Closed BeastBurst closed 1 year ago
For test the speed, you can try $diffz -SD-2m -p-8 -m-6 oldFile newFile outPatFile
or $diffz -SD-2m -p-8 -s-64 oldFile newFile outPatFile
(Note: The default patchFile is not compressed.)
If feasible, it is recommended to use the command line program directly.
'patchStepMemSize' does not need 2GB, 2MB is generally sufficient;
If you call the API yourself, it is recommended to use 'create_single_compressed_diff_block(const hpatch_TStreamInput* newData,...)' and 'patch_single_stream()' ;
filename to 'hpatch_TStreamInput' can use 'hpatch_TFileStreamInput';
If the data is very large, you can use other APIs, which are also faster: 'create_single_compressed_diff_stream()' and 'patch_single_stream()';
Try different 'kMatchBlockSize' values to see what happens.
I am running HDiffPatch on MacBook M2 Max have this settings:
if (_IS_NEED_BSDIFF)
endif
if (_IS_NEED_VCDIFF)
endif