Skip to content

A simple command utility to extract information from the YouTube API v3 for scientific purposes.

License

Notifications You must be signed in to change notification settings

rattletat/yt-scraper

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

YouTube Scraper

made-with-python License: Unlicense version

A simple command utility to extract information from the YouTube API v3 for scientific purposes.

About

This Python based command line utility enables the easy extraction of information from the YouTube API (Version 3). Currently, it supports only a small subset of the API interface and focuses on extracting related videos from given starting points.

Installation

First, make sure you have a recent version of Python 3 installed.

Next, install yt-scraper by using pip:

sudo pip install yt-scraper

Update by adding the --upgrade flag:

sudo pip install --upgrade yt-scraper

Windows users may need to alter the command to:

py -m pip install --upgrade yt-scraper

Mac users may need to alter the command to:

python3 -m pip install --upgrade yt-scraper

API Key

In order to use this program, you will need an official YouTube API key. You can obtain one from this page and use it with the following examples by appending -k <KEY> to them.

Usage

Currently, yt-scraper has two commands: search and config

The first command is used to query the YouTube API, whereas the second is used to configure the default configuration.

The search command starts a video search from one or multiple given starting points. These could be multiple videos orginating from a search (using term), or a provided list of video ids (using input), or just one root video (using id or url).

For example, the following command will return the first video when one searches for cat.

$ yt-scraper search term 'cat'
[STATUS] Result:
 Depth: 0, Rank: 0, ID: hY7m5jjJ9mM
            Title: CATS will make you LAUGH YOUR HEAD OFF - Funny CAT compilation
            Related Videos: []

One can also provide a video id or a video url as a starting point, which is more interesting when combined with the --max-depth option:

$ yt-scraper search id '0A2R27kCeD4' --max-depth 2
 Depth: 0, Rank: 0, ID: 0A2R27kCeD4
            Title: 🤣 Funniest 🐶 Dogs and 😻 Cats - Awesome Funny Home Animal Videos 😇
            Related Videos: ['pc8-8KfIW5c']
     Depth: 1, Rank: 0, ID: pc8-8KfIW5c
                Title: 🦁 Funniest Animals 🐼 - Try Not To Laugh 🤣 - Funny Domestic And Wild Animals' Life
                Related Videos: ['OrJMUNEyZsE']
         Depth: 2, Rank: 0, ID: OrJMUNEyZsE
                    Title: Funniest Videos for Pets to Watch Compilation | Funny Pet Videos
                    Related Videos: []

Additionally, one can specify the number of videos that should be returned on each level by utilizing the --number option. For instance, the following command returns two related videos from a given video (specified by it's url) and additionally one related video from each sibling:

$ yt-scraper search url 'https://www.youtube.com/watch?v=0A2R27kCeD4' --depth 1 --number 2 --number 1
 [STATUS] Result:
  Depth: 0, Rank: 0, ID: 0A2R27kCeD4
             Title: 🤣 Funniest 🐶 Dogs and 😻 Cats - Awesome Funny Home Animal Videos 😇
             Related Videos: ['pc8-8KfIW5c', 'tbyAuT50eu4']
      Depth: 1, Rank: 0, ID: pc8-8KfIW5c
                 Title: 🦁 Funniest Animals 🐼 - Try Not To Laugh 🤣 - Funny Domestic And Wild Animals' Life
                 Related Videos: []
      Depth: 1, Rank: 1, ID: tbyAuT50eu4
                 Title: 😁 Funniest 😻 Cats and 🐶 Dogs - Awesome Funny Pet Animals 😇
                 Related Videos: []

For the sake of brevity, you can shorten --number to -n and --depth to -d.

There are even global commands, too! Global options are specified in front of the command and alter the behavior of all commands. This may not sound very meaningful to you given that there only two commands right now and you are right! But this is likely to change in the future.

For example, to see more output during the program execution, specify --verbose or -v right after yt-scraper:

$ yt-scraper -v search id '0A2R27kCeD4' --max-depth 2

There are many more options that you can make use of. All of them are described in the Options section.

Sometimes you may find yourself struggling with all the possible options. Fortunately, there is the config command for all the lazy typer out there.

Setting a particular default option like the output directory to ~/my_data is as easy as typing

$ yt-scraper config set encoding utf-8

Forgetful? Just double-check by typing get instead of set:

$ yt-scraper config get encoding
[STATUS] The value of 'encoding' is set to 'utf-8'.

