[REL] Combined Library Metadata Agent (Colima)

Dev Community Dev/API Corner
1

I made this for myself a while ago, maybe it is of interest for others.

#What is Colima?

Combined Library Metadata Agent (Colima) in combination with Absolute Series Scanner (ASS) allows you to have movies and tv shows within the same library, something that Plex sadly not supports out of the box. Common scenarios to use this agent for are documentaries, western and Japanese animation libraries.

Colima will look for a unique identificator to distinguish movies from tv shows. The agent pulls data from TheMovieDB for all movies it can find this way. TV Shows are entirely ignored by Colima itself and instead are passed as fallback to the native Plex TVDB agent.

The code is mostly justed based on the Plex implementation of TheMovieDb on github.com.

##Download

Download: https://github.com/defract/Colima.bundle/archive/master.zip
Source: https://github.com/defract/Colima.bundle

##Examples

I always like to see results, so here are two mockups of how a library could look like with Colima (created by using fake files).

Documentaries Library

doculibh3k7l.jpg1375×717

Animation Library

animationlibphkaq.jpg1380×722

##Where is the catch?

As already pointed out, Plex does not natively support movies within a tv show library. Putting movies into the tv show library means they will get seasons, some of their data (writers and directors) are shown differently from native movie libraries. Also, all movies within a tv show library are counted as tv shows for obvious reasons.

#Installation

  1. Find out where your Plex installation is located at, see Plex help article.
  2. Install Absolute Series Scanner (ASS)
  3. Download and copy the unzipped Colima.bundle into the Plug-ins folder of your installation

A Plex server restart might be required for the Agent to appear.

#Configuration

Create a TV Show library in the language of your choice

config016arzj.jpg730×535

Movies and TV Shows have to be in separate folders. Ensure that movies are within a folder with a unique name that will not appear in a movie/show title.

config02yeo1c.jpg728×534

Use Absolute Series Scanner (ASS) to create the entries within Plex. Add the folder name to the movie identificator field (only a single value is allowed).

config03cgs5h.jpg731×532

#Notes

  • currently this has only been tested on windows and Unraid (Linux)
  • if your primary interest is Anime, you might want to check out the awesome HAMA, it offers far greater capabilities in this regard

#Acknowledgements

Special thanks to ZeroQI for ASS/HAMA, I wouldn’t have bothered to make my own thing without them.

1 Like
2

I followed the procedure, however the movies will not show in the new Libraries. Does this still work?

3

Do you have any logs?

I use this for my collection and it still works.

4

I can not get Colima to match a documentary called The King for anything

File name is “The King (2017).mkv” I also tried “The King 2017.mkv”

I tried all search options

2P7pqjZ.png536×409 47.5 KB

I even tried “tt6775942” as a title with all search options

Both ASS & Colima are up to date

Any ideas?

5

Hi krawhitham,

I had a look at this and I could reproduce this behavior.

After some debugging I can say the following:

  1. The movie appears to be released in 2018 instead of 2017 (at least according to TMDB).

  2. In the code I partly used from the official TMDB metadata agent, there is a segment to adjust the match score based on the popularity on TMDB, and this segment is the reason your movie is not matched. The King has such a low popularity value that the (otherwise perfect) match gets a huge penalty.

I currently don’t know if I will change this by default, or make it configurable, but I can tell you how you can change it so that it should work for you.

You have to go to the agent folder, on windows this would be: %userprofile%\appdata\local\Plex Media Server\Plug-ins\Colima.bundle\Contents\Code

Open the “init.py” and comment out one line:

# Adjust score slightly for 'popularity' (helpful for similar or identical titles when no media.year is present)
score = score - (5 * i)

and use a “#” in front of the text to comment out (this code will then be ignored) that line, should look like this:

# Adjust score slightly for 'popularity' (helpful for similar or identical titles when no media.year is present)
#score = score - (5 * i)

As an alternative, I will attach a modifed “init.py” that you could use to replace yours.init.zip (17.7 KB)

6

works great

Thank You

1 Like