Repeated text in extracted closed captions

[Hakunamatata67]

Hi,

I extracted CC from a video file, but the beginning of every subtitles is the same text as the last one:

12
00:00:18,085 --> 00:00:20,187
un jeu terriblement intrigant,
délicieusement tordu
et dangereusement addictif,

13
00:00:20,187 --> 00:00:21,922
délicieusement tordu
et dangereusement addictif,
un jeu à grand déploiement

14
00:00:21,922 --> 00:00:23,724
et dangereusement addictif,
un jeu à grand déploiement
qui a pris le monde entier

15
00:00:23,724 --> 00:00:25,526
un jeu à grand déploiement
qui a pris le monde entier
par surprise, et qui débarque

16
00:00:25,526 --> 00:00:27,327
qui a pris le monde entier
par surprise, et qui débarque
maintenant ici, chez nous.

17
00:00:27,327 --> 00:00:29,129
\hpar surprise, et qui débarque
\hmaintenant ici, chez nous.
Bienvenue dans mon manoir.

18
00:00:29,129 --> 00:00:30,163
\hmaintenant ici, chez nous.
Bienvenue dans mon manoir.
- Le jeu, il est

So is it possible to"clean" the subtitles? "Merge lines with same text" in Subtitle Edit doesn't remove it all.

[2nHxWW6GkN1l916N3ayz8HQoi]

One way to do it is using a custom script. From the sample you posted, could you write what a desired outcome would look like?

[Hakunamatata67]

I think deleting the first two sentences of each line would do. For example:

12
00:00:18,085 --> 00:00:20,187
un jeu terriblement intrigant,
délicieusement tordu
et dangereusement addictif,

13
00:00:20,187 --> 00:00:21,922
un jeu à grand déploiement

14
00:00:21,922 --> 00:00:23,724
qui a pris le monde entier

15
00:00:23,724 --> 00:00:25,526
par surprise, et qui débarque

16
00:00:25,526 --> 00:00:27,327
maintenant ici, chez nous.

17
00:00:27,327 --> 00:00:29,129
Bienvenue dans mon manoir.

18
00:00:29,129 --> 00:00:30,163
- Le jeu, il est

[Subtitles]

Use SubtitleEdit to replace the repeated sentence (assuming they are not relevant)
Edit > Replace >Find What (paste: un jeu terriblement intrigant,) > Replace with: (Leave Blank)
Then go either Replace or
Replace All

Repeat if you have other repeated irrelevant sentences

[Hakunamatata67]

The issue is that ALL sentences are repeated in the next line.

[2nHxWW6GkN1l916N3ayz8HQoi]

I almost wrote a solution but I don't understand what's this

16
00:00:25,526 --> 00:00:27,327
qui a pris le monde entier
par surprise, et qui débarque
maintenant ici, chez nous.

17
00:00:27,327 --> 00:00:29,129
\hpar surprise, et qui débarque
\hmaintenant ici, chez nous.
Bienvenue dans mon manoir.

\h ???

Edit: No idea how that got there but I just removed it. The script is

import re

import pysubs2


def remove_intersection(list1, list2):
    if len(list1) == 0 or len(list2) == 0:
        return list1, list2

    l1 = list(reversed(list1))
    l2 =[list2[0:i2] for i2 in range(1, len(list2) + 1)]

    max_len = 0
    for i1 in range(1, len(l1) + 1):
        l0 = list(reversed(l1[0:i1]))

        if l0 in l2:
            max_len = len(l0)

    if max_len == 0:
        return list1, list2
    return list1, list2[max_len:]


if __name__ == '__main__':
    subs = pysubs2.load("input.srt", encoding="utf-8")
    new_subs = [subs[0]]

    for index in range(0, len(subs) - 1):
        s1 = subs[index]
        d1 = re.split("\\\\N", s1.text, flags=re.IGNORECASE)
        d1 = [re.sub("\\\\[A-Z]", "", dialogue, flags=re.IGNORECASE) for dialogue in d1]
        s2 = subs[index + 1]
        d2 = re.split("\\\\N", s2.text, flags=re.IGNORECASE)
        d2 = [re.sub("\\\\[A-Z]", "", dialogue, flags=re.IGNORECASE) for dialogue in d2]

        lst1, lst2 = remove_intersection(d1, d2)
        new_sub = subs[index + 1].copy()
        new_sub.text = "\\N".join(lst2)
        new_subs.append(new_sub)

    for index in range(0, len(subs)):
        subs[index] = new_subs[index]
    subs.save("output.srt")

