How to download chapters of a youtube video as separate video files?

Question

A New YouTube Feature

Youtube recently enabled the option to add chapters in YouTube videos.

This is done by simply adding the chapter timestamps (start time) in the description of the video.

Related Topics

Using youtube-dl we can

download a portion of a YouTube video by providing the start time and duration.

extract description of a YouTube video as .description files.

QUESTION

Is there a single command to download chapters of a youtube video as separate video files without providing the start time and duration rather it automatically extracts the timestamps from the .description?

---

Solution

You can't do this directly. I have written a Python script which will download the whole file, process the description file and then split the downloaded file into it's component chapters.

Usage - if you save this into a file called "yt_chapters.py" then run

yt_chapters.py --split "https://www.youtube.com/watch?v=m5lp8S-YgrQ"

You can pass multiple download links and it will loop through the list.

The chapter record will be added to the original downloaded MP4.

Requirements: youtubedl, ffmpeg and MP4Box (GPAC). If you don't want to keep the original MP4, or if you don't care about it having the chapters, you can comment out that section. youtubedl and ffmpeg should be in your system path or else change the code below to include the full path to the executables.

#! /usr/bin/env python
# -*- coding: utf-8 -*-

import subprocess, os, argparse, datetime, re, sys

def get_source_name(download_source:str)->str:

    # Get the expected filename in order to process everything down the line
    result = subprocess.run(["youtube-dl", "--get-filename",
                            "-o", "%(title)s", download_source],
                            capture_output=True, encoding='UTF8')
    return result.stdout.rstrip('\n')

def get_best_mp4_single_file(download_source:str, description_file:str,
                             fix_required:bool=False)->list:

    # Download the file as mp4 (in order to use MP4)
    subprocess.run(["youtube-dl", "--write-description", "-f",
                    "bestvideo[ext=mp4]+bestaudio[ext=m4a]/mp4",
                    "-o", "%(title)s.%(ext)s", link_to_download])
    if fix_required:
        input("Fix the description file to have the chapters "
              "on their own lines in the format timecode name "
              "and then press Enter to continue...")
    return read_description_file(description_file=description_file)

def get_worst_mp4_single_file(download_source:str, description_file:str,
                              fix_required:bool=False)->list:

    # Download the file as mp4 (in order to use MP4)
    subprocess.run(["youtube-dl", "--write-description", "-f",
                    "worstvideo[ext=mp4]+worstaudio[ext=m4a]/mp4",
                    "-o", "%(title)s.%(ext)s", link_to_download])
    if fix_required:
        input("Fix the description file to have the chapters "
              "on their own lines in the format timecode name "
              "and then press Enter to continue...")
    return read_description_file(description_file=description_file)

def add_chapters_to_mp4(chapter_file_name:str, name_for_download:str)->None:

    # Use MP4Box to mux the chapter file with the mp4
    subprocess.run(["MP4Box", "-chap", chapter_file_name, name_for_download])

def read_description_file(description_file:str)->list:

    # Read the description file
    # Split into time and chapter name

    list_of_chapters = []

    with open(description_file, 'r') as f:
        # only increment chapter number on a chapter line
        # chapter lines start with timecode
        line_counter = 1
        for line in f:
            result = re.search(r"\(?(\d?[:]?\d+[:]\d+)\)?", line)
            try:
                # result = re.search("\(?(\d+[:]\d+[:]\d+)\)?", line)
                time_count = datetime.datetime.strptime(result.group(1), '%H:%M:%S')
            except:
                try:
                    # result = re.search("\(?(\d+[:]\d+)\)?", line)
                    time_count = datetime.datetime.strptime(result.group(1), '%M:%S')
                except:
                    continue
            chap_name = line.replace(result.group(0),"").rstrip(' :\n')
            chap_pos = datetime.datetime.strftime(time_count, '%H:%M:%S')
            list_of_chapters.append((str(line_counter).zfill(2), chap_pos, chap_name))
            line_counter += 1

    return list_of_chapters

def write_chapters_file(chapter_file:str, chapter_list:tuple)->None:

            #open(chapter_file, 'w').close()

