[go: up one dir, main page]

Skip to content

A Python module to transform subtitle line lengths, splitting into multiple subtitle fragments if necessary.

License

Notifications You must be signed in to change notification settings

peterk/srt_equalizer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

OpenSSF Scorecard PyPI - Downloads

SRT Equalizer

A Python module to transform subtitle line lengths, splitting into multiple subtitle fragments if necessary. Useful to adjust automatic speech recognition outputs from e.g. Whisper to a more convenient size.

This library works for all languages where spaces separate words.

Installing

pip install srt_equalizer

Example

An SRT file containing lines over a certain length can be adjusted to a maximum line length for better readability on screen.

1
00:00:00,000 --> 00:00:04,000
Good evening. I appreciate you giving me a few minutes of your time tonight

2
00:00:04,000 --> 00:00:11,000
so I can discuss with you a complex and difficult issue, an issue that is one of the most profound of our time.

To adjust line length to a maximum length of 42 chars you can use SRT equalizer like this:

from srt_equalizer import srt_equalizer

srt_equalizer.equalize_srt_file("test.srt", "shortened.srt", 42)

...they are split into multiple fragments and time code is adjusted to the approximate proportional length of each segment while staying inside the time slot for the fragment.

1
00:00:00,000 --> 00:00:02,132
Good evening. I appreciate you giving me

2
00:00:02,132 --> 00:00:04,000
a few minutes of your time tonight

3
00:00:04,000 --> 00:00:06,458
so I can discuss with you a complex and

4
00:00:06,458 --> 00:00:08,979
difficult issue, an issue that is one of

5
00:00:08,979 --> 00:00:11,000
the most profound of our time.

Algorithms

By default, this script uses greedy algorithm which splits the text at the rightmost possible space.

An alternative splitting algorithm is halving which will split longer lines more evenly instead of always trying to use maximum line length. This prevents producing lines with isolated word remainders.

Another alternative is the punctuation algorithm that takes punctuation (commas, periods, etc.) into account.

from srt_equalizer import srt_equalizer

# use "greedy", "halving" or "punctuation" for the method parameter
srt_equalizer.equalize_srt_file("test.srt", "shortened.srt", 42, method='halving')

Adjust Whisper subtitle lengths

Is is also possible to work with subtitle items produced from Whisper with the following utility methods:

split_subtitle(sub: srt.Subtitle, target_chars: int=42, start_from_index: int=1) -> list[srt.Subtitle]:

whisper_result_to_srt(segments: list[dict]) -> list[srt.Subtitle]:

Here is an example of how to reduce the lingth of subtitles created by Whisper. It assumes you have an audio file to transcribe called gwb.wav.

import whisper
from srt_equalizer import srt_equalizer
import srt
from datetime import timedelta

options_dict = {"task" : "transcribe", "language": "en"}
model = whisper.load_model("small")
result = model.transcribe("gwb.wav", language="en")
segments = result["segments"]
subs = srt_equalizer.whisper_result_to_srt(segments)

# Reduce line lenth in the whisper result to <= 42 chars
equalized = []
for sub in subs:
    equalized.extend(srt_equalizer.split_subtitle(sub, 42))

for i in equalized:
    print(i.content)

Contributing

This library is built with Poetry. Checkout this repo and run poetry install in the source folder. To run tests use poetry run pytest tests.

To build a new release, create a new tag, build it and publish to pypi:

poetry run pytest tests
git tag v0.1.2
poetry build
poetry publish

If you want to explore the library start a poetry shell.

About

A Python module to transform subtitle line lengths, splitting into multiple subtitle fragments if necessary.

Topics

Resources

License

Security policy

Stars

Watchers

Forks

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •  

Languages