On 1/4/07, jerry gay <[EMAIL PROTECTED]> wrote:
On 1/4/07, Andy Lester <[EMAIL PROTECTED]> wrote:
>
> On Jan 4, 2007, at 11:21 PM, Eric Wilhelm wrote:
>
> >> No. You either have tests that are ordered, or you don't.
> >
> > Stated as if it were some sort of immutable law of the universe!
>
> It is as far as Test::Harness goes. Test::Harness doesn't have any
> sort of idea of what connects tests together.
>
don't we have BEGIN, INIT, and CHECK blocks to deal with this?
You mean, tests have a tlib which they might import their sanity check
from and the test blows chunks if the sanity test notices the state of
the universe isn't copacetic?
use Test::More ...
use sanity;
# tests go here
package sanity;
sanity-check
or die;
END { cleanup ... }
I'd imagine there's some kind of "initialize our test environment by
writing something to a diretory if it doesn't already exist" part that
might exist but I don't know what it looks like and I'd be concerned
about race conditions in the test setup when tests are being run in
parallel.
Josh
