W3C home > Mailing lists > Public > public-rdf-comments@w3.org > May 2013

RE: Format of fragment identifier

From: Markus Lanthaler <markus.lanthaler@gmx.net>
Date: Fri, 31 May 2013 18:26:28 +0200
To: "'Hans Teijgeler'" <hans.teijgeler@quicknet.nl>
Cc: <public-rdf-comments@w3.org>
Message-ID: <014b01ce5e1b$9c24abe0$d46e03a0$@lanthaler@gmx.net>
Cc'ing public-rdf-comments again.


On Friday, May 31, 2013 3:37 PM, Hans Teijgeler wrote:
> Hi Markus,
>  
> One last question, if I may.

Sure

> In the example code you showed, for a /URI a UUID.
>  
> We have fragment identifiers only. Can we use UUIDs there as well? Now
> we add a letter, because UUIDs may start with a number, and we have
> read somewhere, in the impenetrable W3C documentation, that starting
> with a number is not allowed, only with a letter or an underscore.
>  
> What is true?

If you mean relative IRIs consisting of just a fragment identifier (like
#frag) then yes, you can use UUIDs, i.e., start with a number such as in
#1234.

If you are speaking about CURIEs (prefix:suffix) then it depends on the
serialization format. Turtle and JSON-LD e.g. allow it. RDF/XML, which you
used in your examples, doesn't because the XML namespace specification
requires suffixes (LocalPart in the XML namespace spec) to be NCNames which
must start with a letter or an underscore.


--
Markus Lanthaler
@markuslanthaler
Received on Friday, 31 May 2013 16:27:00 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:29:56 UTC