Re-enable test in testGetDeductionsModel0 TestOntModel

["Bruno P. Kinoshita" <ki...@apache.org>]

Hi all, 

I had spotted three tests that looked a bit strange while working on the JENA-380 and made a note to look at them later after the jena3 merge.
I've committed two of them
TestOneToManyMap#testConstruct1 (sorry, used the wrong commit message)TestResourceUtils#testRemoveEquiv
These two were simple, just variables that were created but never used in the tests.
But this one:
TestOntModel#xxtestGetDeductionsModel0
Has the following comment from 2008:
     * this test has been disabled for
     * the time being, since it is not correct as written. However,
     * I'm not removing or changing it just yet, since it is showing up
     * an infelicity in the rule engine that Dave will investigate
     * at some future date.

I have re-enabled (by simply renaming it) it locally and it is working now, but I haven't committed it. Would someone else that knows the code base better like to take a look, and perhaps commit / remove the method? I assume the rule engine had a bug some time ago that was fixed in some commit, but the test was never re-enabled.

CheersBruno

Re: Re-enable test in testGetDeductionsModel0 TestOntModel

[Andy Seaborne <an...@apache.org>]

On 04/05/15 11:40, Bruno P. Kinoshita wrote:
> Hi all,
>
> I had spotted three tests that looked a bit strange while working on the JENA-380 and made a note to look at them later after the jena3 merge.
> I've committed two of them
> TestOneToManyMap#testConstruct1 (sorry, used the wrong commit message)TestResourceUtils#testRemoveEquiv
> These two were simple, just variables that were created but never used in the tests.
> But this one:
> TestOntModel#xxtestGetDeductionsModel0
> Has the following comment from 2008:
>       * this test has been disabled for
>       * the time being, since it is not correct as written. However,
>       * I'm not removing or changing it just yet, since it is showing up
>       * an infelicity in the rule engine that Dave will investigate
>       * at some future date.
>
> I have re-enabled (by simply renaming it) it locally and it is
> working  now, but I haven't committed it. Would someone else that knows
 > the code base better like to take a look, and perhaps
 > commit / remove the method?
> I assume the rule engine had a bug some time ago that was fixed in some
> commit, but the test was never re-enabled.
>
> CheersBruno
>
>
>

Hi Bruno,

I don't know how to interpret the comment for certain but it could be 
saying the test is wrong as written but illustrates an rule engine issue 
at the time.  I don't know whether "not correct as written" means the 
test should fail or whether it is not testing properly.

Based on the "not correct as written", I'd leave it not running for now.

That said, after that length of time what was potentially a "bug" may 
now be considered a "feature"!

	Andy