W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > June 2003

Normalization Checking

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Fri, 06 Jun 2003 11:50:33 +0100
Message-ID: <3EE071F9.8070106@hplb.hpl.hp.com>
To: www-xml-blueberry-comments@w3.org
CC: w3c-rdfcore-wg@w3.org

This comment is prompted by:

http://lists.xml.org/archives/xml-dev/200306/msg00052.html


I have had NFC checking for RDF/XML for many months.

The first announcement message was:

http://lists.w3.org/Archives/Public/www-rdf-interest/2002Jul/0012.html

That version 1.5.0 and onwards, checked plain strings, and IRIs for NFC 
compliance, rejecting documents with data not in NFC (in strict mode; in 
default mode a warning is issued).

Because of changed advice from RDF Core based on changed advice from I18N 
WG, the latest version (Jena2preview3) no longer checks IRIs.

The relationship between this and XML 1.1 NFC checking is not direct.

+ The input documents are XML 1.0 (not 1.1).
+ These are analysed according to the RDF/XML syntax.
+ This generates a set of triples.
+ The triples are checked for NFC compliance according to the rules 
suggested by the RDF Core WDs.

I had hoped that there would be an XML 1.1 system that did the NFC 
checking, according to the XML 1.1 CR rules, since then the last step is 
redundant.

I am pretty sure that there are other implementations of RDF/XML that make 
similar checks.


Frankly, integrating NFC with an XML parser is not hard, given public 
domain libraries such as icu4j.

Jeremy
Received on Friday, 6 June 2003 06:50:51 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:57:54 EDT