            # Write out the chapter file based on simple MP4 format (OGM)
            with open(chapter_file, 'w') as fo:
                for current_chapter in chapter_list:
                    fo.write(f'CHAPTER{current_chapter[0]}='
                            f'{current_chapter[1]}\n'
                            f'CHAPTER{current_chapter[0]}NAME='
                            f'{current_chapter[2]}\n')


def split_mp4(chapters:list, download_filename:str, download_name:str)->None:

    # current_duration = subprocess.run(['ffprobe', '-i', download_filename,
                                       # '-show_entries', 'format=duration',
                                       # '-v', 'quiet', '-of', 'csv="p=0"'],
                                        # capture_output=True, encoding='UTF8')
    current_duration_pretext = subprocess.run(['ffprobe', '-i', download_filename,
                                       '-show_entries', 'format=duration',
                                       '-v', 'quiet'],
                                        capture_output=True, encoding='UTF8')
    current_duration = float(current_duration_pretext.stdout[18:-13])
    m, s = divmod(current_duration, 60)
    h, m = divmod(m, 60)
    current_dur = ':'.join([str(int(h)),str(int(m)),str(s)])
    for current_index, current_chapter in enumerate(chapters):
        # current_chapter will be a tuple: position, timecode, name
        next_index = current_index + 1
        start_time = current_chapter[1]
        try:
            end_time = chapters[next_index][1]
        except:
            end_time = current_dur
        output_name = f'{download_name} - ({current_chapter[2]}).mp4'
        subprocess.run(["ffmpeg", "-ss", start_time, "-to", end_time,
                        "-i", download_filename, "-acodec", "copy",
                        "-vcodec", "copy", output_name])

if __name__ == '__main__':

    parser = argparse.ArgumentParser(description='Download Youtube videos with chapters')

    parser.add_argument('links', metavar='N', type=str, nargs='+',
                        help='List of links to download')

    parser.add_argument("--split", help="Split chapters into individual files",
                        action="store_true")
    parser.add_argument("--test", help="Download worst version to speed up testing",
                        action="store_true")
    parser.add_argument("--fix", help="Pause to allow manual correction of the description file",
                        action="store_true")
    args = parser.parse_args()
    # print(args.links)

    for link_to_download in args.links:

        download_name = get_source_name(download_source=link_to_download)
        download_filename = f'{download_name}.mp4'
        description_file = f'{download_name}.description'
        chapter_file = f'{download_name}_chapter.txt'

        if not args.test:
            chapters = get_best_mp4_single_file(download_source=link_to_download,
                                                description_file=description_file,
                                                fix_required=args.fix)
        else:
            chapters = get_worst_mp4_single_file(download_source=link_to_download,
                                                description_file=description_file,
                                                fix_required=args.fix)

        
        if not chapter_file:
            print("No chapters found")
            sys.exit(1)
        
        write_chapters_file(chapter_file=chapter_file, chapter_list=chapters)

        if args.split:
            split_mp4(chapters=chapters, download_filename=download_filename,
                      download_name=download_name)
        
        add_chapters_to_mp4(chapter_file_name=chapter_file,
                              name_for_download=download_filename)

        # Clean up the text files
        os.remove(description_file)
        os.remove(chapter_file)

Comments

[xrisk]

Thanks a lot. A note of caution: ffmpeg will error if the chapters have slashes in their names. Modify the line generating output_name in that case.

[Mikhail T.]

Why does not this work for this video? It gets downloaded, there are no errors, but no chapter-files either... Opening the video-page in browser, I do see the sections for each song -- and intermissions. Thank you!

[Alan]

@MikhailT. Are you sure you have ffmpeg and MP4Box in your system's path? I have tried with that video and it works for me.

[Mikhail T.]

@Alan, the program is installed as mp4box here on FreeBSD, but yes, it is in the path. Your script runs with no errors (after I changed the MP4Box to lower case), it just does not produce any track-files :( How many have you gotten? I see, that it gets the chapters-file -- will try to debug it. Do you have it on gitlabs or github somewhere, perhaps?

[Alan]

@MikhailT. github.com/alan1world/yt_chapters