vidl

vidl is a script designed to easily download video/audio from anywere, using yt-dlp. It automatically embeds thumbnails to mp3/mp4/m4a files.

vidl will add metadata to mp3 files if it's found. The --no-md option turns this off. title, artist and year metadata is added, but if the URL is a playlist, it also adds album, album artist, track number, track count. If the title contains " - ", vidl often uses what comes before and after it as artist and title respectively. The --dont-extract-md option turns off this behaviour.

Installation

1. Install Python (3.7+ is recommended)
2. Install ffmpeg and ffprobe
3. Run pip3 install vidl.
4. If you're not on macOS or Windows, you need to specify where vidl will download files to in your vidl config file. Run vidl --config-path to see where the config file is. If you're on macOS, I recommend setting up shortcuts for vidl

Updating

vidl is updated by running pip3 install vidl --upgrade --upgrade-strategy eager. If vidl is unable to download a URL, it might be because yt-dlp is outdated. The --upgrade-strategy eager part updates yt-dlp.

Uninstall

Run pip3 uninstall vidl.

To fully uninstall vidl, go to vidl's config file (Run vidl --config-path to see where it is) and delete the folder it's in.

Usage

Examples: vidl https://www.youtube.com/watch?v=ta_ZVS7HkwI

• Downloads the video as mp3, and adds metadata it detects.

vidl mp3 https://www.youtube.com/watch?v=ta_ZVS7HkwI --no-md

• Downloads the video as mp3, without adding metadata.

vidl

• Prints vidl's help menu, which looks like this:

  Download Usage:
      vidl [format] [options] <URL>
  
  Download Options:
      format             bestvideo, bestaudio (default), mp3, mp4, wav, m4a, or opus
      --no-md            Don't add metadata to downloaded files
      --no-smart-md      Don't extract artist and song name from title
      --no-embed         Don't embed thumbnail
      --no-dl            Don't download anything. Usually used with -v
      -v, --verbose      Display all logs
  
  General Options:
      --version          Show version. vidl -v and vidl version works too
      -h, --help         Show this help message. vidl help works too
      --config-path      Show the location of the configuration file
  
  Update:
      pip3 install vidl --upgrade --upgrade-strategy eager
  

Configuration

vidl has a configuration file, which you can find the location of by running vidl --config-path. In it, you can set the download folder and filename template.

If you screw something up, you can delete the file, and the default will be recreated the next time you run vidl.

Custom metadata parsing

vidl has a user_md_parser.py file. By default, it does nothing, but you can configure it to manipulate metadata of songs you download however you like. In my case, I set the "Comment" metadata to "NCS" if the title ends with "[NCS Release]".

Documentation for this can be found in the file itself. The file is in the same folder as vidl's config file, which you can find by by running vidl --config-path. If you screw something up, you can delete the file, and the default will be recreated the next time you run vidl.

Set up shortcuts for vidl (macOS)

You'll be able to select any piece of text, press your chosen shortcut and the link(s) in your selected text will be downloaded! A little tedious to set up, but well worth it.

First, we need to create a macOS Service:

1. Open the Automator app.

2. Choose File > New, and select Service.

3. (TLDR; Add Run Shell Script) In the window that just popped up, there are two columns on the left (if not, click the Library button in the status bar). Select Utilities in the first column, and in the second column, drag Run Shell Script into the main part of the window.

4. Make your settings match these:

   

   If you want the shortcut to only work in one app, select that app instead of any application.

5. In the text box in the "Run Shell Script" box, paste in the following script:

   for f in "$@"
   do
     # behave like a normal terminal window:
     export PATH=/usr/local/bin:$PATH
     source ~/.bash_profile
   
     vidl "$f"
   done

6. Choose File > Save. Type in vidl.

Almost done, you just need to tie a shortcut to the macOS Service you just created:

1. Open the System Preferences app.
2. Go to Keyboard and select the Shortcuts tab.
3. Select Services from the left column, and locate vidl (should be under Internet). Add your preferred shortcut.

API Usage

You can use Python to download a URL. Example:

from vidl import config, dl

def main():
    config.verify_config()
    dl.download({
        # all options are required
        'url': 'https://www.youtube.com/watch?v=82IZ63TU8Fs'
        'file_format': 'mp3',
        'audio_only': True,
        'no_md': False,
        'no_thumbnail_embed': False,
        'no_smart_md': False,
        'no_dl': False,
        'verbose': False,
        'download_folder': config.get_config('download_folder'),
        'output_template': config.get_config('output_template'),
    })
main()

Dev Instructions

Dev Installation

1. Install Python. You may want to install it using pyenv in order to manage Python versions (If Poetry doesn't detect the right version, you can fix it with pyenv).
2. Install ffmpeg and ffprobe
3. Install Poetry
4. Run poetry install to install Python package dependencies.
5. If you're not on macOS or Windows, you need to specify where vidl will download files to in your vidl config file. Run vidl --config-path to see where the config file is.

I recommend running poetry config virtualenvs.in-project true, which makes Poetry store your Python virtual environment inside the project folder. Additionally, it lets VSCode's Python extension detect the virtual environment if you set the python.pythonPath setting to ${workspaceFolder}/.venv/bin/python in your settings.

Running

poetry run vidl

An alternative to poetry run <command> is to run poetry shell to enter the virtual environment's Bash CLI, and then run the command on it's own (e.g vidl).

Releasing a new version

First of all, consider updating the lockfile dependencies by running poetry update, then check if things still work.

1. Bump the version number. <version> can be patch, minor, major or a version number:

   poetry version <version>
   

2. Update CHANGELOG.md
3. Build:

   poetry build
   

4. Commit and create a git tag
5. Create GitHub release with release notes and attach the build files
6. Publish to PyPI:

   poetry publish