I have gotten a merge request a while ago that achieves this, but
- it was for the sync branch, so it does not apply anymore, and
- more importantly, it makes the whole
MLImageFormatReader and the PageIdxTable generic over the voxel type as well.
Instead of creating an enum of readers, I would like to have a get_tile() wrapper that looks up the voxel type and calls the (already generic) get_tile() method with the right type argument.
For that, it makes sense to parse the "ML_IMAGE_DTYPE" tag in the file header on load, and store it inside an enum field.
Probably, the MLImageFormatReader should have a getter method for the voxel type (much like calculateOutputImageProperties in the ML).
I have gotten a merge request a while ago that achieves this, but
MLImageFormatReaderand thePageIdxTablegeneric over the voxel type as well.Instead of creating an enum of readers, I would like to have a get_tile() wrapper that looks up the voxel type and calls the (already generic) get_tile() method with the right type argument.
For that, it makes sense to parse the "ML_IMAGE_DTYPE" tag in the file header on load, and store it inside an enum field.
Probably, the MLImageFormatReader should have a getter method for the voxel type (much like
calculateOutputImagePropertiesin the ML).