>> This could be done using META. However, I am experimenting with using
>> an INDEX attribute on elements in the document BODY together with
>> DFN for the first instance of a term. The INDEX attribute is used
>> to define how the element is to appear in an index and supports
>> hierarchical indexes via a syntactic convention using "::" to separate
>> each level in the hierarchy. I want to use this to index the HTML3
>> specification to increase the effectiveness of the Internet Draft.
> Dave, I'd like to follow up on this a bit, in lieu of the recent
> discussions on REV/REL. How does the INDEX attribute fit into the REV/REL
> framework?
With my intended uses, there isn't a direct connection. I merely want to
build a decent printable index for a document generated from some 70 HTML
files. The INDEX attribute is used in a single pass mechanism to create
a sorted list of index entries. My html2text filter actually makes two
passes through the web of files, but this is in order to insert page numbers
into the table of contents and other cross references within the document.
The index could easily be generated as an HTML file for online use for those
people who prefer "manually created indexes" to simple full text search
mechanisms, if this was considered useful.
> How are you handling this in the HTML3 document?
On the second pass, I treat <a rev=toc href=...> by appending a series
of periods and the page number to the end of the output text line
containing the anchor with the rev=toc. This bit works now as you can
see from the HTML3 I-D. My work in progress extends this to support
cross references within the HTML3 spec which will be mapped to
[see page nnn], and the indexing mechanism as previously described.
-- Dave Raggett <dsr@w3.org> url = http://www.hpl.hp.co.uk/people/dsr
Hewlett Packard Laboratories, Filton Road, | tel: +44 117 922 8046
Bristol BS12 6QZ, United Kingdom | fax: +44 117 922 8924