[rt.cpan.org #96291] t/08taint.t fails on perl 5.20.0

Fri Jun 20 23:56:25 2014: Request 96291 was acted upon.
Transaction: Correspondence added by SISYPHUS
       Queue: Inline
     Subject: t/08taint.t fails on perl 5.20.0
   Broken in: 0.55
    Severity: (no value)
       Owner: Nobody
  Requestors: [email protected]
      Status: open
 Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=96291 >


Inline-0.55_01 was released a couple of days ago, partly to see how t/08taint.t fares following the latest round of amendments to the Inline code.

I've just received the first FAIL report for 0.55_01:
http://www.cpantesters.org/cpan/report/a3185cf8-f82f-11e3-919b-572b4803b899

The culprit is, of course, C/t/08taint.t.

What to do wrt this ?

If it's fixable, and we can fix it, then I guess we should do that.
If we can't fix it, then we need to detect in advance that the failure will occur and either SKIP or TODO the remaining 9 tests.

There's a couple of issues with taking the TODO route:

1) The script is actually dying before all tests have been run - so I don't think TODO'ing the remaining tests is going to prevent the script from being reported as a FAIL. (We can probably workaround that by running the remaining tests inside eval{}.)

2) 3 of the tests are warnings_like() from Test::Warn, and it seems that TODO is ignored for them. That being so, we need to work around that shortcoming, too. (I was unable to find a way to make those 3 tests honor TODO - but that doesn't necessarily mean it can't be done.)

However, I think that if there's no chance of Inline being able to run correctly under -T on a particular system, then we are quite entitled to SKIP those tests on that system.

Thoughts ?

Cheers,
Rob