Re: Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/#review52065 --- This review has been submitted with commit 0f795e66256b7ce0b960b08e1ebe3f608ccd3110 by Alex Merry to branch master. - Commit Hook On March 3, 2014, 3:23 p.m., Alex Merry wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- (Updated March 3, 2014, 3:23 p.m.) Review request for KDE Frameworks. Repository: kimageformats Description --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by some specified amount from the corresponding byte in the expected image, to allow for rounding errors etc. By default, no deviation is permitted, but the XCF tests are allowed a deviation of 1, as the alpha blending can result in rounding errors (depending on whether hardware acceleration is used, for example). In the end, we are not too concerned about a small deviation that is invisible to the human eye. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs - autotests/CMakeLists.txt 5c6508490344ca29097a3f13d01571658ad34786 autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/#review52064 --- This review has been submitted with commit 895305050c6af3e98bbb568c093a34a2a1e486d5 by Alex Merry to branch master. - Commit Hook On March 3, 2014, 3:23 p.m., Alex Merry wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- (Updated March 3, 2014, 3:23 p.m.) Review request for KDE Frameworks. Repository: kimageformats Description --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by some specified amount from the corresponding byte in the expected image, to allow for rounding errors etc. By default, no deviation is permitted, but the XCF tests are allowed a deviation of 1, as the alpha blending can result in rounding errors (depending on whether hardware acceleration is used, for example). In the end, we are not too concerned about a small deviation that is invisible to the human eye. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs - autotests/CMakeLists.txt 5c6508490344ca29097a3f13d01571658ad34786 autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- (Updated March 5, 2014, 12:11 p.m.) Status -- This change has been marked as submitted. Review request for KDE Frameworks. Repository: kimageformats Description --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by some specified amount from the corresponding byte in the expected image, to allow for rounding errors etc. By default, no deviation is permitted, but the XCF tests are allowed a deviation of 1, as the alpha blending can result in rounding errors (depending on whether hardware acceleration is used, for example). In the end, we are not too concerned about a small deviation that is invisible to the human eye. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs - autotests/CMakeLists.txt 5c6508490344ca29097a3f13d01571658ad34786 autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/#review51959 --- Ship it! Ship It! - Kevin Ottens On March 3, 2014, 3:23 p.m., Alex Merry wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- (Updated March 3, 2014, 3:23 p.m.) Review request for KDE Frameworks. Repository: kimageformats Description --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by some specified amount from the corresponding byte in the expected image, to allow for rounding errors etc. By default, no deviation is permitted, but the XCF tests are allowed a deviation of 1, as the alpha blending can result in rounding errors (depending on whether hardware acceleration is used, for example). In the end, we are not too concerned about a small deviation that is invisible to the human eye. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs - autotests/CMakeLists.txt 5c6508490344ca29097a3f13d01571658ad34786 autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- Review request for KDE Frameworks. Repository: kimageformats Description --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by 1 from the corresponding byte in the expected image, to allow for rounding errors etc. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs - autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/#review51782 --- My understanding is that all image formats we support are supposed to be lossless (including JPEG 2000, when quality is set to 100). In other words, if there is a deviation in RGB data, I would consider it as a bug, which the tests should at least warn about. If this is about XCF images: These can contain multiple layers which are alpha-composed during load-time. For those operations, an exception could be made, because blending could even be hardware-accelerated. - Christoph Feck On March 3, 2014, 1:04 p.m., Alex Merry wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- (Updated March 3, 2014, 1:04 p.m.) Review request for KDE Frameworks. Repository: kimageformats Description --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by 1 from the corresponding byte in the expected image, to allow for rounding errors etc. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs - autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 116567: Implement fuzzy image matching in readtest
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/116567/ --- (Updated March 3, 2014, 3:23 p.m.) Review request for KDE Frameworks. Changes --- Only do fuzzing for XCF images. Repository: kimageformats Description (updated) --- Implement fuzzy image matching in readtest Images are converted to ARGB32 format, then each byte (ie: each pixel channel) in the read image is allowed to deviate by some specified amount from the corresponding byte in the expected image, to allow for rounding errors etc. By default, no deviation is permitted, but the XCF tests are allowed a deviation of 1, as the alpha blending can result in rounding errors (depending on whether hardware acceleration is used, for example). In the end, we are not too concerned about a small deviation that is invisible to the human eye. Extract QImage::Format parsing into its own header Use the array-of-strings suggested by David Faure so that only one list has to be maintained instead of three. Diffs (updated) - autotests/CMakeLists.txt 5c6508490344ca29097a3f13d01571658ad34786 autotests/readtest.cpp dec9686e38389b04296fdf176db9fb8c1f3a56a4 tests/format-enum.h PRE-CREATION tests/imagedump.cpp 4b38c07d151d9bcb895f49a76e2bd03ddee41487 Diff: https://git.reviewboard.kde.org/r/116567/diff/ Testing --- imagedump still works. Most tests still pass; note that the non-alpha pic tests fail without https://git.reviewboard.kde.org/r/116568/diff/ as the wrong format (ARGB32 instead of RGB32) is constructed. This should make the xcf tests pass again on Jenkins. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
