A simple podcast client that runs on the CLI.
Latest commit 4ba3235 May 30, 2016 @rafaelmartins Merge pull request #4 from karkhaz/all-flag
Added --all flag to get all configured podcasts



marrie is a simple podcast client that runs on the Command Line Interface. It is pretty simple and just does the basic tasks: fetch and play podcast chapters listed on a feed.


marrie is available on the PyPI Package Index, and you can install it using pip:

# pip install marrie

If you are a Gentoo user, you can install it using your favorite package manager:

# emerge -av media-sound/marrie


With marrie installed, you'll need a configuration file, with your list of feeds and the commands to be used to fetch and play the podcast chapters.

To create it, just run the main script without arguments:

$ marrie

This command will raise an error, but will create a sample configuration file for you at ~/.marrie. Edit it as follows.

[config] section

The main section of the configuration file.

  • fetch_command: The command used to fetch the chapters. The default command will use wget. If you want to change it, make sure that the variables %(file)s and %(url)s are correctly used on your command.
  • player_command: The command used to play the chapters. The default command will use mplayer. If you want to change it, make sure that the variable %(file)s is correctly used on your command.
  • media_dir: The directory where the chapters will be stored. Defaults to ~/podcasts. A subfolder will be created for each feed.

[podcast] section

The section with the URLs of your feeds. Each key=value pair represents a feed URL. The key is the identifier of the feed (it will be used later to choose the feed to be used by marrie) and the value is the URL of the feed.

my_podcast = http://example.org/feed.rss

You can add as many feeds as you want.

Basic usage

After having marrie installed and configured, you should syncronize your feeds:

$ marrie --sync

If you want to sync a single feed, you just need to provide its identifier:

$ marrie --sync my_podcast

marrie will not provide any auto-sync feature. It is simple! If you want to syncronize periodically, feel free to create a cron job on your system.

With a fresh list of feeds in cache, you'll want to download a chapter. If you just want to get the latest chapter published in a feed, type:

$ marrie --get my_podcast

If you want to download a specific chapter, you'll need to list all the available chapters first:

$ marrie --list my_podcast

Remote files available for "my_podcast" (reverse sorted by date):

    1: chapter_100.mp3 (Fri, 19 Dec 2014 03:00:00 +0000)
    2: chapter_99.mp3 (Fri, 19 Dec 2014 02:00:00 +0000)
    3: chapter_98.mp3 (Fri, 19 Dec 2014 01:00:00 +0000)

If you want to download the chapter_99.mp3, just type:

$ marrie --get my_podcast 2

Where 2 is the numeric identifier of the wanted chapter.

After the download, you'll want to play the chapter. To play the latest chapter fetched by marrie (that isn't exactly the latest chapter published on the feed, it is actually the latest chapter that was downloaded to your filesystem), just type:

$ marrie --play my_podcast

To play a specific chapter (after download it, obviously), you'll need to list the chapters available on your disk:

$ marrie --list my_podcast
Fetched files available for "nageral" (sorted by name):

    1: chapter_1.mp3
    2: chapter_2.mp3
    3: chapter_2.mp3

This is the same command used to list the available chapters to download. The output will contains both lists.

If you want to play chapter_2.mp3, just type:

$ marrie --play my_podcast 2

Where 2 is the numeric identifier of the wanted chapter.

If you are bored with the absence of new chapters of your favorite podcast, you can ask marrie to play a random old chapter for you. It will pick any of the chapters that were previously fetched and will play for you:

$ marrie --play-random my_podcast

That's it. This is pretty much everything that marrie can do for you!


You can send patches to my email address:

rafael@rafaelmartins.eng.br .

Patches should be created against the Git repository:


Any patch that heavily increases the complexity of the script will be rejected!