Re: Turtle LC Draft

On Fri, Jun 22, 2012 at 6:32 AM, Ivan Herman <ivan@w3.org> wrote:
>
> On Jun 22, 2012, at 15:20 , Gavin Carothers wrote:
>
> [skip]
>
>>
>>> - Appendix A says: "The character encoding of the embedded Turtle will match the HTML documents encoding.". Isn't this in contradiction to the fact that Turtle must be UTF-8? Formally, that means a turtle parser cannot just take the content of the <script> element and parse it, it has to make it sure that it is converted into UTF-8 first. Propose: add a remark to A.2 that the content of the <script> element must be converted to UTF-8 before being parsed by a Turtle parser.
>>
>> Not exactly true. If parsing from HTML the parser may need to simply
>> accept a character stream rather than a byte stream. Providing
>> specific instructions on what to do with the encoding feels a bit too
>> restrictive on how it could be parsed. For example a JavaScript parser
>> would likely see the character stream from the DOM and not even care
>> about the encoding. If you used libhtml5 or another DOM parser to
>> parse the HTML the same would also be true.
>
>
>
> I do not want to get into a bike-shed-painting exercise here, the whole section being non-normative anyway. But isn't there a contradiction in the sense that the Turtle document specifies Turtle to be UTF-8, and, according to the current text, the script content may well be, say, UTF-16 'cause it was written in China or, worse, can be one of the non-UTF windows encoding? I guess something has to be said in the text, even if it may be less prescriptive than what I wrote...

Added some language clearing up parsing from DOM vs parsing from bytes.

>
> Ivan
>
> ----
> Ivan Herman, W3C Semantic Web Activity Lead
> Home: http://www.w3.org/People/Ivan/
> mobile: +31-641044153
> FOAF: http://www.ivan-herman.net/foaf.rdf
>
>
>
>
>

Received on Friday, 22 June 2012 17:52:11 UTC