Content of input.srt

12
00:00:18,085 --> 00:00:20,187
un jeu terriblement intrigant,
délicieusement tordu
et dangereusement addictif,

13
00:00:20,187 --> 00:00:21,922
délicieusement tordu
et dangereusement addictif,
un jeu à grand déploiement

14
00:00:21,922 --> 00:00:23,724
et dangereusement addictif,
un jeu à grand déploiement
qui a pris le monde entier

15
00:00:23,724 --> 00:00:25,526
un jeu à grand déploiement
qui a pris le monde entier
par surprise, et qui débarque

16
00:00:25,526 --> 00:00:27,327
qui a pris le monde entier
par surprise, et qui débarque
maintenant ici, chez nous.

17
00:00:27,327 --> 00:00:29,129
\hpar surprise, et qui débarque
\hmaintenant ici, chez nous.
Bienvenue dans mon manoir.

18
00:00:29,129 --> 00:00:30,163
\hmaintenant ici, chez nous.
Bienvenue dans mon manoir.
- Le jeu, il est

Content of output.srt

1
00:00:18,085 --> 00:00:20,187
un jeu terriblement intrigant,
délicieusement tordu
et dangereusement addictif,

2
00:00:20,187 --> 00:00:21,922
un jeu à grand déploiement

3
00:00:21,922 --> 00:00:23,724
qui a pris le monde entier

4
00:00:23,724 --> 00:00:25,526
par surprise, et qui débarque

5
00:00:25,526 --> 00:00:27,327
maintenant ici, chez nous.

6
00:00:27,327 --> 00:00:29,129
Bienvenue dans mon manoir.

7
00:00:29,129 --> 00:00:30,163
- Le jeu, il est

Install python
https://www.python.org/downloads/
and pysubs

pip install pysubs2

Put the input.srt file with your subtitles in it and run the script near that file with

python script.py

[Subtitles]

The issue is that ALL sentences are repeated in the next line.

Yeah I realized that.
Can you tell us which application did you use to extract the subtitles?
Perhaps a different application would have given you a better solution.
Try to load the video file using Subtitles Edit. If it finds subtitles it will OCR it or display them all immediately.

[Hakunamatata67]

I almost wrote a solution but I don't understand what's this]

Wow thank you so much, I'll try it

[Hakunamatata67]

The issue is that ALL sentences are repeated in the next line.

Yeah I realized that.
Can you tell us which application did you use to extract the subtitles?
Perhaps a different application would have given you a better solution.
Try to load the video file using Subtitles Edit. If it finds subtitles it will OCR it or display them all immediately.

The CC subtitles are extracted from a MKV file as in this topic. I had to use clever FFmpeg-GUI as gMKVExtractGui didn't work.

[pcspeak]

The software hasn't been updated for a few years.
I find it still works, most times, for me.
https://www.videohelp.com/software/ccextractor
This is the command line I use.
ccext.cmd

"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv -o "my video.srt"

Cheers.

[usually_quiet]

i think the video could have "rollup style" closed captions. Although I have no experience using the command line version of CCExtractor, the command line version has some settings that deal with rollup style captions.

[pcspeak]

Hakunamatata67
You can try this collection and see what you get. The first attempt makes for a much LARGER file.

ccext2.cmd

"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv" --dru -o "my video-direct_rollup.srt"
"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv" --norollup -o "my video-norollup.srt"
"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv" --ru1 -o "my video-ru1.srt"
"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv" --ru2 -o "my video-ru2.srt"
"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv" --ru3 -o "my video-ru3.srt"

Cheers.

[Subtitles]

Have you tried to load the video file using Subtitle Edit? If it finds subtitles it will OCR it or display them all immediately.
File > Open > Select the file
You can also try to use Subtitle Edit to create the subtitles from scratch using Subtitle Edit
video > Audio to text (Whisper)
If the audio is of good quality you can expect good results. Any errors can be corrected using the file that you already have.

[Hakunamatata67]

Hakunamatata67
You can try this collection and see what you get. The first attempt makes for a much LARGER file.

ccext2.cmd

"C:\CCExtractor 0.94\ccextractorwinfull.exe" "my video.mkv" --norollup -o "my video-norollup.srt"

Cheers.

Thanks, this one worked for me.