[postgis-devel] CUnit

Paul Ramsey pramsey at cleverelephant.ca
Tue Dec 16 15:51:34 PST 2008


I'll wait 24 hours before hacking into the configure script :) But if
everyone remains high on unit tests, I'll get this framework more
properly integrated.

Question on directory location: right now it's in the root... either
it should be in something like 'tests' with 'regress' or perhaps it
should be under liblwgeom, since it will be 100% liblwgeom related.

P.

On Tue, Dec 16, 2008 at 3:28 PM, Mark Leslie <mrk.leslie at gmail.com> wrote:
> Agreed.  I haven't worked with cunit much at all, so I can't speak to the
> technology, but unit tests are worth their weight in blood, sweat and beers.
>  It would be a huge effort to write tests for all existing functionality,
> but I think it would be reasonable to require new functionality to have
> decent test coverage.  Then we can add tests for existing functionality as
> we fix it.
>
> Hard-coding the path seems like an immediate problem.  Just grabbing cunit
> with apt-get I see that it installed into /usr as I would expect.  If this
> is definable in configure it won't be a problem.
>
> Mark Leslie
> Geospatial Software Architect
> LISAsoft
>
> -------------------------------------------------------------
> Ph: +61 2 8570 5000 Fax: +61 2 8570 5099 Mob: +61
> Suite 112, Jones Bay Wharf 19-21 Pirrama Rd Pyrmont NSW 2009
> -------------------------------------------------------------
>
> LISAsoft is part of the A2end Group of Companies
> http://www.ardec.com.au
> http://www.lisasoft.com
> http://www.terrapages.com
>
>
> Kevin Neufeld wrote:
>>
>> I think having the ability to perform unit tests is, as you say,
>> invaluable.  As the code base grows, it becomes harder to manage / easier to
>> break some portion of the codebase that depends on other portions.  If we
>> can swing it, I vote to introduce the harness.
>>
>> We could add a "make code-check" or something to the main Makefile.
>>
>> -- Kevin
>>
>> Paul Ramsey wrote:
>>>
>>> FYI, I've committed in the early bits of code on my line crossing
>>> project. Most interesting to the development community is the contents
>>> of 'cunit', which is a cunit test harness directly working on
>>> liblwgeom. It was really invaluable for testing the code, I must say.
>>> Good work Mark on breaking out liblwgeom, it has all worked as
>>> advertised so far.
>>>
>>> I'm interested in thoughts on how we should handle cunit tests, or if
>>> we should even have them at all. Right now there is a hardcode
>>> dependency on have cunit installed in /usr/local, but obviously it is
>>> still early days.
>>>
>>> Paul
>>> _______________________________________________
>>> postgis-devel mailing list
>>> postgis-devel at postgis.refractions.net
>>> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>>
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at postgis.refractions.net
>> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>



More information about the postgis-devel mailing list