[Webware-discuss] More thoughts on testing

[Chuck E. <ec...@mi...>]

A friend & colleague of mine and I were discussing testing of web sites. I 
pointed out that I currently scan the HTML output of my web application for 
important substrings, but that in the future I want to do more detailed 
analysis on the contents. For example, if the page has a table of numbers, 
I'd like to be able to work with those numbers to do comparisons, check for 
"greater than zero", etc.

He pointed out that I could reformulate my output as XHTML instead of HTML 
and then with a few lines of code, suck in the output as a Pythonic data 
structure which I could address and traverse.

Sounds pretty good to me. Some questions:

* Has anyone in the Webware or python-web-modules community tried this 
testing approach?

* Does anyone know of a resource describing how various browser versions 
react to XHTML? (I've breezed through the XHTML sites for a couple hours 
today and found everything *except* that.)

-Chuck