Added a new parameter to the read tests called `perceptive-fuzz`.
The parameter, when active, modifies the fuzziness value based on the alpha value of the pixel. The more transparent the pixel, the more the fuzziness value increases.
We have found that some image manipulation functions give different results depending on the architecture (we think it is differences in rounding). These differences can become problematic with small alpha values when there are several image conversions from normal alpha to premultiplied alpha (and vice versa).
In particular, the offending plugin is XCF.
The parameter should be set if and only if necessary. CMakeList has not been modified to allow it to be enabled on all format images (you can still try it from the command line). To use it, you need to set it in the JSON file of the image that has problems (after careful analysis).
More info about the issue on #18
This MR also fixes a bug in `fazzeq()`: it only compared 1/4 of the image.
Below is the same XCF image rendered on AMD64 and PowerPC:
- AMD64:
- PowerPC:
The image is visually the same because the differences are with very low alpha and therefore are negligible. The patch proposed with this MR is useful in these cases.
Fixes the following warning:
```
/home/daniel/kimageformats/src/imageformats/dds.cpp: In function ‘qfloat16 readFloat16(QDataStream&)’:
/home/daniel/kimageformats/src/imageformats/dds.cpp:1037:11: warning: ‘void* memcpy(void*, const void*, size_t)’ copying an object of non-trivial type ‘class qfloat16’ from an array of ‘quint16’ {aka ‘short unsigned int’} [-Wclass-memaccess]
1037 | memcpy(&f16, &rawData, sizeof(rawData));
| ~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from /usr/include/qt6/QtCore/qmetatype.h:14,
from /usr/include/qt6/QtCore/qobject.h:18,
from /usr/include/qt6/QtCore/qiodevice.h:10,
from /usr/include/qt6/QtGui/qimageiohandler.h:9,
from /usr/include/qt6/QtGui/QImageIOPlugin:1,
from /home/daniel/kimageformats/src/imageformats/dds_p.h:13,
from /home/daniel/kimageformats/src/imageformats/dds.cpp:12:
/usr/include/qt6/QtCore/qfloat16.h:46:7: note: ‘class qfloat16’ declared here
46 | class qfloat16
| ^~~~~~~~
```
Should also fixes the following failed tests under PowerPC (32-bits):
```
INFO : rgba16dx10.dds: converting rgba16dx10.dds from RGBA16FPx4 to ARGB32
FAIL : rgba16dx10.dds: differs from rgba16dx10.png
expected data written to rgba16dx10.dds-expected.data
actual data written to rgba16dx10.dds-actual.data
```
```
INFO : rgba_f16.dds: converting rgba_f16.dds from RGBA16FPx4 to ARGB32
FAIL : rgba_f16.dds: differs from rgba_f16.png
expected data written to rgba_f16.dds-expected.data
actual data written to rgba_f16.dds-actual.data
```
- Added a class to read and write minimal exif metadata.
- JXL plugin uses EXIF metadata to load/save the resolution of the image (like GIMP).
- JXL plugin uses EXIF metadata to set/store text metadata and date/time.
- Enable info display in Dolphin (JXL File -> Properties -> Details on a JXL file, see image below).
- Enabled read test to check also image metadata and resolution.
{width=401 height=357}
The issue was identified by OSS Fuzz and the feature was not covered by our tests.
- Added earth-cubemap.dds under MIT licenses taken from [Open Toolkit library](https://github.com/mono/opentk/tree/main/Source/Examples/Data/Textures)
- Fix a wrong image size returned by a cubemap image
- Read test skips .license files
Let Qt rotate the image when the ImageAutotransform option is set to true.
In tests it also solves the image size control with the value returned by the options with certain rotations.
EXR, HDR, JXR and PFM formats support High Dynamic Range images (FP values grater than 1).
In summary, here is the list of changes:
EXR, HDR, JXR and PFM: When working with FP formats, the clamp between 0 and 1 is no longer done.
EXR: Removed old SDR code and conversions. Due to the lack of a QImage Gray FP format, Gray images are output as RGB FP (recently added code for Qt 6.8 has been removed).
PFM: Due to the lack of a QImage Gray FP format, Gray images are output as RGB FP.
HDR: Added rotation and exposure support.
With this patch, EXR, JXR, HDR, PFM behave like Qt's TIFF plugin when working with FP images.
While working on MR !230 I noticed that the options read I entered into several plugins could not be read after reading the image.
**The patch fixes problems reading options in plugins and adds option checking in the readtest.cpp.**
In particular, the reading test does the following additional actions:
- reads options before reading the image;
- compare the options read with the options returned by the reader after reading the image;
- compares the format and size of the returned image with the format and size returned by the reader.
Qt 6.8 will introduce native support for the CMYK (8-bit) format.
With this patch you will finally be able to correctly see the colors of CMYK images with ICC profile.
The testing part has been updated with the addition of an (optional) json file for each image to test. Inside you enter which image to use depending on the Qt version.
In short:
- Added native CMYK suport to PSD reader
- CMYK with alpha is converted using QColorSpace in a RGBA image
- Read tests changed to use the correct comparison image based on the Qt version
- Fixed also XCF tests: now works with all Qt version (see also [QTBUG-120614](https://bugreports.qt.io/browse/QTBUG-120614))
- Work around for CCBUG: 468288
Plugin to read RAW camera images based on LibRAW.
- Supersedes MR !86
- Support to LibRaw 0.20 and 0.21-Beta
- Support to multi-shot images: use imageCount(), jumpToImage() to select the wanted shot
- By default generates 16-bits sRGB images using camera white balance and interpolation AHD
- Should fix CCBUG: 454208: on my Debian with KF5.96 and the pulgin installed, I see the preview of all my RAW files (ARW included) in Dolphin
News compared to V1 (MR !86)
- Fix possible stack overflow due to the huge size of LibRaw class
- Fix image allocation with Qt 6 (make use of QImageIOHandler::allocateImage())
- Support to XMP metapacket
- Support to quality option. For e.g. you can focus on quality (q = 10) or speed (q = 1)
- oss-fuzz available [here](https://github.com/mircomir/oss-fuzz/tree/raw_fuzz/projects/kimageformats)
Unfortunately none of them pass since it seems they can't load a png,
save it to their format with loseless quality and read it back and get
exactly the same contents than the png
According to relicensecheck Brad is OK with changing LGPLv2 to LGPLv2+,
which is required to be compatible with the LGPL-2.1-or-later licensed
source files.
Qt also has a tga image plugin so unless we make sure ours is used first
tests are not testing what they should
On a side note their plugin fails our tests so someone with enough time
should report the failures to them
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.
REVIEW: 116567
Two main tests: read the file and compare to a given image (loaded from
png), and write a file out and compare to a previously-written image.
The write test in particular is kind of fragile, but it should help
catch some errors where a code change was not intended to affect the
output of the plugin.
Tests are included for all the writable formats, as well as every
readable format that GIMP can write to (and so I could easily get an
example of).
REVIEW: 115430
