Re: buildpages script for www.gnu.org/software/hurd

[Ognyan Kulev]

Reinhard Mueller wrote:
> * header.html - the header that is included at the very start of every
> English page, including the menu and all such stuff.
> * header.de.html - the header that is included at the very start of
> every German page, including the menu and all such stuff
> * header.??.html - the header that is included at the very start of
> every page in language ??
> * footer[.??].html - the footer of HTML pages

Personally, I'm more keen on template system, having single template 
file for each language.  It's good such template files to have 
information about translation of language names too.  Templates in XML 
and processing with DOM in Perl probably will do the work.

> * <pagename>[.??].html - the content of the page <pagename> in the given
> language, for example whatsnew.de.html or whatsnew.html for English.

AFAIK GNU Web server just shows what is in CVS.  So you don't have 
separate source directory and output directory.  That's why in another 
mail I talked about script that extracts content, title, etc. from 
generated page and regenerates it using template.

> And yes, I really believe you can do *anything* in a shell script :-)

Well, Perl is definately better for text processing stuff :-P  Although 
you do it very well with shell script :-)

I'm not the person who will decide what will go in, so don't think about 
above things as requirements.  I'm not sure if I can show implementation 
of my vision for this week.

Regards,
ogi