Re: [XSLT30] collection lang processing should refer to BCP47 [I18N-ISSUE-400][I18N-ACTION-467]

Addison, and I18N

I have been asked to respond to this comment on behalf of the XSLT Working Group.

Thank you for pointing out the identity of the algorithm for matching the language requested in xsl;sort and the algorithm described in RFC 4647 BCP 47, which I am sure is no coincidence.

The XSL WG has resolved to add a reference to RFC 4647 to the XSLT specification. Because a normative reference would require considerable care in defining exactly what rules we are referencing and exactly what parameters apply, we have decided for avoidance of risk to make this a non-normative reference, but phrased in such a way that users wanting a more precise description of the algorithm than the somewhat informal presentation in the XSLT specification are encouraged to turn to the RFC for further detail.

Michael Kay
for XSLT WG


> On 10 Oct 2015, at 21:10, Phillips, Addison <addison@lab126.com> wrote:
> 
> Hello QT,
> 
> The Internationalization (I18N) WG has been cleaning up our Tracker recently and we found that we inadvertently did not forward you one of our comments during the Last Call period on your document [0]. Although that LC is long closed, I have been asked to forward the comment in case you find it useful and/or are able to address it anyway. The comment is tracked here: [1] and reads:
> --
> In Section 13.1.3 I find:
> 
> --
> The lang attribute indicates that a collation suitable for a particular natural language should be used. The effective value of the attribute must either be a string in the value space of xs:language, or a zero-length string. Supplying the zero-length string has the same effect as omitting the attribute. If a language is requested that is not supported, the processor may use a fallback language identified by removing successive hyphen-separated suffixes from the supplied value until a supported language code is obtained; failing this, the processor behaves as if the lang attribute were omitted.
> --
> 
> This describes a fallback that is identical to RFC4647 Basic Filtering [1] used in BCP47. However, no reference is provided to BCP47.
> 
> I'd suggest explicitly referencing BCP47 here and specifying fallback by reference.
> --
> 
> Best regards (for I18N),
> 
> Addison
> 
> [0] http://www.w3.org/TR/xslt-30/#collating-sequences
> [1] https://www.w3.org/International/track/issues/400
> 
> Addison Phillips
> Principal SDE, I18N Architect (Amazon)
> Chair (W3C I18N WG)
> 
> Internationalization is not a feature.
> It is an architecture.
>

Received on Thursday, 29 October 2015 22:11:00 UTC