Getting FFProbe Information With Python

Question

I've been attempting to figure this out for forever now (I'm new to programming) and I can't figure it out.

I'm attempting to build a script that will test the file, and give me output from which I can get information like "Audio Format" that I can then put into the filename. However, I can't even get the script to return any file info. I've hit a wall at inserting an input file...

So at this point I just need help getting it to spit out info based on the argvs I've thrown in. Hopefully I'll be able to figure out how to parse the audio info from that.

My attempt that seems to be close:

#!/usr/bin/python
import os, sys, subprocess, shlex, re
from subprocess import call
def probe_file(filename):
    p = subprocess.Popen(['/opt/local/bin/ffprobe', '-show_format', '-pretty', '-loglevel quiet', -i filename], stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
    print filename
    print p.communicate()
[probe_file (f) for f in os.listdir('.') if not f.startswith('.')]

Answer 1

The json output from ffprobe is simple enough to parse in python without having to install a third-party library.

#!/usr/bin/env python

import argparse
import json
import subprocess
import sys
from pathlib import Path
from typing import NamedTuple


class FFProbeResult(NamedTuple):
    return_code: int
    json: str
    error: str


def ffprobe(file_path) -> FFProbeResult:
    command_array = ["ffprobe",
                     "-v", "quiet",
                     "-print_format", "json",
                     "-show_format",
                     "-show_streams",
                     file_path]
    result = subprocess.run(command_array, stdout=subprocess.PIPE, stderr=subprocess.PIPE, universal_newlines=True)
    return FFProbeResult(return_code=result.returncode,
                         json=result.stdout,
                         error=result.stderr)

Example usage, echoing the resulting json string:

if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='View ffprobe output')
    parser.add_argument('-i', '--input', help='File Name', required=True)
    args = parser.parse_args()
    if not Path(args.input).is_file():
        print("could not read file: " + args.input)
        exit(1)
    print('File:       {}'.format(args.input))
    ffprobe_result = ffprobe(file_path=args.input)
    if ffprobe_result.return_code == 0:
        # Print the raw json string
        print(ffprobe_result.json)
        
        # or print a summary of each stream
        d = json.loads(ffprobe_result.json)
        streams = d.get("streams", [])
        for stream in streams:
            print(f'{stream.get("codec_type", "unknown")}: {stream.get("codec_long_name")}')

    else:
        print("ERROR")
        print(ffprobe_result.error, file=sys.stderr)

So:

