RE: RFC2518bis: xml:lang (2.6)

> From: Lisa Dusseault [mailto:ldusseault@xythos.com]
> Sent: Thursday, June 27, 2002 8:41 PM
> To: Julian Reschke
> Cc: w3c-dist-auth@w3c.org
> Subject: RE: RFC2518bis: xml:lang (2.6)
>
>
> OK, so it looks like we may need some language explaining that due to
> the definition of xml:lang, it can't appear more than once in the same
> scope.  That's a good simplification.

Well, the *attribute* xml:lang can appear more than once, but each and every
*element* is always only in the scope of one specific xml:lang attribute.

XML: "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." In XPath-spec: the closest xml:lang attribute on
"ancestor-or-self".

> I still think it's wise to say where the xml:lang attribute must appear
> in order for the server to know that it must be stored along with the
> property.  I don't think it's enough to say that any xml:lang whose

It must be stored when it's present. So if xml:lang is in scope (no matter
where), it needs to be stored with the property. I agree that RFC2518 should
be clearer about this.

> scope includes the property value counts.  Some XML parsers could put
> "xml:lang" on the root element of the document by default, even when

No compliant XML parser will place an xml:lang anywhere unless explicitly
asked to do that. Reporting xml:lang on the root element when it wasn't
present in the parsed document clearly would be a parser bug.

> that language value is actually inappropriate to the language value of
> certain properties.  I do not believe the server should store that value

If it's inappropriate where it was sent, then it shouldn't have been sent.
If it hasn't been sent and is reported by the XML parser nevertheless,
that's a bug in the parser.

> in that case.

Julian

Received on Thursday, 27 June 2002 14:51:23 UTC