Re: [xsl] complex positioning problem

Subject: Re: [xsl] complex positioning problem
From: Geert Josten <Geert.Josten@xxxxxxxxxxx>
Date: Wed, 03 Nov 2004 14:57:06 +0100
Because more difficult examples are too complex to handle?

Two things:


First, the author-year class (already implemented) is arguably more complex (though a different problem). Being able to get (Doe, 1999a, b) is not easy, but it works.

Second, the algorithm for numbers isn't that hard (in plain English): take a list of distinct db:biblioref/@linkend values and number them based on order of occurrence (unless they should be numbered based on author order in the bib list). I'll take a look at your explanation.

Yes, but we agreed that your sample is too simple and gives wrong impressions, even for the 'simple' case of numbered references. Didn't we?


I don't say the author-year problem is easier. Contrary, I agree it is even more difficult, though quite different.

For the footnote style, I need to identify the first occurrence of every distinct db:biblioref/@linkend value.

You will see that the get-bibref-position templates together with the unique-bibrefs variable gives you a solution. Feel free to make a nice XSLT 2 version out of it.


I really have no programming background, and my time to spend on this limited (I have a day job!). I'm someone with an itch to scratch. XSLT 2.0 has features that are very useful for this sort of thing, and Saxon is a good processor.

Hey, I would expect most members on this list to have at least a part time job. :)


Yes, XSLT 2 is a large improvement over XSLT 1, but as long as clients limit the freedom of programmers like me, I have to fall back on what I can use for the problem at hand. And it is not always possible to use Saxon. :(

I think what I've done so far is difficult to do with 1.0. My tendency is to want to say it might be better to use Python or Perl to do some of the pre-processing I currently do in XSLT 2.0. However, I'm not religious about it: I'd be happy if interested people with more general XSLT skills could help me get the code modified such that it could also work in 1.0 contexts (we had a discussion about this a week ago).

I would prefer languages or parser that can handle grammars, but I get the picture.. ;)


As soon as I have some time to dig into your code, I will see how far one could get with XSLT 1.

Grtz,
Geert

Current Thread