db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew McIntyre" <mcintyr...@gmail.com>
Subject Re: Searching the Derby Documentation
Date Fri, 16 Nov 2007 22:28:53 GMT
On Nov 14, 2007 6:25 PM, Rick Hillegas <Richard.Hillegas@sun.com> wrote:
>
> Laura Stewart wrote:
> >
> > I think that what we really need to invest in is changes to the code
> > that generates the docs, so that an index can be created with the
> > output.  Many index entries are already in the DITA files for the
> > documentation and are not exposed because the code that generates the
> > output is not setup to support index generation...
>
> +1 to the idea of generating an index. The old Cloudscape index used to
> be very helpful and I think that a lot of that work has been preserved
> in the existing index entry markers.

DITA 1.4 appears to have fairly complete back-of-the-book index generation:

http://dita-ot.sourceforge.net/SourceForgeFiles/doc/release_14.html

Although I think you're referring to the old Cloudscape global index
which was definitely very useful. I suppose it might be possible with
XSLT and lots of time to combine the XML for each book into one
massive XML and generate a global index with links into the specific
books for the indexterms, at least for the HTML and HTML book formats.
Relative filesystem URLs might not work for PDFs, but I wouldn't rule
it out completely. At any rate, there's no built-in multi-book index
feature in the DITA toolkit as far as I'm aware, so a global index
would be something we'd have to build ourselves, once we migrated to
DITA 1.4 and got back-of-the-book indexes working correctly.

There are two problems with regular indexes, though. One is that there
hasn't been an ASL release of the DITA toolkit since 1.2.2, for
whatever reason, and the back-of-the-book index support for PDFs may
not have been implemented for PDFs generated by the old PDF / FOP
targets that were deprecated with DITA-OT 1.3.1:

http://dita-ot.sourceforge.net/SourceForgeFiles/doc/release_13.html

The current PDF2 targets rely on the Idiom FO plugin (ok, looks like a
BSD derivative although it's got a strange clause at the end) but also
on the RenderX XEP rendering engine, whose free version has a license
that is very restrictive and not open source friendly. So, we might
not get back-of-the-book indexes even if we upgraded to DITA 1.4
without some additional effort on our part.

andrew

Mime
View raw message