johnnyjohnandthefunkybunch / ytextractor

For hackweek 2017

Hack Week 2017 - YT Extractor

This Python script will take in a JSON file with YouTube video URLs and timestamps and will extract the main video and cut the videos to create the clips.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Prerequisites

• Python version 2.7.10 (Tested on)
• Have ffmpeg installed
• Dependant on youtube-dl (Install through pip install requirements.txt)

Note: If youtube-dl fails, then you should get the most updated version of youtube-dl and not the one in the requirements.txt.

Installing

Clone the project and change directories.

git clone [email protected]:JohnnyJohnAndTheFunkyBunch/ytextractor.git

Create virtual environment and activate it

virtualenv ytextractor
source ytextractor/bin/activate

Install the necesssary packages

pip install -r requirements.txt

Running the tests

There is an example.json file that can be used to test the program. To run the program enter the following:

python ytextractor.py example.json

If there were no errors and the video folder looks like this:

videos/
├── Fox_Snow_Dive_-_Yellowstone_-_BBC_Two
│   ├── clip1.mp4
│   ├── clip2.mp4
│   ├── clip3.mp4
│   ├── clip4.mp4
│   └── main.mp4
└── Lee_Sunbin_-_FILA_Dance
    ├── clip1.mp4
    └── main.mp4

Then you're gucci.

Documentation

Very simple to use the only argument you can give it, is the name of the JSON file with the information. Each video will contain it's own folder with the title of the video. Inside the folder will be a main.mp4 video which is the full video and clipX.mp4 videos which will be the clips.

All of this will be in the videos folder.

How to run

Feed it a JSON file

python ytextractor.py myjsonfile.json

JSON File

Here is an example of a JSON file:

{
    "data": [
        {
            "clips": [
                {
                    "start": "00:00:03",
                    "length": "00:00:07"
                },
                {
                    "start": "00:00:08",
                    "length": "00:00:03"
                },
                {
                    "start": "00:00:10",
                    "length": "00:00:10"
                },
                {
                    "start": "00:00:15",
                    "length": "00:00:02"
                }
            ],
            "video_url": "https://www.youtube.com/watch?v=dP15zlyra3c"
        },
        {
            "clips": [
                {
                    "start": "00:00:02",
                    "length": "00:00:30"
                }
            ],
            "video_url": "https://www.youtube.com/watch?v=YeGWmq6ahUw"
        }
    ]
}

data

The data key contains all an array of all video objects. A video object contains clips and video_url.

video_url

The video_url is the standard YouTube URL of the video

clips

clips is an array of start and length values that determines which clips to extract from the video

start

start should be a string of the format 00:00:00 that determines the start of your clip.

length

length should be a string of the format 00:00:00 that determines the length of your clip. Hence if you want to get a clip from 00:00:03 to 00:00:10 then your start would be 00:00:03 and your length would be 00:00:07.