Configuration

Search options Default Description
-n, --number 1 Number of the videos fetched per level (can be specified several times)
-d, --max-depth 0 Number of recursion steps to perform.
-k, --api-key Required The API key that should be used to query the YouTube API v3.
-o, --output-dir Optional Path to the directory where output files are saved
-f, --output-format csv Specifies the file format of output files.
-N, --output-name Optional Specifies the file name or prefix of output files.
-r, --region-code de Return only videos which are unrestricted in the given region.
-l, --lang-code de Return videos mostly relevant to a specified language.
-s, --safe-search none Filter sensitive or restricted videos.
-e, --encoding utf-8 Transform fetched text to another encoding.
-u, --unique False Do not process seen videos again
-i, --include All Specify field to export (can be used several times)
-x, --exclude None Specify field not to export (can be used several times)
Global options Default Description
-c, --config-path System-specific Specifies a configuration file. For details, see configuration.
-v, --verbose False Shows more output during program execution.
-V, --version Optional Shows the current program version and exits

More information can be found by adding the --help option to commands or reading the YouTube API manual.

Old-fashioned people, who do not like the config command, can manually configure the program by editing the config.toml file. It is secretly used and altered when using the config command. Entered values are used in all future queries as long as they are not overwritten by actual command line options.

For example, to always use the API key ABCDEFGH and a search depth of 3, where on each level one video less is returned, just create following configuration file:

config.toml

api_key = "ABCDEFGH"
number = [ 4, 3, 2, 1 ]
depth = 3
verbose = true

An example toml is included: config.toml

Then put this file in your standard configuration folder. Typically this folder can be found at the following location:

  • Mac OS X: ~/Library/Application Support/YouTube Scraper
  • Unix: ~/.config/youtube-scraper
  • Windows: C:\Users\<user>\AppData\Roaming\YouTube Scraper

If the folder does not exist, you may need to create it first.

Quota costs

Google dramatically lowered the maximum number of API requests per day in recent years. Currently, they impede a daily usage of 10,000 quota points, which corresponds to around 100 videos. One can create multiple accounts though, and specify them in the config.toml file like this:

api_key = ["<API Key 1>", "<API Key 2>", "..."]

Furthermore, specify the --unique option to avoid duplicate processing of frontier nodes. Keep in mind, however, that the resulting graph will have a maximum degree of n.

Release History

  • 0.2.X

  • 0.3.X

  • 0.4.X

    • New command search
  • 0.5.X

    • Option --depth renamed to --max-depth
    • Video attributes, such as title, description, channel are fetched
    • More consistent option handling
  • 0.6.X

    • New export feature: csv
    • New command: config
    • New API options: region-code, lang-code and safe-search
  • 0.7.X

    • New --version option
    • New --encoding option
    • New --export-name option
    • New --unique option
    • New input method by importing a file or reading from stdin
    • Added prompt when encountering an API error
  • 0.8.X

    • New --format sql SQLite export
    • New config where command
    • Multiple API keys with automatic key switching is now possible
    • New --include option
    • New --exclude option
    • Renamed input argument to file
    • Piping of urls is now possible

Roadmap

Every of these features is going to be a minor patch:

  • Add node video data attributes, such as title and description.
  • Add possibility to specify more than one API key to switch seamlessly.
  • Add possibility to query more than 50 videos on one level.
  • Add youtube-dl integration for downloading subtitles.
  • Add a testing suite.
  • Add export functionality to CSV and SQLlite.
  • Add more information about quota to README

Contributing

If you come across any bugs or have a suggestion, please don't hesitate to file an issue.

Contributions in any form are welcomed. I will accept pull-requests if they extent yt-scraper's functionality.

To set up the development environment, please install Poetry and run poetry install inside the project. A test suite will be added soon.

In general, the contribution process is somewhat like this:

  1. Fork it ($ git clone https://github.com/rattletat/yt-scraper)
  2. Create your feature branch ($ git checkout -b feature/fooBar)
  3. Commit your changes ($ git commit -am 'Add some fooBar')
  4. Push to the branch ($ git push origin feature/fooBar)
  5. Create a new Pull Request

Author

Michael Brauweiler

License

This plugin is free and unemcumbered software released into the public domain.

For more information, see the included UNLICENSE file.