Hi, Chad. Thanks for the help.

Chad wrote:
> Right now the only test we run is the parser tests which
> is just a big list of expected input/output for the parser.
> It's by no means complete and only covers the parser.
>   

That's interesting. Some tests are better than no tests, for sure.

Luckily, we have a giant corpus of known-good inputs and outputs, so it 
would be fun to do some mutation testing to find the holes in the parser 
coverage. At a conference yesterday I asked around to see if anybody had 
heard of a mutation testing tool (like http://jester.sourceforge.net/) 
for PHP and regexes, but alas, nobody had.

If somebody gets interested in that, let me know, and I'm glad to dig 
further.

> The stuff in /t and /tests should be ignored...as they're
> wildly out of date and completely unmaintaned.

Good to know. If I get to the point where I have time to write tests, 
I'll ask here further about the best way to go about it.

William

_______________________________________________
Wikitech-l mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Reply via email to