Skip to content

removes ads from subtitle files cleanly.

Notifications You must be signed in to change notification settings

ChristianMalazarte/subcleaner

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Subcleaner

Subcleaner is a python3 script for removing ads from .srt subtitle files. The script is more sophisticated than a simple search and delete per line and can use different regex profiles for different languages. Once the script have identified ad-blocks they get removed and the remaining blocks get re-indexed.

The script can also determine the language in the script and inform if the actual language doesn't match up to the subtitle language label. This is optional. It uses python langdetect package to detect what actual language is in the subtitle. However, running the language detection program takes a couple of seconds extra (depending on hardware). so if you run a batch job be prepared for the extra time.

works well with Bazarr directly installed and in a docker container.

Installing

Cloning and running with python3 should work. You can also make the script executable, the shebang is already in place

cd /opt

git clone https://github.com/KBlixt/subcleaner.git

cd subcleaner

Then install the default config by testing to run the script with:

python3 ./subcleaner.py -h

Or if you make it executable:

./subcleaner.py -h

the script comes with a default config and default regex profiles for English and Swedish. Read the config section further down for more information about customization.

Windows:

It should be the same method, although you'll have to figure out how to clone the project and install python3.

Bazarr

Unlock the scripts full potential by running it after downloading a subtitle from Bazarr. Enable custom post-processing and use the command:

python3 /path/to/subcleaner/subcleaner.py "{{subtitles}}" -s (note the quotation)

It should work right out the gate provided the paths and permissions are set up correctly. If you wish to enable language checking simply add:

-l {{subtitles_language_code2}}

It doesn't really do anything currently. It logs a WARNING in the logfile if the language doesn't match with the labeled. in the future I hope it'll automatically delete miss-labeled languages and add them to the bazarr blacklist in order to automatically trigger a re-download.

in the bazarr log it should confirm that the script ran successfully or give you an error message that tells you what's wrong. if nothing is output then you've probably set the script path wrong.

Docker

If you run Bazarr in a docker container, as you should, make sure the Bazarr container have access to the script directory. Either mount /opt/subcleaner directly into the container as a volume or install the script inside the Bazarr config directory.

Setup

Install the default config simply by running the script once or copy the default config into the script root directory. With the subcleaner.conf file installed you can modify the settings within it. the config file contains instructions what each of the settings does.

Regex:

Under the regex directory you can set up custom language profiles with regex for ad detection. If you need help to set up custom profiles there is a README in the directory to guide you.

If you make a useful regex profile for a non-default language, PLEASE let me know!

I'll review it and add it to the included default profiles. And it'll help out others that use that language. :)


Thank you :)

Please, If you find any issues or have any questions feel free to open an issue or discussion.


Future (possibly):
  • add functionality to restore false positives more easily and approve deletion of warning blocks.

  • ASS support?

  • Automatic subtitle deletion if language don't match label. (need bazarr to blacklist removed files for this to be implemented)

About

removes ads from subtitle files cleanly.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%