[Insight-developers] ITK Dashboard : Failing tests in Applications : BMP and RGB assumption

[Luis Ibanez]

Hi Bill,

A change has been committed to disable these two BMP-related tests:

http://public.kitware.com/cgi-bin/viewcvs.cgi/ConvertBetweenFileFormats/CMakeLists.txt?root=InsightApplications&r1=1.10&r2=1.11&sortby=date

    Thanks,


         Luis


---------------------------------------------------------
On Sun, Apr 26, 2009 at 8:23 AM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
> This is the bug I fixed:
> http://public.kitware.com/Bug/view.php?id=7773
>
> I would remove those tests from the application.
>
> Bill
>
> On Sat, Apr 25, 2009 at 4:07 PM, Luis Ibanez <luis.ibanez at kitware.com> wrote:
>> Hi Bill,
>>
>> I'm tracking the two failing tests from InsightApplications in
>>
>>                           redwall.kitware.
>>
>> http://www.cdash.org/CDash/viewTest.php?onlyfailed&buildid=319965
>>
>> They are:
>>
>> a) ccvnt_from_bmp_unsigned_char
>> b) compare_bmp_unsigned_char
>>
>> The test (b) fails as a secondary effect of (a) failing.
>>
>> The test (a) fails because the input BMP image is reported to be
>> an image of RGB  pixels and 3 components, despite the fact that
>> it has depth 8 bits per pixel.
>>
>> The tool
>>
>> InsightApplications/ConvertBetweenFileFormats/castconvert.cxx
>>
>> doesn't support RGB as a pixel type.
>>
>> This test started failing on October 2008:
>> http://www.cdash.org/CDash/viewTest.php?buildid=193504&onlyfailed
>>
>> Failing on
>> http://www.cdash.org/CDash/testDetails.php?test=10733286&build=193504
>>
>>
>> This seems to be related to your modifications of the file
>>
>>           Insight/Code/IO/itkBMPImageIO.cxx
>> http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/IO/itkBMPImageIO.cxx?root=Insight&r1=1.28&r2=1.29
>>
>>
>> Where the BMPImageIO reader, by default declares all images
>> to be of RGB type, even if they are of 8 bits.
>>
>>   1) Is this the correct behavior for BMP ?
>>
>>   2) Is the error in our creation of the BMP file: image_out.bmp  ?
>>
>>   3) Should we remove this test from Insight Applications
>>
>>
>>  Thanks for any hints,
>>
>>
>>      Luis
>>
>