v3.2.2

Changelog:

• Fixed storage_path issue, should now give actual path rather than empty string.

v3.2.0

Changelog:

• Added URL and do_not_copy options to video_duration.
• Moved _copy_video_to_video_dir() and _create_required_dirs_and_check_for_errors() to a new file, "videocopy.py" to facilitate the last point.
• Adjusted test_videoduration.py and test_videohash.py to facilitate this and ensure that it's tested properly.
• Removed support for Python versions <=3.7 to comply with yt-dlp support.

v3.1.0

Changelog:

• Roughly halved average hashing time by using video_duration to limit FramesExtractor.detect_crop as suggested here.
• Added do_not_copy option to VideoHash. If set to False it will copy the video into temporary storage, it's set to True by default. This also saves on memory and improves performance. As suggested here.
• Fixed stdint issue with subprocesses as suggested here.
• Added new wiki to README which can be found here.

v3.0.3

Changelog:

• Changed the Windows test workflow to use ffmpeg-installer instead of pyunpack
• Amended the README

v3.0.2

Changelog:

• Forked videohash
• Changed package name to videohash2
• Switched "imagedominantcolor" to "imagedominantcolour"
• Updated Image.ANTIALIAS to Image.Resampling.LANCZOS to comply with Pillow update
• Removed Python 3.6 support and added Python 3.11 support.