Re: <time> values in HTML5

On Mon, Nov 21, 2011 at 6:46 PM, Tantek Çelik <tantek@cs.stanford.edu> wrote:
> 2. Provide input to the next iteration of such existing type systems (eg XML Schema 1.2). Thus I would not hold back or modify any (even imminent) CR drafts.

I think this is reasonable. It's also worth noting that datetimes are
not the only conceptual datatype whose lexical space (both in the
validity sense and in the ability to process sense) differs between
XSD and HTML5.

I haven't updated this since 2008, so it no longer matches the
implementation, but http://hsivonen.iki.fi/html5-datatypes/ may be of
interest. The implementation is at
https://bitbucket.org/validator/syntax/src/4854426e8897/relaxng/datatype/java/src/org/whattf/datatype

The lexical space of anyURI is all finite sequences of XMLChars. The
lexical space for URLs in HTML5 is narrower, depends on the charset of
the document and depends on the URL scheme (see the html5-datatype
spec above for my attempt to deal with the URL scheme issue).

Booleans in XSD have a lexical space that accept "0" or "1", but ARIA
states that take "true" and "false" don't take "0" or "1":
http://www.w3.org/TR/wai-aria/complete#aria-required (There are also
ARIA states that take more than 2 values two of which are "true" and
"false" and the others aren't "0" or "1"!)

The lexical space for various XSD number types includes leading and
trailing whitespace as valid. However, leading and trailing space
typically isn't valid in HTML5. However, HTML5 processing often
ignores leading space characters and trailing garbage.

There may be other issues of similar nature that I failed to remember right now.

-- 
Henri Sivonen
hsivonen@iki.fi
http://hsivonen.iki.fi/

Received on Monday, 5 December 2011 06:58:59 UTC