$ ./ffprobe.py -i input.mp4
File:       input.mp4

{
  "streams": [
    {
      "index": 0,
      "codec_name": "h264",
      "codec_long_name": "H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10",
      "profile": "High",
      "codec_type": "video",
      "codec_time_base": "1001/48000",
      "codec_tag_string": "avc1",
      "codec_tag": "0x31637661",
      "width": 1920,
      "height": 1080,
      "coded_width": 1920,
      "coded_height": 1088,
      "has_b_frames": 2,
      "pix_fmt": "yuv420p",
      "level": 40,
      "color_range": "tv",
      "color_space": "bt709",
      "color_transfer": "bt709",
      "color_primaries": "bt709",
      "chroma_location": "left",
      "refs": 1,
      "is_avc": "true",
      "nal_length_size": "4",
      "r_frame_rate": "24000/1001",
      "avg_frame_rate": "24000/1001",
      "time_base": "1/24000",
      "start_pts": 0,
      "start_time": "0.000000",
      "duration_ts": 442440,
      "duration": "18.435000",
      "bit_rate": "4970031",
      "bits_per_raw_sample": "8",
      "nb_frames": "442",
      "disposition": {
        "default": 1,
        "dub": 0,
        "original": 0,
        "comment": 0,
        "lyrics": 0,
        "karaoke": 0,
        "forced": 0,
        "hearing_impaired": 0,
        "visual_impaired": 0,
        "clean_effects": 0,
        "attached_pic": 0,
        "timed_thumbnails": 0
      },
      "tags": {
        "creation_time": "2020-05-20T15:37:59.000000Z",
        "language": "und",
        "handler_name": "L-SMASH Video Handler",
        "encoder": "AVC Coding"
      }
    },
    {
      "index": 1,
      "codec_name": "aac",
      "codec_long_name": "AAC (Advanced Audio Coding)",
      "profile": "LC",
      "codec_type": "audio",
      "codec_time_base": "1/48000",
      "codec_tag_string": "mp4a",
      "codec_tag": "0x6134706d",
      "sample_fmt": "fltp",
      "sample_rate": "48000",
      "channels": 2,
      "channel_layout": "stereo",
      "bits_per_sample": 0,
      "r_frame_rate": "0/0",
      "avg_frame_rate": "0/0",
      "time_base": "1/48000",
      "start_pts": 0,
      "start_time": "0.000000",
      "duration_ts": 885760,
      "duration": "18.453333",
      "bit_rate": "137011",
      "max_bit_rate": "140304",
      "nb_frames": "865",
      "disposition": {
        "default": 1,
        "dub": 0,
        "original": 0,
        "comment": 0,
        "lyrics": 0,
        "karaoke": 0,
        "forced": 0,
        "hearing_impaired": 0,
        "visual_impaired": 0,
        "clean_effects": 0,
        "attached_pic": 0,
        "timed_thumbnails": 0
      },
      "tags": {
        "creation_time": "2020-05-20T15:37:59.000000Z",
        "language": "und",
        "handler_name": "L-SMASH Audio Handler"
      }
    }
  ],
  "format": {
    "filename": "input.mp4",
    "nb_streams": 2,
    "nb_programs": 0,
    "format_name": "mov,mp4,m4a,3gp,3g2,mj2",
    "format_long_name": "QuickTime / MOV",
    "start_time": "0.000000",
    "duration": "18.453333",
    "size": "11779747",
    "bit_rate": "5106826",
    "probe_score": 100,
    "tags": {
      "major_brand": "mp42",
      "minor_version": "0",
      "compatible_brands": "mp42mp41isomavc1",
      "creation_time": "2020-05-20T15:37:59.000000Z"
    }
  }
}

video: H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10
audio: AAC (Advanced Audio Coding)

Answer 2

Here is technique I think it is both simple to use and easy to parse (tested with ffmpeg 3.x):

import subprocess
import xml.etree

def ffprobe(executable, filename):
    '''Runs ``ffprobe`` executable over ``filename``, returns parsed XML

    Parameters:

        executable (str): Full path leading to ``ffprobe``
        filename (str): Full path leading to the file to be probed

    Returns:

        xml.etree.ElementTree: containing all parsed elements

    '''

    cmd = [
        executable,
        '-v', 'quiet',
        '-print_format', 'xml', #here is the trick
        '-show_format',
        '-show_streams',
        filename,
        ]

    return xml.etree.ElementTree.fromstring(subprocess.check_output(cmd))

The data available comes from a string representation that looks like this:

