[freeciv-data] XML for help texts? (Was: Historical background on wonders)

[Mathias Hasselmann <Mathias.Hasselmann@xxxxxx>]

Jeff Mallatt wrote:

> - Rather than forcing our poor, overworked translators to re-translate
> every Wonder help text, I put your texts in a new entry called
> "historytext".  This also allows for more flexibility in displaying the
> information -- however, for now, I simply changed the code to concatenate
> the help and history texts.
> 
> - Unfortunately, the formatting requirements for text in the rulesets is
> rather strict, and is undocumented.  One thing, each line needs to end with
> a "\" (this also implies no trailing whitespace).  I went through every
> entry, fixing formatting problems.
> 
> - And, I noticed and corrected one content error: the Statue of Liberty
> does not stand on Ellis Island, it's on Liberty (formerly Bedloe's) Island.

Because the statements above,
because my plan to embed a lightweight HTML engine into the help dialog
because the wish to include pictures
and finally because of the idea to create a printable documentation:

Shouldn't we convert the help texts in to a XML format?

By using XSLT stylesheets those pages easily could converted into
XHTML, plain text and LaTeX or PS/PDF.

The major problem I see is how to keep translations synchronized...

Parsing could be done with the lightweight libxml aka. gnome-xml
(http://www.xmlsoft.org/) by Daniel Veillard...

For HTML rendering we could use XmHTML (available for Xt *and* as
GtkXmHTML also for Gtk+) or GtkHTML (for Gtk+ only).

If someone cares about memory footprint we could still have a
ASCII text browser available as compile time option or as dll 
(as far I know autoconf handles platform independent creation
 of them...)

Ciao,
Mathias
-- 
PGP: 1024-Bit DSA: ID 55E572F3, 1024-Bit RSA: ID EAAF7CF1
WWW: http://www.informatik.hu-berlin.de/~hasselma/
     http://www.dominanz-haus.de/