Re: ITS Inheritance

Hi Yves all,

I made a mistake / forgot s.t. in my last mail.

On Tue, 24 Jan 2006 12:26:58 +0900, Felix Sasaki <fsasaki@w3.org> wrote:

> On Tue, 24 Jan 2006 08:34:43 +0900, Yves Savourel  
> <ysavourel@translate.com> wrote:
>
>>
>> Just some thoughts about inheritance:
>>
>> The current specification does not describe precisely how in situ  
>> values are inherited, especially for attribute. Is it working
>> exactly the same way as xml:lang or differently?:
>
> I would say as xml:lang :
> "The intent declared with xml:lang is considered to apply to all  
> attributes and content of the element where it is specified, unless  
> overridden with an instance of xml:lang on another element within that  
> content."

Of course we define default behavior for each data category, see the table  
at  
http://www.w3.org/International/its/itstagset/itstagset#datacat-description  
. For translatability, this behavior is
"Textual content of element, including  content of child elements, but  
excluding attributes"

I would propose to have a subsection 4.1. "default section and inheritance  
for data categories", which encompasses everything in sec. 4. before the  
current subsection 4.1, and in addition a paragraph on inheritance:

"There is no interplay between default inheritance and XPath expressions  
in selector attributes. That is, an XPath expression like
<its:documentRule datacat="yes" datacatSelector="//*"/>
should always select the textual content of the child elements, but not  
the attributes. So if the data category would be "directionality", we  
would need an XPath expression
<its:documentRule dir="ltr" dirSelector="//* | //@*"/>
to get the default behavior. If the data category is "translatability",  
the XPath expression would be
<its:documentRule translate="yes" translateSelector="//*"/>

What do you think? An alternative would be to say that *there is* an  
interplay between inheritance and XPath expressions in selector  
attributes, but IMO that would be very messy.
Another comment below.

> That is,
>
>>
>> Example 1:
>>
>> <p its:translate="yes" alt="text">text <img alt="text"/> text</p>
>>
>> This makes translatable the content of <p> but the values for both alt  
>> atr strill not transltable, right?
>
> no, the attributes would be translatable as well, if we follow xml:lang

this was wrong, since I forgot the default behavior. Sorry for that :(

My other comment on the issue of handling inheritance, if we don't have  
alternative values (see below) still makes some sense, I think. What do  
you all think? I guess this is not a very technical issue, but a question  
of what users would prefer.

>
>>
>>
>> Example 1:
>>
>> <p its:translate="yes" translateSelector="@alt" alt="text">text <img  
>> alt="text"/> text</p>
>>
>> This makes the alt attribut of <p> translatable, but not the values of  
>> the alt attribute in <img>, right?
>
> yes.
>
> There is another issue with inheritance:
> - its:translate="yes" - in situ - is overridden by its:translate="no" -  
> in situ at a child element.
> - what should happen with e.g. its:Locinfo, e.g.
> <p its:locInfo="info no 1">text <img its:locInfo="info no2">  
> alt="text"/> text</p>
> As for the <img> element, does both "info no 1" and "info no 2" apply,  
> or only one of these? I guess the general question is: how to handle  
> inheritance if we don't have *alternative* values, that is s.t.  
> different  from its:translate, its:dir etc. Should the information of  
> the its:locInfo attribute be additive or alternative?
>
Cheers,
Felix

Received on Tuesday, 24 January 2006 13:22:54 UTC