Get FFFmpeg Metadata In Python

Getting metadata with the ffmpeg module, easier than using subprocess and capturing and parsing the output.

Get Titles

The key bit is

info = ffmpeg.probe(x)
fmt = info['format']
tags = fmt['tags']
title = tags['title']

though you have to check for key errors (so I use try...except. FFmpeg throws a ffmpeg._run.Error exception if e.g. the file is not found.

The complete script

#!/usr/bin/env python

import ffmpeg
from glob import glob
from icecream import ic; ic.configureOutput(includeContext=True)
import re
import json
import sys

fails = []
out = []

args = sys.argv[1:]
if len(args) == 0:
    # default to all .m4a files in current directory
    m4as = glob("*.m4a")

for x in m4as:
  try:
    info = ffmpeg.probe(x)
  except Exception as e:
    ic("probe",e)    
    fails.append(x)
    continue
  try:
    fmt = info['format']
    tags = fmt['tags']
    title = tags['title']
  except KeyError as e:
    ic(f"Failed to get tags",e)
    fails.append(x)
    continue

  m = re.search(r"\d{1,3}",title)
  if not m:
    print("No number in {x}")
    fails.append(x)
    continue

  n = m.group()
  out.append(y := [n,title,x])
  print(*y)

with open("titles.json","wt") as f:
  json.dump(out,f)
with open("fails.json","wt") as f:
  json.dump(fails,f)