Re: XInclude language fixup

On Sat, 11 Feb 2023 at 18:01, Norm Tovey-Walsh <[email protected]> wrote:

> Hello world,
>
> Following up on a bug report (only tangentially related), I find myself
> reading https://www.w3.org/TR/xinclude-11/#language again and I’m having
> trouble working out what’s intended (yes, I know I was the editor for
> 1.1, but it was a long time ago and it’s remarkable how much
> specification prose that is “obvious” when you write it and review it
> can lose its obviousness over time).


Is there no inheritance Norm?
From the doc element?

Regards




>
> In particular, I’m trying to work out if/when the value of xml:lang is
> the empty string.
>
> Suppose I have:
>
> <doc xmlns:xi="http://www.w3.org/2001/XInclude"
>      xml:lang="en">
>   <xi:include href="en.xml" fragid="element(/1/1)"/>
>   <xi:include href="de.xml" fragid="element(/1/1)"/>
>   <xi:include href="xx.xml" fragid="element(/1/1)"/>
> </doc>
>
> where en.xml is:
>
>   <chap xml:lang="en"><p>English</p></chap>
>
> de.xml is:
>
>   <chap xml:lang="de"><p>Deutsch</p></chap>
>
> and xx.xml is:
>
>   <chap><p>Something</p></chap>
>
> After XInclude processing, if xml:lang fixup is performed, the first two
> paragraphs are obvious, I think:
>
> <doc xmlns:xi="http://www.w3.org/2001/XInclude"
>      xml:lang="en">
>   <p>English</p>
>   <p xml:lang="de">Deutsch</p>
>   …
> </doc>
>
> But what about the third? I can’t decide if 4.7.6 says it should be
>
>   <p xml:lang="">Something</p>
>
> or if 4.7.6. only admits the possibility that I might have specified
> <chap xml:lang="sp"><p xml:lang="">Something</p></chap> to reset
> the paragraph to, I guess, the “default” value.
>
> My XInclude processor currently produces
>
> <doc xmlns:xi="http://www.w3.org/2001/XInclude"
>      xml:lang="en">
>   <p xml:lang="en">English</p>
>   <p xml:lang="de">Deutsch</p>
>   <p>Something</p>
> </doc>
>
> I’m not too worried about the redundant xml:lang on first paragraph, it
> seems harmless. But should I put xml:lang="" on the last paragraph?
>
> I think the answer is “yes” but I wonder if anyone else feels strongly
> that I’m wrong.
>
> FWIW, I’m also confused about the last paragraph in that section:
>
>   If an xml:lang attribute information item is already present, it is
>   replaced by the new attribute.
>
> How is that ever going to happen? I think that’s just…incoherent.
>
>                                         Be seeing you,
>                                           norm
>
> --
> Norman Tovey-Walsh <[email protected]>
> https://nwalsh.com/
>
> > Change is a dangerous game, but one you have to play.
>
-- 
Dave Pawson
XSLT XSL-FO FAQ.
Docbook FAQ.