 <ffprobe>
   <streams>
     <stream index="0" codec_name="h264" codec_long_name="H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10" profile="Constrained Baseline" codec_type="video" codec_time_base="1/60" codec_tag_string="avc1" codec_tag="0x31637661" width="560" height="320" coded_width="560" coded_height="320" has_b_frames="0" sample_aspect_ratio="0:1" display_aspect_ratio="0:1" pix_fmt="yuv420p" level="30" color_range="tv" color_space="bt709" color_transfer="bt709" color_primaries="bt709" chroma_location="left" refs="1" is_avc="true" nal_length_size="4" r_frame_rate="30/1" avg_frame_rate="30/1" time_base="1/90000" start_pts="0" start_time="0.000000" duration_ts="498000" duration="5.533333" bit_rate="465641" bits_per_raw_sample="8" nb_frames="166">
       <disposition default="1" dub="0" original="0" comment="0" lyrics="0" karaoke="0" forced="0" hearing_impaired="0" visual_impaired="0" clean_effects="0" attached_pic="0" timed_thumbnails="0"/>
       <tag key="creation_time" value="2010-03-20T21:29:11.000000Z"/>
       <tag key="language" value="und"/>
       <tag key="encoder" value="JVT/AVC Coding"/>
     </stream>
     <stream>...</stream>
   </streams>
   <format filename="/Users/andre/Projects/qnap/librarian/librarian/data/movie.mp4" nb_streams="2" nb_programs="0" format_name="mov,mp4,m4a,3gp,3g2,mj2" format_long_name="QuickTime / MOV" start_time="0.000000" duration="5.568000" size="383631" bit_rate="551193" probe_score="100">
     <tag key="major_brand" value="mp42"/>
     <tag key="minor_version" value="0"/>
     <tag key="compatible_brands" value="mp42isomavc1"/>
     <tag key="creation_time" value="2010-03-20T21:29:11.000000Z"/>
     <tag key="encoder" value="HandBrake 0.9.4 2009112300"/>
   </format>
 </ffprobe>

Answer 3

One-liner in a terminal: ffprobe -v quiet -print_format json -show_format -show_streams "foobar.mp4" > "foobar.mp4.json"

Hence, you may simply:

#!/usr/bin/python3
import subprocess
import json


def probe_file(filename):
    command_line = ['ffprobe', '-v', 'quiet', '-print_format', 'json', '-show_format', '-show_streams', f"{filename}"] 
    p = subprocess.Popen(command_line, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    print(filename) # * Sanity check *
    out, err =  p.communicate()
    if len(out)>0: # ** if result okay
        print("==========output==========")
        result = json.loads(out)
    else:
        result = {}
    if err:
        print("========= error ========")
        print(err)
    return result


result_json = probe_file('/foo/bar.mov')
print(json.dumps(result_json, indent=2))

NB: I have used this example. You may clone the code from github and use it as a module:

#python3
    from ffprobe_json import ffprobe

    ffprobe_result, error = ffprobe(file_path="/your/media_file.mp4", quiet=True)

Answer 4

I made some improvements to @mikebridg code to fit my needs, here it is if useful for anyone. If someone has comments about code style/mistakes, it's welcome... I've never had any external code review :S

class FFProbeResult:

    def __init__(self, return_code: int = None, output: str = '', error: str = '', format=None):
        self.return_code = return_code
        self.output = output
        self.error = error
        self.format = format
        self._output_as_dict = None

    def get_output_as_dict(self):
        if self._output_as_dict is None:
            if self.format == 'json':
                self._output_as_dict = json.loads(self.output)
            elif self.format == 'flat':
                output = [e.split('=') for e in self.output.strip().split('\n')]
                self._output_as_dict = {key_val[0]: key_val[1] for key_val in output}
            else:
                raise ValueError("ffprobe format '%s' not supported to build dict" % self.format)
        return self._output_as_dict

    def to_json_file(self, path: Path, mode='w', **kwargs):
        """
        :param mode: file open mode
        :param kwargs: kwargs for pathlib.Path().open()
        """
        path = path if isinstance(path, Path) else Path(path)
        with path.open(mode, **kwargs) as f:
            json.dump(self.get_output_as_dict(), f, indent=4)
            logger.debug('Dumped ffprobe output into %s', path)


