[m-rev.] Fix coverage monitors in Morphine

Erwan Jahier Erwan.Jahier at imag.fr
Tue Mar 5 20:52:03 AEDT 2002


On Tue, 5 Feb 2002, Erwan Jahier wrote:

I will commit that change posted a month ago.

BTW, I can't ``cvs diff -u'' anymore. I vaguely remember a message posted
on a mercury list saying that the cvs network stuff have changed in some
ways, but from what I can read on

http://www.cs.mu.oz.au/research/mercury/information/developers/remote_cvs.html

nothing seems to have changed. So quid ?

> --
> Estimated hours taken: 10
>
> Branches: main, release
>
> Fix the monitors that were computing coverage rates in Morphine.  The
> problem was the following. In order to check that two successes and
> one failure occur for a multi predicate, I was looking at two exits
> and one fail; but this is wrong since, of course, the two exits can
> be produced by different calls, and since all multi predicates ends
> up with a fail event. To get it right, I need to associate the call
> number to exit and fail events. To do that, I maintain at exit ports
> the list of call numbers; when an exit event occurs, I consider it as
> covered iff the current call number is in the list. On the contrary,
> I consider a failure as covered at a fail port iff the current call
> number is not in the list. This also holds for semidet and nondet
> procedures.
>
> extras/morphine/non-regression-tests/queens.exp:
> 	Update the new expected outputs.
>
> extras/morphine/source/call_site_cov.in:
> extras/morphine/source/pred_cov.in:
> 	Change the code so that it does what it is supposed to do.
>
> 	Also use a map instead of a list to store what has to be
> 	covered for each procedure.
>
> extras/morphine/source/generate_pred_cov.m:
> extras/morphine/source/generate_call_site_cov.m:
> extras/morphine/source/coverage_util.m:
> 	Generate an initialize/1 predicate that uses map instead of lists.
>
> extras/morphine/source/coverage.op:
> 	Coverage monitors now output assoc lists.
>
>

--
R1.


--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------



More information about the reviews mailing list