W3C home > Mailing lists > Public > www-validator@w3.org > September 2001

Re: Machine readable format for validation results

From: Gannon Dick <gdick@gte.net>
Date: Sun, 30 Sep 2001 11:25:49 -0500
Message-ID: <000f01c149cc$91fb3500$ef89e041@home2>
To: <www-validator@w3.org>
>The XML format introduced at <URL:http://valet.webthing.com/xml/>
>and implemented at the Site Valet has now been checked in to OpenSP
>via CVS, and so is a likely future for w3-validator too.  If you have
>any comments on it, now would be a good time to make them.

While I'm familiar with SP, I generally use the MS parser.  Slower and unclean, I know, I know ...
Anyway, it will take me some time to get up to speed.

Could you fix a couple of irritants though?
1) The <html> tag is not closed it's <html .
2)a Put a PUBLIC declaration on the sp: namespace.  My SYSTEM is not your SYSTEM :)
2)b The sp: namespace needs a xmlns:sp="someurl" attribute.

While a command line interface for the OpenSP (of SP) tool set is a very good start, automation will require that you are able to run batches with one HTTP submission.  That is the point (one anyway) behind wrappers such as SOAP.  Returning an XHTML document is not as elegant as returning the same wrapper, but may be just as good provided that the batch data is grouped so that a local transform can access individual DOM elements.  Any error messages should be individual XML elements or tree fragments, for the same reason.  To make your output as useful as possible, you need to make re-building a DOM as simple as possible.  Analysis of the batched return can thereby ignore the ordering inherent in a SAX event parser like XT.

 Anyway, like I said, I'll need to work on it.


Received on Sunday, 30 September 2001 12:20:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:13:59 GMT