[mapserver-dev] cunit / RFC 12

[Paul Ramsey]

Tamas,

The msautotest suite seems optimized to run the whole mapserver
executable over a test set of data. And I agree there's a huge need
for that kind of testing ("integration testing" I think is the term).
I want to run tightly focussed calls against the actual functions that
make up my driver: if I send in this string, does it parse right? does
the base64 decoder work on this case? ("unit testing", I think). It
doesn't look like msautotest does this (though, a cunit element could
be added to it, so that msautotest turns over the cunit tests, but it
would still be a cunit element).

Paul

On Mon, Nov 3, 2008 at 6:45 AM, Tamas Szekeres <szekerest at gmail.com> wrote:
> Paul,
>
> How about adding these tests to the msautotest suite? As far as I
> remember to the related discussion, the most substantial conclusion
> was that we wouldn't like to maintain various kind of testing
> environments in parallel.
> For more information about the msautotest topic see:
> http://mapserver.gis.umn.edu/development/Tests/regression_testing
>
> Best regards,
>
> Tamas
>
>
>
> 2008/11/3 Paul Ramsey <pramsey at cleverelephant.ca>:
>> Devs,
>>
>> As I work through the postgis driver, I'm wanting to test individual
>> components of it (data defn parsing, SQL generation, geometry parsing,
>> etc), without invoking a whole Mapserver run... sort of like unit
>> testing. I might just write a wee program to do that, but I was
>> wondering if a more formal approach is worth pursuing?
>>
>> I see RFC 12 brought cunit to the plate, but that's as far as it went.
>> Certainly, getting coverage of the code base this late in the game is
>> going to be hard/impossible, but for new functionality, it has to be
>> useful to have some unit tests in place as one works, no?
>>
>> http://mapserver.gis.umn.edu/development/rfc/ms-rfc-12/
>>
>> Any interest / disinterest in pursuing this?
>>
>> Paul
>> _______________________________________________
>> mapserver-dev mailing list
>> mapserver-dev at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/mapserver-dev
>>
>