Re: [xml-dev] Why Use anyURI

[Michael Kay <mike@saxonica.com>]

 On 15/09/2010 2:29 PM, MURATA Makoto (FAMILY Given) wrote:
>>  In XSD 1.1 the WG 
>> admitted defeat and changed the spec so the lexical space is exactly the 
>> same as xs:string.
> I'm confused. 
You're not the only one.

>  If xsd:anyURI in XSD 1.1 allows everything, why does W3C
> pulish a TR for LEIRI?  

I'm not sure how those two things are related? Are you saying, why
didn't XSD 1.1 choose to make xs:anyURI correspond to the LEIRI
definition? Perhaps because some people want to be even more liberal
than the LEIRI definition, for example permitting Windows filenames, or
the Sun-defined JAR file "URIs" which break all the URI specifications.
> Does xsd:anyURI capture LEIRIs or does it
> capture IRIs? 
Both.
>  Does xsd:anyURI in XSD 1.1 allow the following?
>
> 	The space character
> 	"<" (U+003C), ">" (U+003E) and '"' (U+0022)
> 	Unwise characters "\" (U+005C), "^" (U+005E), "`" (U+0060), "{"
> 	(U+007B), "|" (U+007C) and "}" (U+007D)
> 	The controls (C0 controls, DEL and C1 controls, U+0000 - U+001F
> 	U+007F - U+009F)
> 	The Bidi formatting characters (U+200E, U+200F, U+202H-202E)
> 	Specials (U+FFF0-FFFD)
> 	Tags (U+E0000-E0FFF)
> 	Non-characters (U+FDD0-FDEF, …)
> 	Surrogate code units (U+D800-U+DFFF) 
All of the above provided they are "finite-length sequences of zero or
more characters (as defined in [XML])". That doesn't include unmatched
surrogates or U+0000, and I'd have to check for "non-characters", but I
think it includes all the others.

Michael Kay
Saxonica