the getPhrasesFromTranscript function will leave out the last partial phrase if the number of words in the transcript is not divisible by 10. As a result the vtt file will be missing the last 1-9 words from the transcript.

note: bug might also exist in getPhrasesFromTranslation function

The get_time_code function sometimes delivers the microseconds incorrectly causing the following caption to appear before the previous caption is removed.

I corrected the issue by overriding the function with the code below:

Format and return a string that contains the converted number of seconds into WebVTT format

def get_time_code( aws_timecode ):
timedelta = datetime.timedelta(seconds=float(aws_timecode))

hours = (timedelta.seconds // 3600) % 24
minutes = (timedelta.seconds // 60) % 60
seconds = timedelta.seconds % 60
microseconds = timedelta.microseconds / 1000
return str( "%02d:%02d:%02d.%03d" % (hours, minutes, seconds, microseconds ))	   

Thank you for sharing the code. But I cannot find vtt.py file to convert json to vtt file. Any idea?

Hi! I'm interested in using these tools, but I don't see the srt.py file mentioned in the README.