Problems testing w/ seconds at runtime

[clemcrock]

Hello,

I constantly run into this problem when trying to test assertions
against time w/ seconds. Since there can be a tiny lapse time between
when a test is fired off and the value is tested your test can be off by
a second, therefor causing it to fail.

For instance. I have an assert select test that will randomly fail
because the dynamic time being tested against will have lapsed a second.

assert_select "pubDate", 6.days.from_now.gmtime.strftime("%a, %d %b %Y
%I:%M:%S GMT")

Now, about 50% of the time, this test will fail and give you a result
like such:

<"Sun, 11 Nov 2007 21:15:48 GMT"> expected but was
<"Sun, 11 Nov 2007 21:15:47 GMT">.

Any ideas on how I can make this test less brittle?

Thanks,
Eric

 

[stevexff]

Maybe you could convert the returned time to seconds since the epoch (or some other standardised numeric value), then check that it was less than (now + 6 days + 5 seconds) and more than (now + 6 days - 5 seconds). This would add a bit of tolerance to your test...

Steve

[small]"Every program can be reduced by one instruction, and every program has at least one bug. Therefore, any program can be reduced to one instruction which doesn't work." (Object:erlDesignPatterns)[/small]