Waldek Hebisch <[email protected]> writes:

> Martin Rubey wrote:
>> Waldek Hebisch <[email protected]> writes:
>> 
>> > Attached is a patch that adds support for expected failures to
>> > testing framework.  IMHO currently testing framework has
>> > several problem but this was probably the worst one.  
>> 
>> I'll try not to take this personal.  I always thought that my package
>> was quite an improvement about *no* package
>>
>
> Sure.  Maybe we should say more about achivements (and current
> framework is one of them).  But certainly we need to discuss
> problems, without that it is hard to improve things...

No problem, and thanks for the kudos!

>> I assume that what you add is a grand total over all testcases - or is
>> it a grand total over all testsuites?  Please make this clearer in the
>> output.
>> 
>>    
>> =============================================================================
>>    testsuite | testcases: failed (total) | tests: failed (total)
>>    bugs2009                    0    (19)               0    (69)
>>    unexpected failures: 0
>>    expected failures: 5
>>    unexpected passes: 0
>>    total tests: 69
>> 
>
> I want total of all tests: if one test of a testcase is expected
> failure and say three other pass classifying the whole testcase
> as expected failure would be (IMHO) misleading.  So I want
> summary which ignores testsuite/testcase boundaries.  

OK.  I agree that this was misleading.  I mainly tried to copy Christian
Aistleitner's unittest package for Aldor.  So, the rules where:

* if a test fails, the whole testcase fails (i.e., testcases are a very
  small unit) I guess that "expected" failures don't make a testcase
  fail now, right?

* above that, there are only testsuites.

> Note that my plan is to have 'make check' rule in Makefiles which
> gives a single compact result summarizing all tests.  Since we run
> each test file separately the first step is to get summaries for
> files.  

My assumption was actually that the files are all run in the same image,
and statistics() would then output a grand total for all testsuites,
too.

But that was probably too optimistic.

> Later we can add some (probably awk) script to collect data
> from files and produce final numbers.  My main concern is that numbers
> are easy to parse for programs, I hope that human will have little
> need to look at them.  We can add a line and extra header to make
> clear that the new summary is separate from the old one:
>   ...
>   ==========================================================
>   File summary.
>   unexpected failures: 0
>   expected failures: 5
>   unexpected passes: 0
>   total tests: 69

OK.

Martin
-- 
You received this message because you are subscribed to the Google Groups 
"FriCAS - computer algebra system" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/fricas-devel?hl=en.


Reply via email to