Re: DAVENPORT: explanation of indexing

[Norman Walsh <ndw@xxxxxxxxxx>]

/ Mark Galassi <rosalia@xxxxxxxxxx> was heard to say:
| My expectations might be a bit off, because I sort of would expect to
| put <indexterm> tags in my body, and then have Norm's stylesheets
| automatically generate the index at the end of the book for me :-)
| That's the way it works in TeXinfo, where you can say something like
| @cindex{Screen -- physical properties} anywhere you want in the text,
| and the concept index will have a reference to the text in which you
| embedded the @cindex{} call.

That's the way that <indexterm> is supposed to work.  The problem with
Jade is that it's loosely coupled from the actual backends, so there's
no way for Jade to know the page numbers.  Ideally Jade should embed
the terms in the RTF or TeX and let the real processor generate them.
But the RTF backend doesn't support the hidden text fields that would be
required.  (Sebastian, can this be made to work in the TeX backend?)

| But looking at the examples in the DocBook Element reference, I find
| an example of an <index> with no references to a body of text.
| Looking at <indexterm>, I see nothing about the processing
| expectations.

Very few people actually build indexes in SGML because few systems are
tightly coupled with the backend.  I have in mind putting together a
solution for the HTML case, though, if that helps at all... ;-)

                                        Cheers,
                                          norm
-- 
Norman Walsh <ndw@xxxxxxxxxx>      | Those parts of the system that you
http://nwalsh.com/                 | can hit with a hammer are called
                                   | hardware; those parts that you can
                                   | only curse at are called software.


 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist