W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > July to September 2009

Re: Exact format for XML Literals?

From: Bijan Parsia <bparsia@cs.man.ac.uk>
Date: Mon, 14 Sep 2009 11:44:38 +0100
Message-Id: <6B9E40C1-528B-404D-936C-8D7AA7AD6DC9@cs.man.ac.uk>
Cc: Ivan Herman <ivan@w3.org>, "Seaborne, Andy" <andy.seaborne@hp.com>, Axel Polleres <axel.polleres@deri.org>, W3C SPARQL Working Group <public-rdf-dawg@w3.org>
To: Bijan Parsia <bparsia@cs.man.ac.uk>
On 14 Sep 2009, at 09:40, Bijan Parsia wrote:
[a lot of nonsense]

Sorry, I had massive me/Google fail and ended up on an earlier  
working draft of the RDF concepts doc wherein the lexical space was  
*not* restricted to canonicalized forms.

So there's no contradiction...just the weird case that you can have  
syntactically legal RDF/XML documents who's, er..., literal form  
contains parseType=literals which are syntactically not in the  
lexical form (*but*, are not mal formed!)

This makes it easy to miss that the lexical form is wrong (in yet  
another way...malformed literal handling is similarly error hiding --  
I'm not against error hiding per se, but these are quite nasty ways  
of hiding).

It also means that you could easily ignore the parsing requirement to  
canonicalize without anyone noticing for quite some time: Your  
*input* doesn't have to be canonical and your *serialization* doesn't  
have to be canonical.

We could really fix it. If not this, then what? When?

Cheers,
Bijan.
Received on Monday, 14 September 2009 10:40:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:08:27 GMT