Re: [docbook-apps] Re: Request For Clarification: Indexterm processingin auto-index generation.

[Dave Pawson <davep@dpawson.co.uk>]

Cavicchio_Rob@emc.com wrote:
> Dave Pawson [mailto:davep@dpawson.co.uk] wrote:
> 
>> My basic point is that the author should identify duplicates
>> (perhaps from viewing the presented format) and do something about it.
> 
> 
> I disagree with this, because it forces the author to presume a specific
> output format when creating the source. 

That provided by docbook XSLT?


If you have a long section, you
> might want to index the first and last paragraphs of it with the same
> index term. When a PDF is generated, those paragraphs could easily fall
> on two different pages and thus should have two entries in the index.

The issue is (AFAIK) html. fo output follows print pretty well I think.


> But when HTML topics are generated, you simply want to collapse these
> two entries into one and point it to the topic.

You may, others clearly don't. Choose the xml:id of the closest
ancestor section to obtain this?


> 
> I think the transforms should be smart enough to collapse identical
> entries in HTML topics. However, if there is only one instance of a
> particular index term in a topic, then I agree that it should go
> directly to the anchor point rather than the top of the page.

Which is becoming almost a programming language for indexing?

They can't be collapsed if they point to quite different
points in the document? Or at least it wouldn't make
sense to me to do that.





regards

-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk