RFR: 8316497 : ColorConvertOp - typo for non-ICC conversions needs one-line fix

Renjith Kannath Pariyangad rkannathpari at openjdk.org
Thu Jan 11 06:56:26 UTC 2024


On Wed, 10 Jan 2024 19:50:45 GMT, Sergey Bylokhov <serb at openjdk.org> wrote:

>> Can this patch be covered by the new test?
>
>>I'm not sure… We want the filter to take another path, there could be a list of filters applied, if I understand @mrserb correctly. Sergey could be able to provide a more detailed guidance.
> 
> The current test validates two code paths:
>  - wrapper>icc_color_space->wrapper
>  - icc_color_space->icc_color_space->icc_color_space
> 
> The color space in the middle is always icc_color_space: https://github.com/openjdk/jdk/pull/16895/files#diff-70b19b2642d6d3f44904de8b6eb2993e1c97320e3476898c4372db364c4288b7R130
> 
> If we will use the wrapper for the middle as well we will cover this code path:
> https://github.com/openjdk/jdk/pull/16895/files#diff-e3d6eea060882cab00827c00e1a83b0e0a5b2a31fa9a9aa2122841bbd57c4a6dL853

@mrserb ,

Did you mean to use the wrapper for the middle like `ColorSpace mid = createCS(ColorSpaceSelector.WRAPPED_PYCC);` , So we can achieve : 
                               **Current**                                                                                 **New**
    wrapper->icc_color_space->wrapper                                >                  wrapper->wrapper->wrapper                          
    icc_color_space->icc_color_space->icc_color_space         >                  icc_color_space->wrapper->icc_color_space

Correct me if I am wrong

-------------

PR Comment: https://git.openjdk.org/jdk/pull/16895#issuecomment-1886405188


More information about the client-libs-dev mailing list