin came through. Hmm, I wonder which I'll
> Finish first :)
> 
> > However it gets done is good.
> > 
> > -David
> 
> On it,
> 
> Gav...

Mime
View raw message