W3C home > Mailing lists > Public > public-vocabs@w3.org > September 2012

Re: Question about Microdata to RDF Note and lang attribute

From: Gregg Kellogg <gregg@greggkellogg.net>
Date: Wed, 19 Sep 2012 17:19:44 -0400
To: KANZAKI Masahide <mkanzaki@gmail.com>
CC: "public-html-data-tf@w3.org" <public-html-data-tf@w3.org>, "public-vocabs@w3.org" <public-vocabs@w3.org>
Message-ID: <0F95EAB0-41A8-4A24-8C91-DA40167DBABA@greggkellogg.net>
On Sep 11, 2012, at 8:29 PM, KANZAKI Masahide <mkanzaki@gmail.com> wrote:

> Gregg, thanks for clarifying the intention of the microdata to RDF
> spec ([1]). OK, so let me clarify the terms.
> Re: HTML5 spec [2],
> (A) I believe "the language of a node" is described here in the sense
> of CSS3 :lang() selector which "uses the UAs knowledge of the
> document's semantics to perform the comparison" [3].
> (B) On the other hand, "The lang IDL attribute" is, IMHO, the lang
> attribute value of the element itself, in the sense of CSS3 |=
> operator which "only performs a comparison against a given attribute
> on the element" [3]. DOM lang property follows this, too.
> An HTML comsumer needs both (A) and (B), and these are not the same.
> For most spec readers, I think, "set from the lang IDL attribute of
> the property element" (in [1]) means B, i.e. "set from DOM lang
> property value of the element". (BTW, original microdata spec
> described it as "set from the language of the element" [4], which is
> closer to what you say [1]'s intention).

Now that I'm back from vacation, and I can look at this further, I agree that the meaning of the lang IDL attribute is unclear. As I'm about to release an updated working draft anyway, I've changed the language as follows:

The value is a plain literal created from the value with language information set from the language of the property element.
See The lang and xml:lang attributes in [HTML5] for determining the language of a node.

There is similar language preceding this for using a non-conforming <time> element value.


> Thank you for taking time to talk with me.
> cheers,
> [1] http://www.w3.org/TR/microdata-rdf/
> [2] http://www.w3.org/TR/2011/WD-html5-20110525/elements.html#the-lang-and-xml:lang-attributes
> [3] http://www.w3.org/TR/css3-selectors/#lang-pseudo
> [4] http://www.w3.org/TR/2011/WD-microdata-20110525/#rdf
> 2012/9/12 Gregg Kellogg <gregg@greggkellogg.net>:
>> On Sep 11, 2012, at 1:52 AM, KANZAKI Masahide <mkanzaki@gmail.com> wrote:
>>> Gregg, thanks for the quick response.
>>> 2012/9/11 Gregg Kellogg <gregg@greggkellogg.net>:
>>>> The HTML IDL attribute for .lang includes the @lang context of the element,
>>>> including its ancestors. From [2]:
>>>> [[[
>>>> To determine the language of a node, user agents must look at the nearest
>>>> ancestor element (including the element itself if the node is an element)
>>>> that has a lang attribute in the XML namespace set or is an HTML element and
>>>> has a lang in no namespace attribute set. That attribute specifies the
>>>> language of the node (regardless of its value).
>>>> ]]]
>>> I wonder "the language of a node" is different from IDL attribute,
>>> which is separately described at the end of section as
>>> [[
>>> The lang IDL attribute must reflect the lang content attribute in no namespace.
>>> ]]
>>> AFAIK, all modern browsers return "" for the DOM lang value of <cite>
>>> and <span> elements in microdata to RDF example. Isn't this "the lang
>>> IDL attribute of the property element" ?
>> We'll, my interpretation is that the lang content attribute of an element is what's described in the first quote:
>> [[[... That attribute specifies the language of the node ]]], so to determine the language attribute of a node, consider the nearest ancestor element having an @lang attribute. The IDL .lang method then reflects this value. This is certainly consistent with the intention of setting @lang on an ancestor, and having it remain in-scope for descendant nodes until another @lang attribute is introduced. There's also symmetry with xml:base (in the XHTML variety), which is used to influence the lookup of relative URLs.
>> In any case, the microdata-rdf spec makes it clear through examples what the intended behavior is, and the test suite will ultimately provide tests that help validate this.
>> Thanks for the note. If you think some clarification is needed in the normative text, we're prepping an update release for additionalType anyway; however, I think that if there's any mis-interpretation, it's in the HTML spec.
>> Gregg
>>> cheers,
> -- 
> @prefix : <http://www.kanzaki.com/ns/sig#> . <> :from [:name
> "KANZAKI Masahide"; :nick "masaka"; :email "mkanzaki@gmail.com"].
Received on Wednesday, 19 September 2012 21:20:30 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:25 UTC