How to get a valid milliseconds using Python

I have a subtitle of a two-digit milliseconds, how do I get a valid three-digit milliseconds?

example before

1
00:00:06,10 --> 00:00:08,05.
Let go, let go.

2
00:00:24,21 --> 00:00:29,24.
This is Dateline: Springfield.

3
00:00:30,08 --> 00:00:32,18.
Homer and Marge were

4
00:00:40,03 --> 00:00:45,08.
high school sweethearts.

5
00:00:45,13 --> 00:00:49,05.
She saved me from a dateless prom.

after , using Subtitle Edit It gives you a valid milliseconds

1
00:00:06,417 --> 00:00:08,208
Let go, let go.

2
00:00:24,875 --> 00:00:29,999
This is Dateline: Springfield.

3
00:00:30,333 --> 00:00:32,750
Homer and Marge were

4
00:00:40,125 --> 00:00:45,333
high school sweethearts.

5
00:00:45,542 --> 00:00:49,208
She saved me from a dateless prom.

I used a script pysubs2 https://pypi.org/project/pysubs2/

import pysubs2
subs = pysubs2.load("my_subtitles.srt", encoding="utf-8")
subs.shift(s=0.0)
for line in subs:
    line.text = "{\\be1}" + line.text
subs.save("Output_subtitles.srt")

Only supplies a zero at the end of the timestamp

00:00:06,10 --> 00:00:08,05 to 00:00:06,100 --> 00:00:08,050

Note I know code provides seconds and milliseconds But all I want is to calculate the duration milliseconds
like use Subtitle Edit

Solution

What you want can be done with regular expression and re.sub from the standard python library.

import math
import re


subtitles = """
1
00:00:06,10 --> 00:00:08,05.
Let go, let go.

2
00:00:24,21 --> 00:00:29,24.
This is Dateline: Springfield.

3
00:00:30,08 --> 00:00:32,18.
Homer and Marge were

4
00:00:40,03 --> 00:00:45,08.
high school sweethearts.

5
00:00:45,13 --> 00:00:49,05.
She saved me from a dateless prom.
"""


def modify_sub_match(match):
    fps = 24  # set the value for fps, adjsut for values different from 24!

    # calculate miliseconds based on fps
    miliseconds = math.floor(1000 * int(match.group(2)) / fps) - 1

    # return the match group 1, i.e: xx:xx:xx, concatenate the calculated miliseconds
    # and format with a leading zeros, in case the value is less than 100
    return f"{match.group(1)}{miliseconds:03d}"


# (\d{2}:\d{2}:\d{2},)(\d{2}) is a regular expresion that matches xx:xx:xx,xx
modified_subtitles = re.sub(r"(\d{2}:\d{2}:\d{2},)(\d{2})", modify_sub_match, subtitles)

# print the value of `modified_subtitles`
print(modified_subtitles)

# save `modified_subtitles` to output_subtitles.srt
with open("output_subtitles.srt", "w") as file:
    file.write(modified_subtitles)