Subtitle Tools

Online tools to sync, fix and convert subtitle files.

The following repositories are part of Subtitle Tools:

• https://github.com/SjorsO/sup

Installation

Required php extensions:

• gd for working with images extracted from sup files

Required binaries:

• Vobsub2srt for converting sub/idx to srt
• Uchardet v0.0.6 for detecting text encoding
• Tesseract v3 for OCR'ing sup images
  • Tesseract v3 traineddata

Installation

cp .env.example .env

# Fill in the .env file

composer install

php artisan key:generate
 
php artisan migrate
 
npm i && npm run dev

Laravel version

The following link can be used to keep the application in sync with Laravel's master branch: compare to laravel master

General information

• Language codes use ISO 639-2 (same as Tesseract)

Queues

Subtitle Tools runs four separate queue workers. The supervisor config that keeps these workers running is listed below. The queue workers that handle sub/idx and sup OCR jobs are nice. OCR'ing is a CPU heavy job, it will consume all available processing power of the thread it is running on. Making them nice ensures that web requests and other more important work is given priority.

Supervisor config

[program:st-worker-default]
process_name=%(program_name)s_%(process_num)02d
command=php /var/www/st/current/artisan queue:work --queue=default,A500 --sleep=2 --tries=1
numprocs=3
autorestart=true
user=www-data

[program:st-worker-broadcast]
process_name=%(program_name)s_%(process_num)02d
command=php /var/www/st/current/artisan queue:work --queue=broadcast --sleep=2 --tries=2
autorestart=true
user=www-data

[program:st-worker-1]
process_name=%(program_name)s_%(process_num)02d
command=nice php /var/www/st/current/artisan queue:work --queue=A100,A200,A300,A400,A500 --sleep=2 --tries=1
autorestart=true
user=www-data

[program:st-worker-2]
process_name=%(program_name)s_%(process_num)02d
command=nice php /var/www/st/current/artisan queue:work --queue=B200,A100,A200,A300,A400,A500 --sleep=2 --tries=1
autorestart=true
user=www-data

Processing all jobs

While debugging, you can run the following command to process all jobs:

artisan queue:work --queue=broadcast,default,B200,A100,A200,A300,A400,A500 --sleep=2 --tries=1

Jobs per queue

The list below shows which queue runs which jobs. The order in which the queues are handled can be found in the supervisor config above.

Jobs per queue

• default
  • All FileJobs
• broadcast
  • All events
• A100
  • BuildSupSrtJob
• A200
  • ExtractSupImagesJob
• A300
  • OcrImageJob
• A400
  • OcrImageJob (when the job in queue A300 takes too long, it is re-dispatched on this lower priority queue with a higher timeout)
• A500
  • RecordUploadedFileMimesJob and other low priority diagnostic jobs
• A500
  • CollectSupMetaJob
  • CollectStoredFileMetaJob
• B200
  • ExtractSubIdxLanguageJob

Adding a FileGroup + FileJob tool

• Make a new controller that extends FileJobController, add routes and views
• Make a new job that extends FileJob

Adding a new plain-text format

• Create a NewFormat class that extends the abstract TextFile
• Make it use either the WithFileContent trait or the WithFileLines trait
• If it can be converted to Srt, implement the TransformsToGenericSubtitle interface
• Add the NewFormat class to the $formats array in \App\Subtitles\TextFileFormat

Contributing

Feel free to open a pull request if you want to add a new feature, or if you want to help improve the code, design, text, seo, or any other part of the website.

If you want want to discuss an idea before implementing it, you can contact me by email, twitter, or by opening an issue.

License

This project is open-source software licensed under the MIT license