def ffprobe(file_path, ffprobe_format="json", format_optn='', log_level='error') -> FFProbeResult:
    assert ffprobe_format in ['json', 'flat'], "format must be json or flat, not %s" % ffprobe_format
    format_optn = '=' + format_optn if format_optn else format_optn
    command_array = ["ffprobe",
                     "-v", log_level,
                     "-print_format", ffprobe_format + format_optn,
                     '-show_programs',
                     "-show_format",
                     "-show_streams",
                     f"{file_path}"]
    try:
        result = subprocess.run(command_array, stdout=subprocess.PIPE, stderr=subprocess.PIPE, universal_newlines=True)
    except Exception as e:
        logger.critical("ffprobe failed to run on %s, with the following error: '%s'\n"
                        " check first that cmd is in your path",
                        file_path, result.stderr, exc_info=True)
        raise e

    return FFProbeResult(return_code=result.returncode,
                         output=result.stdout,
                         error=result.stderr,
                         format=ffprobe_format)

Answer 5

You can use ffprobe-python module (https://github.com/gbstack/ffprobe-python). This module is compatitable with Python 3.

Or install it using pip

pip install ffprobe-python

Usage

from ffprobe import FFProbe
metadata = FFProbe("test.mp4")

Answer 6

There are a few problems in your code:

1. args list to Popen has last argument as -i filename which is a syntax error use '-i '+filename instead
2. shell=True is usually not needed and is unnecessary burden.

Other than that it seems to be working, are you not seeing output after fixing #1 ?

Edit: Looks like you are having problem with ffprobe commandline, so I installed it and changes you require are:

1. My ffprobe (ffprobe 0.7.3-4:0.7.3-0ubuntu0.11.10.1) doesn't seems to accept -i flag, input file is just passed as last argument.
2. you need to pass -loglevel and option of loglevel quiet as separate arguments i.e. [..., '-loglevel', 'quiet',..]

So after these changes here is a sample script:

#!/usr/bin/python
import os, sys, subprocess, shlex, re
from subprocess import call
def probe_file(filename):
    cmnd = ['ffprobe', '-show_format', '-pretty', '-loglevel', 'quiet', filename]
    p = subprocess.Popen(cmnd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    print filename
    out, err =  p.communicate()
    print "==========output=========="
    print out
    if err:
        print "========= error ========"
        print err

probe_file('drop.avi')

And I see the correct output:

==========output==========
[FORMAT]
filename=drop.avi
nb_streams=1
format_name=avi
format_long_name=AVI format
start_time=0:00:00.000000
duration=0:00:06.066667
size=660.000 Kibyte
bit_rate=891.217 Kbit/s
[/FORMAT]

========= error ========
ffprobe version 0.7.3-4:0.7.3-0ubuntu0.11.10.1, Copyright (c) 2007-2011 the Libav developers
  built on Jan  4 2012 16:08:51 with gcc 4.6.1
  configuration: --extra-version='4:0.7.3-0ubuntu0.11.10.1' --arch=amd64 --prefix=/usr --enable-vdpau --enable-bzlib --enable-libgsm --enable-libschroedinger --enable-libspeex --enable-libtheora --enable-libvorbis --enable-pthreads --enable-zlib --enable-libvpx --enable-runtime-cpudetect --enable-vaapi --enable-gpl --enable-postproc --enable-swscale --enable-x11grab --enable-libdc1394 --enable-shared --disable-static
  libavutil    51.  7. 0 / 51.  7. 0
  libavcodec   53.  6. 0 / 53.  6. 0
  libavformat  53.  3. 0 / 53.  3. 0
  libavdevice  53.  0. 0 / 53.  0. 0
  libavfilter   2.  4. 0 /  2.  4. 0
  libswscale    2.  0. 0 /  2.  0. 0
  libpostproc  52.  0. 0 / 52.  0. 0
Unsupported codec with id 114 for input stream 0

Answer 7

There is FFProbe wrapper for Python (https://pypi.org/project/ffprobe/).

You can install it easily: (sudo) pip install ffprobe

A typical usage:

from ffprobe import FFProbe
metadata = FFProbe("example.mp4")

An alternative to FFProbe is pymediainfo library (https://pymediainfo.readthedocs.io/en/stable/). You can find usage in the documentation site.