RE: [xsl] DocBook XSL and Internationalization

Subject: RE: [xsl] DocBook XSL and Internationalization
From: "Bill Burns" <bburns@xxxxxxxxxxxxxx>
Date: Mon, 19 Mar 2007 11:16:29 -0600
I realize now that the DocBook list would've been a more appropriate venue
for these questions. My apologies. I will direct any other questions I have
on this topic to that list.

Jirka responded:

> > - Default fonts
> Uhm, don't understand your question

Where fonts might be called out in the XSL, are different default values
predefined for languages that might require other characters than those
available in, say, a font that supports Latin characers. For example, in a
Japanese document, if the output doesn't fully support Unicode (for example,
HTML Help), do the style sheets specify some font that would be appropriate
for JA output? For an Arabic document, does the XSL allow setting defaults
for both Arabic and LTR content? I just want to understand whether fonts are
defined in the XSL in such as way that output in such circumstances happens
by default or needs to be tweaked to ensure proper support.

> > header, and footer word/character order
> Uhm, don't understand your question

Depending on the locale, the placement of autogenerated elements in headers,
footers, xrefs, and other items might vary because of the syntax of a
particular language. For example a page number might occur before, after, or
between two words. Is this something that is already designed into DocBook
XSL, or is this a customization that would have to be provided by a

>> - Chapter, section, item numbering

>Also localized to 59 languages using simple templates. 
>You can override templates yourself. There are some limitations 
>in respect to Chinese needs.

I was thinking of something that might not use Arabic numerals. The only
language I can think of off hand is Arabic. There's also the matter of order
that affects the items mentioned above and might also affect headings and
table-of-contents entries. Clearly the latter is not an issue if a TOC entry
pulls text from a heading that is already in the correct order. Anyway, I
didn't want to leave out the possibility that this might have potential
ramifications for localized output.

> > - Word breaking
> This is handled by formatter.

My understanding was that there are some Java classes that handle this
outside of what AntennaHouse or XEP might provide. Is this not accurate? If
so, would these have to be called from the tool set, or would they be
extensions to one of the formatters?


Bill Burns
Verbum Communications, Inc.
Microsoft Help MVP

Current Thread