IDN test: system identifiers

Hi, I've created three test pages using IDNs in a "system identifier".

1 - That works as expected with the W3C validator:
http://hmdmhdfmhdjmzdtjmzdtzktdkztdjz.googlepages.com/IDN-XML-test.html

2 - That fails as expected with the W3C validator (unencoded UTF-8 IRI):
http://hmdmhdfmhdjmzdtjmzdtzktdkztdjz.googlepages.com/IDN-IRI-test.htm

3 - That fails with the validator (%-encoded UTF-8 system identifier):
http://hmdmhdfmhdjmzdtjmzdtzktdkztdjz.googlepages.com/IDN-XML-test.htm

IMO it's no validator bug, but a bug in the XML 1.0 specification.  It
should say something in the direction of "a system identifier is an IRI
encoded as URI following the procedure specified in RFC 3987 (or its
successor)".

 Frank

Received on Thursday, 1 November 2007 13:45:31 UTC