It tries to be as flexible as possible, but still requires a bit of configuration (including regexes, but if metadata is embedded into the files, it can be skipped).
I just released v3.0, making videos first-class data, and scanning + metadata matching faster.