Merge Files with Chapters

mergechapters.py

import datetime
import json
import os
import subprocess
import sys

#############
### USAGE ###
#############

if len(sys.argv) < 4:
    print("Usage:")
    print("{} [input file] [input file] [output file]".format(sys.argv[0]))
    print("")
    print("Both files are assumed to have their chapters")
    print("entered correctly and completely.")
    sys.exit(0)

########################
### Get Chapter List ###
########################

def getChapterList(videoFile):
    # Get chapter list as JSON
    result = subprocess.run(
        ["ffprobe", "-print_format", "json", "-show_chapters", videoFile],
        capture_output=True,
        text=True
    )
    # Load the JSON
    fileJson = json.loads(result.stdout)['chapters']
    # Map to Python object:
    # {
    #     id: 1
    #     start: 123.456
    #     end: 789.012
    # }
    chapters = list(map(
        lambda c: {
            'index': c['id'], 
            'start': float(c['start_time']), 
            'end': float(c['end_time']), 
            'title': c['tags']['title']}, 
        fileJson))
    return list(chapters)

########################
### Video 1 Duration ###
########################
 
# Get the duration of the first video
result = subprocess.run(
    ["ffprobe", "-v", "error", "-show_entries", "format=duration", "-of", "default=noprint_wrappers=1:nokey=1", sys.argv[1]], 
    capture_output=True,
    text=True
)
# Get the result and trim off the trailing newline.
file1duration = float(result.stdout.rstrip())
print("{} duration is {} seconds.".format(sys.argv[1], file1duration))

############################
### Video 2 Chapter List ###
############################

def chapterMigrator(chapter):
    startTime = chapter['start']
    endTime = chapter['end']
    
    offsetStartTime = file1duration + startTime
    offsetEndTime = file1duration + endTime
    
    return {'index': chapter['index'], 'start': offsetStartTime, 'end': offsetEndTime, 'title': chapter['title']}

video2rawchapters = getChapterList(sys.argv[2])
# Migrate these chapters to be offset from the end of the first file
video2chapters = list(map(chapterMigrator, video2rawchapters))
print("{} has {} chapters.".format(sys.argv[2], len(video2chapters)))

###########################
### Get file 1 metadata ###
###########################

result = subprocess.run(
    ["ffmpeg", "-i", sys.argv[1], "-f", "ffmetadata", "-"],
    capture_output=True,
    text=True
)

metadata = result.stdout

##################################
### Append file 2 chapter list ###
##################################

metadataFileName = "metadata.txt"

# Note the timestamps are in milliseconds, and should be integers.
for c in video2chapters:
    metadata += f"""
[CHAPTER]
TIMEBASE=1/1000
START={int(c['start'] * 1000)}
END={int(c['end'] * 1000)}
title={c['title']}"""

with open(metadataFileName, "w") as metadataFile:
    metadataFile.write(metadata)

############################
### Join two video files ###
############################

fileListFileName = "files.txt"

fileList = f"""
file {sys.argv[1]}
file {sys.argv[2]}"""

with open(fileListFileName, "w") as fileListFile:
    fileListFile.write(fileList)

if os.path.exists(sys.argv[3]):
    os.remove(sys.argv[3])

print("Joining {} and {} into {}...".format(sys.argv[1], sys.argv[2], sys.argv[3]))

result = subprocess.run(
    ["ffmpeg", "-f", "concat", "-i", fileListFileName, "-i", metadataFileName, "-map_metadata", "1", "-c", "copy", sys.argv[3]],
    capture_output=False
)

print("...file {} created.".format(sys.argv[3]))

# Clean up.
os.remove(metadataFileName)
os.remove(fileListFileName)

Just FYI ffmpeg does indeed seem to miss this concat feature as acknowledged in their bug tracker

@eugenesvk Huh, thanks for pointing that out.

This most recent activity does not instill confidence, however:

Changed 2 years ago by dextro

There were some minor quirks with file names and dropped audio streams but otherwise this worked a treat, thanks for sharing and saving me a lot of time and frustration!

By the way if you want to make sure it copies all audio streams, change line 128 like so:

    ["ffmpeg", "-f", "concat", "-i", fileListFileName,"-i", metadataFileName, "-map_metadata", "1",  "-map", "0", "-c", "copy", sys.argv[3]],

Basically adding -map 0 just before the copy command, more info here if you're interested how it works: https://trac.ffmpeg.org/wiki/Concatenate#samecodec

Hey Guys!
I am getting Error below:

Traceback (most recent call last):
File "./mergechapters.py", line 55, in
text=True
File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.7_3.7.2544.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 488, in run
with Popen(*popenargs, **kwargs) as process:
File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.7_3.7.2544.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 800, in init
restore_signals, start_new_session)
File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.7_3.7.2544.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 1207, in _execute_child
startupinfo)
FileNotFoundError: [WinError 2] The system cannot find the file specified

@xcomprs Sorry, but I don't have any installation of Windows available to me, so I'm not sure what's going on here. :(

@xcomprs

Hey Guys! I am getting Error below:

Traceback (most recent call last): File "./mergechapters.py", line 55, in text=True File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.7_3.7.2544.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 488, in run with Popen(*popenargs, **kwargs) as process: File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.7_3.7.2544.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 800, in init restore_signals, start_new_session) File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.7_3.7.2544.0_x64__qbz5n2kfra8p0\lib\subprocess.py", line 1207, in _execute_child startupinfo) FileNotFoundError: [WinError 2] The system cannot find the file specified

I ran into the same issue, it turns out I didn't have ffprobe installed. It normally comes with ffmpeg, but if you installed ffmpeg manually, you might not have it.

You can check by opening up command prompt and running

ffprobe -version

If you get an error saying ffprobe isn't recognized, you need to install it.

It should be installed if you run

choco install ffmpeg

in command prompt with Administrator privileges.

I made the following revisions to my copy to help handle whitespace better:

-    ["ffmpeg", "-f", "concat", "-i", fileListFileName, "-i", metadataFileName, "-map_metadata", "1", "-c", "copy", sys.argv[3]],
+    ["ffmpeg", "-f", "concat", "-safe", "0", "-i", fileListFileName, "-i", metadataFileName, "-map_metadata", "1", "-c", "copy", sys.argv[3]],

fileList = f"""
- file {sys.argv[1]}
- file {sys.argv[2]}"""
+ file '{sys.argv[1]}'
+ file '{sys.argv[2]}'"""

Fixes "Unsafe File Name" issue that I had. That may be what our windows friend was suffering...hard to say without more info

On my version of ffmpeg (4.1.11-0), I get an error unless I specify -f ffmetadata prior to -i metadata.txt. Combining this with other suggestions above yields:

    ["ffmpeg", "-f", "concat", "-safe", "0", "-i", fileListFileName, "-f", "ffmetadata", "-i", metadataFileName, "-map_metadata", "1", "-map", "0", "-c", "copy", sys.argv[3]],

Works wonderfully on Arch Linux, thank you!

Based on previous edits by @jojo2357

+ from uuid import uuid4

- fileListFileName = "files.txt"
+ metadataFileName = f"metadata-{str(uuid4())}.txt"

- fileListFileName = "files.txt"
+ fileListFileName = f"files-{str(uuid4())}.txt"

- file '{sys.argv[1]}'
- file '{sys.argv[2]}'"""
+ file '{os.path.abspath(sys.argv[1])}'
+ file '{os.path.abspath(sys.argv[2])}'
+ """

Resolves an issue I had where the file would delete any existing files I had that happened to be named "files.txt" or "metadata.txt"