[Gdal-dev] Re: GDAL 1.3.1 Alpha 2 Released
Charlie Savage
cfis at interserv.com
Sun Oct 2 14:39:36 EDT 2005
I just meant having all the tests in one place, i.e., gdalautotest. For
Ruby, I just created a Ruby directory and then ported some of the Python
tests to Ruby. But I think you should write the tests that are most
appropriate for the language.
I tend to agree with you that I'd like the tests in the GDAL CVS project
(makes them more obvious), but I think that a matter of debate
(integrated tests or separate tests) without a clearly better answer.
Charlie
Ari Jolma wrote:
> Frank Warmerdam kirjoitti:
>
>> On 10/1/05, Charlie Savage <cfis at interserv.com> wrote:
>>
>>
>>> Check out gdalautotest from CVS. There is a test-suite in Python, and a
>>> partial one in Ruby (its slowly growing).
>>>
>>> Ari, maybe this is where the Perl tests should go to stay consistent?
>>>
>>
>> Charlie,
>>
>> I have prepared a 1.3.1 snapshot of the autotest suite at:
>>
>> http://www.gdal.org/dl/gdalautotest-1.3.1.tar.gz
>>
>> A small autotest capability within the swig/perl directory is fine.
>> However, at the point where it needs significant amounts of data
>> or other complexity I would appreciate it moving into the
>> gdalautotest tree.
>>
>>
>
> I confess I haven't looked at the autotest suite at all. The test code
> (test.pl) in swig/perl is a top-down, boot-strap test. It is a top-down
> test in the sense it tries to test everything that's available through
> the Perl API in a general sense. It currently overlooks very many
> details and does not check against programmers errors (because many
> mistakes still cause a segfault). It is a boot-strap test in the sense
> that it creates a gdal dataset or an ogr datasource first with known
> properties and then opens it and checks if the properties come back the
> same. I'm planning to put more tests for checking that the interface
> works as it should (typemaps are ok) and that it is more fool-proof.
>
> gdalautotest-1.3.1.tar.gz is 283k and gdal-1.3.1a2.tar.gz is 6.5M. A
> stupid question comes to mind: why are they separate?
> Testing is of course a large subject but I'd like to have a good
> self-test an integrated part of the Perl module as is common (required
> in fact) in Perl world (the normal install procedure is perl
> Makefile.PL; make; make test; make install). The current test.pl is 15k
> and will probably never be more than 30k.
>
> Charlie, what do you mean exactly with the "stay consistent"?
>
> All the best,
>
> Ari
>
>
>> Best regards,
>> --
>> ---------------------------------------+--------------------------------------
>>
>> I set the clouds in motion - turn up | Frank Warmerdam,
>> warmerdam at pobox.com
>> light and sound - activate the windows | http://pobox.com/~warmerdam
>> and watch the world go round - Rush | Geospatial Programmer for Rent
>>
>> _______________________________________________
>> Gdal-dev mailing list
>> Gdal-dev at lists.maptools.org
>> http://lists.maptools.org/mailman/listinfo/gdal-dev
>>
>>
>
>
More information about the Gdal-dev
mailing list