...It might have been discussed before, but I still have this thought anyway:
Isn't it possible to use our nice crawling feature of Cocoon to check
each link? So, basically we could start Cocoon as a servlet, start the
crawler and the crawler reports every link (URI) that doesn't produce
a valid response (throws an exception). I'm not sure, but that should
only be a bunch of lines and would imho help a little bit.
WDYT?
I was thinking about a similar thing which would list all test links and allow us to make a list of what's been tested or not before a release.
But you're right, just checking that no links return an error page would be an improvement already.
Note that some samples (in "system pages" IIRC) do return exceptions by design, you might need to be able to define these special cases.
-Bertrand
