similar:
- pdfgrep
- https://gist.github.com/ColonolBuendia/314826e37ec35c616d70506c38dc65aa
# considerations
- matching on mime (magic bytes) instead of filename