W3C home > Mailing lists > Public > www-international@w3.org > January to March 2006

Re: question about IRI spec

From: Jeremy Carroll <jjc@hpl.hp.com>
Date: Tue, 10 Jan 2006 18:13:24 +0000
Message-ID: <43C3F944.8010709@hpl.hp.com>
To: Chris Lilley <chris@w3.org>
CC: Martin Duerst <duerst@it.aoyama.ac.jp>, "www-international@w3.org" <www-international@w3.org>, public-iri@w3.org


I think you mint the fragment.

Noting %41 is a percent escape sequence that is legal but inappropriate 
for minting, then:

http://example.org/foo#         does not have the relevant warning
#toto%41                        does
http://example.org/foo#toto%41  does
#toto                           does not
http://example.org/foo%41       does
http://example.org/foo%41#toto  does

Examination of the relative reference and the base URI and the resolved 
URI allow one to see where the problem occurred.

Ditto for say NFKC.

Jeremy



Chris Lilley wrote:
> On Tuesday, January 10, 2006, 4:14:06 PM, Jeremy wrote:
> 
> JC> I think in both RFC 3987 and RFC 3986 there are three effective levels 
> JC> of instruction: MUST force, SHOULD force, and something I am calling 
> JC> minting force (i.e. applied on generation not receipt).
> 
> A meta-question - does the concatenation of a URI with a fragment to
> make a new URI constitute minting?
> 
> In RFC 2396, the fragment was not part of the URI so in theory no new
> URI was generated.
> 
> In RFC 3986 and thus RFC 3987, the fragment is part of the URI.
> 
> So, if http://example.org/foo (or http://example.net/foo# ) already
> exist, does concatenation to produce http://example.org/foo#toto (or
> http://example.net/foo#toto ) constitute minting a new URI/IRI or is it
> re-using an existing one?
> 
> 
> 
Received on Tuesday, 10 January 2006 18:13:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 19:17:06 GMT