> (I'm looking to extract audio clips from video files by using the subtitles and timing information, to create listening flashcards. Since Anki doesn't really do this, not reusing their code isn't really a big deal.)

Just fyi, I recently hacked something up that did exactly this, using a bit of csv parsing in python and ffmpeg to extract the clips. It wasn't too hard to write, though I can see it becoming more difficult if you want more sophistication.