This is a frame extractor and encoder for my Scratch project: (https://scratch.mit.edu/projects/614473645/editor). It uses (simple) run-length encoding to compress the image data for Bad Apple into a Scratch-sized file which is able to fit within the 5MB project JSON limit. Since the individual frames consist largely of black and white only, I quantize the pixels to one of either (255, 255, 255) or (0, 0, 0) depending on its brightness. Then, it is run-length encoded to convert frames into sequences of like pixels; for instance, the first frame is "10800B". The corresponding data can be recovered from within the Scratch project.
In order to get the video, run the command (provided you have Python and pip installed, or youtube-dl) in get_video.sh; subsequently, extracting the frames is as simple as running extract_frames.sh. The frames are then extracted into the frames directory, from where you can run the Python file:
python to_string.py. Note that, to run the frame extractor, you will need to have ffmpeg installed.
If you just wish to use the data obtained from this method, then feel free to just download the data.txt file. :)