srt-translator

Batch translate subtitles srt files using deepL api

• minimise number of characters to translate by removing sequence number, timecode and can optionally remove HTML tags and hearing impaired annotations
• translation is done on concatenated packet of subtitles (1000 lines). It preserves context by avoiding line by line translation, improve translation execution time and avoid exceeding deepl text length limit
• automatically fix some punctuation glitch after translation

/!\ you must request a free or paid deepL api key on the deepL web-site /!\

Developed by Fabrice Deshayes for my friend Didier Martini

installation

prerequisites

• miniconda (which install conda + python + pip)
• git

You can check that all is ok by checking conda, python, pip and git version

conda --version
conda 23.5.2

python --version
Python 3.10.12

pip --version                                                                                                                                            ─╯
pip 23.1.2 from C:\Users\fabri\scoop\apps\miniconda3\current\lib\site-packages\pip (python 3.10)

git --version
git version 2.41.0.windows.2

setup

clone project somewhere on you disk

git clone https://github.com/Fabrice-Deshayes-aka-Xtream/srt-translator.git

using powershell on Windows terminal, go inside srt-translator folder and execute setup.ps1 script. you can also right-click "execute with powershell".

cd srt-translator
.\setup.ps1

that's all folks! this setup script has created a dedicated conda environment for this project and install all required libraries using pip. No need to activate the environment, it will be done automatically by using the run script.

usage

put your .srt files to translate into the batch/todo folder

using powershell on Windows terminal, go inside srt-translator folder and execute run.ps1 script. you can also right-click "execute with powershell".

cd srt-translator
.\run.ps1

Example of execution output

you can request a free deepL api key on the deepL web-site

After execution, processed files have be moved to batch/done folder, and the translation result files are in batch/result folder.

You can also put your file in batch/planned and launch the program which count the number of chars to translate in files

cd srt-translator
.\run-count.ps1

configuration

some variables can be changed on config.py file if needed. please note that only deepl translation_engine is working. Microsoft one's has not been tested yet.

# plannedPath is the place where your srt files planned to be processed are located to count number of chars
plannedPath = "batch/planned"

# todoPath is the place where your srt files which need to be translated are located
todoPath = "batch/todo"

# each files in todoPath will be translated using the defined targetLang and the result will be put in resultPath
# result files have the same name as input file with targetLang suffix
resultPath = "batch/result"

# files found in todoPath are moved to donePath after translation
donePath = "batch/done"

# process only files with this extension
fileExt = "*.srt"

# source lang of file to translate (keep None for automatic lang detection)
sourceLang = None

# target lang for translation (source lang is automatically detected)
targetLang = "FR"

# rename file in todopath with origin lang suffixed (detected or forced)
suffixDoneFileWithDetectedLang = True

# rename file translated with target lang suffixed
suffixResultWithTargetLang = False

# define encoding for input files (see https://docs.python.org/3/library/codecs.html#standard-encodings)
input_encoding = "utf_8_sig"  # default is utf_8_sig (utf-8 with BOM)

# define encoding for result files (see https://docs.python.org/3/library/codecs.html#standard-encodings)
result_encoding = "utf_8_sig"  # default is utf_8_sig (utf-8 with BOM)

# remove deaf annotations from translated result (text inside parenthesis (...) or hooks [...])
removeDeafAnnotations = True

# remove HTML/XML tags from subtitles before translation (text like <...>)
removeTags = True

# formality possible values
# default (default)
# more - for a more formal language
# less - for a more informal language
# prefer_more - for a more formal language if available, otherwise fallback to default formality
# prefer_less - for a more informal language if available, otherwise fallback to default formality
formality = "prefer_less"