Re: [docbook-apps] I18N

[Dave Pawson <davep@dpawson.co.uk>]

On Thu, 26 Aug 2010 10:01:27 -0700
"Bob Stayton" <bobs@sagehill.net> wrote:

> Hi Dave,
> I guess I need some clarification of your terminology here.  When you
> say "odd words", I presume that means the text is mixed language
> text, with foreign words mixed with English text, rather than foreign
> content isolated in elements?

Not even that Bob. It's a school text book, maths. It refers to ....
Rand ....   SA uses , as decimal separator, etc. So it is 'localization'
rather than I18N I guess. I'm thinking about <phrase
lang='en-US'>dollar</   etc, hence  'odd words', rather than text
blocks. 


  And when you say "boilerplate", are
> you referring to the generated text specified in the stylesheet files
> common/*.xml, which is used for generating "Chapter" labels and
> such?  

Yes, in this case 'labels' are not wanted, but the processing model
may be adaptable? 




> 
> With mixed text of odd words, I'm not sure how useful it is to
> separate out the content into separate files.  Certainly marking
> content with <phrase lang="xx"> might be helpful, depending on what
> you need to do with it.

My thought was an identity tranform 'per locale' such that I could
replace the phrases with text 'looked up' in a file similar to
common/*.xml and replace it prior to output format processing.



-- 

regards 

-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk