RE: XForms localization

The primary means for localization that we put into XForms is based on
external instance data and the HTTP "Accept-Lang" header.

For example (namespaces omitted)

<head>
 <model>
  <instance src="my-data.xml" />
  <instance src="labels.xml" id="labels" />
  ...
 </model>
</head>

 <body>
  <input ref="cc/expiry"><label ref="instance('labels')/expiry" /></input>
  ...
</body>

Assume that the my-data.xml XML tree looks like this:
  <order>
    <cc>
     <number/>
     <expiry/>
     <name/>
     <type/>
    </cc>
    ...
   </order>

The labels.xml tree that comes back from the server will depend on what data
the server chooses to provide.
It can look at the "Accept-Lang" header in the HTTP request to determine
what actual resource (file) on the server side to use.
For example, for en-US it might look like this
  <labels>
    <expiry>Expiration Date</expiry>
    ...
  </labels>
and for en-CA it might look like this
  <labels>
    <expiry>Expiry Date</expiry>
    ...
  </labels>
and for fr it might look like this:
<labels>
    <expiry>Date d'échéance</expiry>
    ...
  </labels>

A sprinkling of xml:lang attributes in the server files might help a server
sort out exactly what XML to return in response to the labels.xml query, but
in my experience translations are often produced by different individuals so
combining the language strings into one file doesn't help matters.

If you choose do to so however, you might combine the data together like
this:
  <labels>
    <expiry xml:lang="en-US">Expiration Date</expiry>
    <expiry xml:lang="en-CA">Expiry Date</expiry>
    <expiry xml:lang="fr">Date d'échéance</expiry>
    ...
  </labels>
and then write a server-side expression that uses the following XPath
expression in server-side XSLT to return only the labels matching the
Accept-Lang header:
   /*[xml:lang=$acceptLang]

A more complicated XPath and XSLT template could return all of the matching
languages, in order of preference, and the XForms first-node rule would
return the first matching value in the <label ref=...> expression in the
form, and that would take care of the case where labels are not fully
translated, so there would be a fallback. 

Leigh.

-----Original Message-----
From: T. V. Raman [mailto:tvraman@us.ibm.com] 
Sent: Friday, February 06, 2004 11:40 AM
To: azydron@xml-intl.com
Cc: www-forms@w3c.org
Subject: XForms localization




not that I am aware of --- I think we conciously (or perhaps
unconciously) put anything that is likely to get rendered to the
user as element content and used attributes exclusively 
for "qualifying" things e.g., appearance for controls, styling,
binding etc etc 



>>>>> "Andrzej" == Andrzej Zydron <azydron@xml-intl.com> writes:
    Andrzej> Hi,
    Andrzej> 
    Andrzej> I am a member of the OASIS XLIFF (XML Localization
    Andrzej> File Format) TC. I have been charged with
    Andrzej> investigating the localization issues regarding
    Andrzej> XForms. I have a fairly good grasp of the XForms
    Andrzej> elements that contain text data:
    Andrzej> 
    Andrzej> <label> <hint> <help> <alert> <value> <message>
    Andrzej> <setvalue>
    Andrzej> 
    Andrzej> Are there any XForm attributes that could contain
    Andrzej> translatable text?
    Andrzej> 
    Andrzej> Thanks in advance,
    Andrzej> 
    Andrzej> AZ
    Andrzej> 
    Andrzej> 
    Andrzej> --
    Andrzej> 
    Andrzej> 
    Andrzej> email - azydron@xml-intl.com smail - Mr. A.Zydron 24
    Andrzej> Maybrook Gardens, High Wycombe, Bucks HP13 6PJ
    Andrzej> Mobile +(44) 7966 477181 FAX +(44) 870 831 8868 www
    Andrzej> - http://www.xml-intl.com
    Andrzej> 
    Andrzej> This message contains confidential information and
    Andrzej> is intended only for the individual named.  If you
    Andrzej> are not the named addressee you may not disseminate,
    Andrzej> distribute or copy this e-mail.  Please notify the
    Andrzej> sender immediately by e-mail if you have received
    Andrzej> this e-mail by mistake and delete this e-mail from
    Andrzej> your system.  E-mail transmission cannot be
    Andrzej> guaranteed to be secure or error-free as information
    Andrzej> could be intercepted, corrupted, lost, destroyed,
    Andrzej> arrive late or incomplete, or contain viruses.  The
    Andrzej> sender therefore does not accept liability for any
    Andrzej> errors or omissions in the contents of this message
    Andrzej> which arise as a result of e-mail transmission.  If
    Andrzej> verification is required please request a hard-copy
    Andrzej> version. Unless explicitly stated otherwise this
    Andrzej> message is provided for informational purposes only
    Andrzej> and should not be construed as a solicitation or
    Andrzej> offer.
    Andrzej> 
    Andrzej> 

-- 
Best Regards,
--raman
------------------------------------------------------------
T. V. Raman:  PhD (Cornell University)
IBM Research: Human Language Technologies
Architect:    Conversational And Multimodal WWW Standards
Phone:        1 (408) 927 2608   T-Line 457-2608
Fax:        1 (408) 927 3012     Cell: 1 650 799 5724
Email:        tvraman@us.ibm.com
WWW:      http://almaden.ibm.com/u/tvraman
AIM:      TVRaman
GPG:          http://www.almaden.ibm.com/cs/people/tvraman/raman-almaden.asc
Snail:        IBM Almaden Research Center,
              650 Harry Road
              San Jose 95120

Received on Friday, 6 February 2004 15:27:54 UTC