- From: Brett Bieber <brett.bieber@gmail.com>
- Date: Tue, 9 Jan 2007 07:41:08 -0600
- To: "Karl Dubost" <karl@w3.org>
- Cc: www-validator@w3.org
On 1/9/07, Karl Dubost <karl@w3.org> wrote: > Hi Brett, > That's cool. > > Le 4 janv. 2007 à 07:41, Brett Bieber a écrit : > > http://pear.php.net/package/Services_W3C_HTMLValidator/ > > More context from the page > > Description > This package provides an object oriented > interface to the API of the W3 HTML Validator application > (http://validator.w3.org/). With this package you can connect > to a running instance of the validator and retrieve the > validation results (true|false) as well as the errors and > warnings for a web page. > > By using the SOAP 1.2 output format from the validator, you > are returned simple objects containing all the information > from the validator. With this package it is trivial to build > a validation system for web publishing. > > > > Usage example: > > <?php > > require_once 'Services/W3C/HTMLValidator.php'; > > $v = new Services_W3C_HTMLValidator(); > > var_dump($v->validate('http://www.unl.edu/')); > > ?> > > > Have you thought about a cache system? > In a workflow, the validation would occur only if the content of the > page or one of the ancestors has been really modified. To avoid to > request validation of pages or content already validated. > The process you describe is something which I've thought about implementing, but not within the context of my PEAR library. My need for the library is to build an automated validation of pages for departments at a university. We've assembled a QA committee to approve sub-sites which adopt our University's web templates, and the items which can be machine validated will be handled before the QA committee examines a site in detail. To that end, we will build a small system to handle the workflow you describe - but a system like this would be more of an application instead of a re-usable library. (In short, I question that PEAR would approve it to be added to the repository) Caching the results could easily be handled by PEAR's Cache_Lite package, and it wouldn't take much time to add in. This is probably the route I will take when building our automated system. I DO plan on publishing what we build and distributing it through our own PEAR channel (http://pear.unl.edu/). But keep in touch if you or anyone else is interested in an automated site validation system built in PHP. -- -Brett Bieber http:saltybeagle.com aim:ianswerq
Received on Tuesday, 9 January 2007 13:41:27 UTC