Fix the invalid timestamp strings in the downloaded danmaku ass files

fix_downkyi_ass.py

# Fixes the invalid timestamp string in the downloaded danmaku ass files 
# from Downkyi (https://github.com/leiurayer/downkyi). A pull request (https://github.com/leiurayer/downkyi/pull/219)
# has been opened on the repo. This script can be used to fix the downloaded 
# ass files before it is merged.
# 
# The script has only been tested with limited cases and provided AS-IS.
# The author will NOT be responsible for any data loss for using the script.
#
# Usage: python ./fix_downkyi_ass.py <paths_to_ass_folder>

import sys
import os
import os.path
import re

DIALOGUE_LINE = re.compile(r'Dialogue: ((?:.*?,){9})(.*)')
TIMESTAMP = re.compile(r'([0-9]{1,2}):([0-9]+):([0-9]{2})\.([0-9]{2})')

def fix_ass_timestamp(path):
  print(f'Processing "{path}"...')

  if os.path.isdir(path):
    print(f'{path} is a folder. Will process ass files directly in it.')
    for f in os.listdir(path):
      full_file_path = os.path.join(path, f)
      if f.endswith('.ass') and os.path.isfile(full_file_path):
        fix_ass_timestamp(full_file_path)
    return

  if os.path.isfile(path):
    if not path.endswith('.ass'):
      print('Unsupported file extension.')
      return
    
    buffer = []
    with open(path, 'r', encoding='utf-8') as input:
      for l in input:
        l = l.strip()
        match = DIALOGUE_LINE.fullmatch(l)

        if not match:
          buffer.append(l)
          continue

        (
          layer,
          start,
          end, 
          style,
          name,
          margin_l,
          margin_r,
          margin_v,
          effect,
        ) = match.group(1).strip()[:-1].split(',')
        text = match.group(2)
        new_start = fix_timestamp(start)
        new_end = fix_timestamp(end) 

        buffer.append(
          f'Dialogue: {layer},{new_start},{new_end},{style},{name},{margin_l},{margin_r},{margin_v},{effect},{text}'
        )
    
    with open(path, 'w', encoding='utf-8') as output:
      output.write('\n'.join(buffer))
    
    print('Done.')


def fix_timestamp(time_str):
  match = TIMESTAMP.fullmatch(time_str)
  if not match:
    print(f'Invalid timestamp: {time_str}')
    raise Exception('Error in matching timestamp string.')

  h, m, s, f = match.groups()
  new_m = int(m) % 60
  return f'{h}:{new_m:02d}:{s}.{f}'

if __name__ == '__main__':
  paths = sys.argv[1:]

  for path in paths:
    fix_ass_timestamp(path)