typewalk.py

#!python3

"""
A simple python script to find all file types in a folder structure.
call: typewalk.py <folder> [lfs]
optional argument 'lfs' to run 'git lfs track *.ext' for each new BINARY FILE extension

I used this for making a git repo out of the full Android SDK/NDK.
"""

import os, sys
from pathlib import Path
print(sys.argv)

root = Path(sys.argv[1])

exts = []
for f in root.rglob("*"):
    if f.is_file() and f.suffix != '' and f.suffix not in exts:
        exts.append(f.suffix)
        with os.popen('file -b --mime-type ' + str(f), 'r') as ff:
            ft = ff.read()
            print(f.suffix, 'text' if ft.startswith('text') else 'binary', ft)
            if not ft.startswith('text') and len(sys.argv) > 2 and sys.argv[2] == 'lfs':
                os.system('git lfs track "*{}"'.format(f.suffix))