Dockerized Over Drive Loader

odmload is a dockerized tool for downloading audiobooks from Over Drive. It's intended to help you use Over Drive the same way as the old official program.

Before we start anything, make sure you have Libby up and running for all of your libraries. Ask your library staff if you don't know how to do this. This will also activate your Overdrive accounts for each library; you need to know the websites and logins for each individual library. To look them up, go to the central Over Drive site and look up your own libraries. As I mentioned, you will need to be able to log in; this is usually something simple like your library card number and the last 4-5 digits of your phone number, but your library staff will help you.

Make sure you have Docker, Docker Compose, and Python installed. These instructions will assume you're using a Git supported command line; if you're using Windows, you may want to use the Git Bash command line. I'll wait for you to get that all done, and my apologies that you may have to dig a bit.

Clone this repository into a directory of your choice. Github has a guide.

And finally for setting up your machine, pick a place to store the books after they're downloaded, and a place to store them while they're being downloaded. I use Audiobookshelf, so I was able to simply use that for the final path. You will want it to have enough space for the number of books you typically have checked out at a time.

Books downloaded will be placed in the download directory under a subdirectory "libby", with each book stored in a directory named only with its Libby ID (a simple way to keep book folders short and unique - I am open to contributions of code with better suggestions). Naturally, this means if you are using an audiobook manager you will need to give it metadata since it's not in the folder structure. I have implemented Audiobookshelf metadata and (as before) I'm open to taking code contributions for other systems like Plex or Jellyfin.

NOTE: although Libby can report when loans are due, this program does not currently implement automatic deletion of due books. You are responsible for complying with the terms of service, and I have tagged the books as "Odmpy-NG" to make that easy to do.

Run the following command to get the current tested versions of the two dependencies this uses. This will be a patched version of odmpy, which correctly logs into libby and can get the list of books you have checked out; as well as odmpy-ng, which can actually fetch the books.

git submodule update --init --recursive

Use pip to install odmpy from the submodule of the same name. Even if you have a patched version running this update is needed to get library websiteIds.

python3 -m pip install ./odmpy

You'll now need your Libby authentication code. Open a browser and go to your own Libby config page and choose "Copy to another device", then choose one of the devices listed, either Sonos or Android will work, and here's a link.

With that code in mind, run odmpy to get the list of books you have checked out. If it isn't set up yet, not a problem; run it just like this and it will prompt you for the Libby code, and once it's all set up, it'll display your list of books.

Your libby code may time out ... if so, just run it again. No worries.

odmpy libby --reset
odmpy libby

Next, create a config file for odmpy-ng. If you've run odmpy-ng on its own before, you'll want to copy the config file from the previous run; put it into odmpy-ng/config/config.json. Don't worry if you don't have one.

The next step will write to odmpy-ng/config/config.json, so if you have something there, make sure you have a safe copy.

Next, we'll run the configuration tool:

./odmload.py configure

Once that finishes, edit odmpy-ng/config/config.json and make sure you set the login pins for each library (your librarians should be helpful here). Also, you may also want to set some of the encoding options, I always get all of the extra metadata, and I also prefer to skip reencoding because it takes a long time (so that's the least tested part for me).

We're almost there. Now prepare two folders; one a temporary folder for in-progress downloads, and one for completed downloads. I simply point this app to the same folder all of my Audibookshelf collections are in; because this app saves each book in a path that looks like libby/<id>/..., the Libby books don't get mixed up with other sources.

export AUDIOBOOK_TMP=/your/path/to/a/folder
export AUDIOBOOK_FOLDER=/a/path/to/plenty/of/space

Go to libby and check out some books. Run the app and it will download them! If the download works, it'll show up in your download folder. If something goes wrong, it'll keep the files in your tmp folder, including a log file. It'll try again the next time you run it, but if it fails twice in a row, it'll give up and mark the tmp folder with a file named bad.

./odmload.py