- From: Felix Sasaki <fsasaki@w3.org>
- Date: Fri, 12 Oct 2007 10:22:55 +0900
- To: Yves Savourel <ysavourel@translate.com>
- CC: public-i18n-its@w3.org
Hi Yves,
Yves Savourel wrote:
> Hi Richard,
>
> I'm worried about the edit changes to 'translate flag' in BP5.
> (http://www.w3.org/International/its/techniques/its-techniques.html#DevTransOver)
>
> One does not override 'flags' but rules. To me the idea of 'flags' is associated to making something, in XML, most likely with an
> attribute. But this BP is about overriding rules. If anything here we use 'flags' to override rules and defaults.
>
> I know this is about not using 'translatability'. But to me using 'translate flag' is worst than 'translatability'. 'Translate
> information' maybe, ut does not sound good either.
>
>
> And (unrelated to above) one questions for XPath experts:
>
> In Example 6 we have:
>
> <its:translateRule selector="//*[lang('zxx')] | //@*[lang('zxx')]" translate="no"/>
>
> Is //@*[lang('zxx')] really useful? Since the attributes have the same language as the elements were they are, such case should be
> caught by //*[lang('zxx')] already, no?
>
I think this is not only an XPath question, the answer depends ITS:
since "Translate" information does not inherit to attributes, you need
to select attributes directly in an XPath expression to specify
Translate information for all attributes with the language zxx, i.e.
eventually overriding other global rules for attributes. Hence you need
//@*[lang('zxx')] in the above. The inheritance you are describing
("attributes have the same language as the elements were they are") is
about xml:lang, i.e. it's a different beast.
Felix
Received on Friday, 12 October 2007 01:23:17 UTC