Xkb api

[Russell Shaw]

Alan Coopersmith wrote:
> Since I work at one of the aforementioned old consortium members
> who already had a Framemaker license, I fired it up and did "save as"
> of the XKB docs into HTML, XML & MIF formats.   Unfortuantely, the
> XML doesn't seem to identify the DTD and appears to be a Framemaker
> specific one, but hopefully between the three you can get closer to
> usable docs.
> 
> I've posted them at:
>     http://people.freedesktop.org/~alanc/xkb/
> 
> The xkbdoc.tar.bz2 contains all three formats, just as you see in the
> other directories.
> 
> I did get a couple warnings about undefined cross-references and unknown
> tags, but a quick glance looks like most of the content was converted - you
> can use the old postscript versions to verify.
> 
> I can do similar conversions for other FrameMaker docs in the specs if 
> anyone is interested in doing anything useful with them.

Hi,
Thanks. The documents are far too large for me to convert manually to
docbook, and i can't find any conversion tool to do it. It would take
me weeks to convert XKBProto.xml. The style sheet is DSSSL, and the dtd
seems to be a proprietory filemaker thing.

There is an extremely large number of empty and unused anchor tags like:

   <A NAME="pgfId-851021"></A>

*Every* entity in XKBProto.xml has one, which makes the corresponding
html take a *long* time to load into a browser or into open-office.
Deleting all these tags made a dramatic speedup.

What would be a suitable file format to store converted documents
in if docbook isn't feasible? I can save the html as open-office .sxw.

Maybe it's just better to forget the whole thing and remember not
to use a proprietory file format next time something is documented.
Atleast the output is available as html and postscript. The html is
really what i was after.