How to make sense from these rubbish subtitles

[Gautier]

Hello, I just ripped a subtitle off a release. But look at the garbage I'm getting: https://copy.com/TQ8jTLDUIfFZ7yeb/bbc.srt?download=1

Two different ways of extracting gave me the same results. Every word of every line is added one at a time. Is there a tool that can make sense of this, and collapse the individual lines?

Code: Select all

1 00:00:06,680 --> 00:00:06,920 Britain's 2 00:00:06,920 --> 00:00:07,160 Britain's railway 3 00:00:07,160 --> 00:00:07,520 Britain's railway - 4 00:00:07,520 --> 00:00:07,760 Britain's railway - the 5 00:00:07,760 --> 00:00:07,920 Britain's railway - the oldest 6 00:00:07,920 --> 00:00:08,120 oldest and 7 00:00:08,120 --> 00:00:08,320 oldest and one 8 00:00:08,320 --> 00:00:08,640 oldest and one of 9 00:00:08,640 --> 00:00:08,920 oldest and one of the 10 00:00:08,920 --> 00:00:09,120 oldest and one of the busiest 11 00:00:09,120 --> 00:00:09,600 oldest and one of the busiest in 12 00:00:09,600 --> 00:00:09,760 oldest and one of the busiest in the

[MarcoTC]

Instead of trying to repair it, maybe you should find a better way to rip it.
What do you use now?

[vankasteelj]

That would only take a few lines of script, if there is no tool to "collapse" those, I can wrap up a node.js script pretty easely to do so (but I rather not if there already is a tool for it^^). PM if needed, not sure i'll see an answer here.

[Gautier]

Hi MarcoTC. The rip is not at fault, it's the source file which seems to be like this. So the rip just extracts the garbage which is already present. Still... I have no clue how they managed to come up with a subtitle file like this one. Some piece of subtitling or encoding software must have screwed up quite badly.

No problems vankasteelj, I already found another file which I downloaded directly from the BBC. Just have to resync it now... But it would be nice to find a software that can handle these kind of situations. Subtitles from French television suffer from similar duplicates as well. