A python script to strip SRT captions file into a non-timed TXT transcript. Provided an SRT file as input, the script preserves the caption blocks structure in an SRT file input, while merging consecutive lines inside each caption block. Then, the script removes all SRT formatting, and produces a clean TXT file without line index markers, timing…

convert-srt-to-transcript-txt.py

import sys
import re

def merge_lines(srt_file):
    # Read in the SRT file
    with open(srt_file, 'r') as f:
        srt = f.read()
    
    # Merge consecutive lines in each caption block
    merged_srt = re.sub(r'\d+\n(\d\d:\d\d:\d\d,\d+ --> \d\d:\d\d:\d\d,\d+\n)?(.*?\n\n)', lambda match: match.group(2).replace('\n', ' ') + '\n', srt, flags=re.S)
    
    return merged_srt

def clean_srt(srt):
    # Remove SRT formatting (index markers, timing lines, speaker indication lines)
    clean_srt = re.sub(r'\d+\n\d\d:\d\d:\d\d,\d+ --> \d\d:\d\d:\d\d,\d+\n', '', srt)
    clean_srt = re.sub(r'^>> ', '', clean_srt, flags=re.M)
    clean_srt = re.sub(r'>> ', '', clean_srt)
    clean_srt = re.sub(r'\n\d+\n', '\n', clean_srt)
    
    return clean_srt

if __name__ == '__main__':
    # Get the SRT file path from the command line arguments
    srt_file = sys.argv[1]
    
    # Merge consecutive lines in each caption block
    merged_srt = merge_lines(srt_file)
    
    # Remove SRT formatting (index markers, timing lines, speaker indication lines)
    clean_srt = clean_srt(merged_srt)
    
    # Write the cleaned text to a new file with the same name as the SRT file, but with the extension changed to .txt
    with open(srt_file[:-4] + '.txt', 'w') as f:
        f.write(clean_srt)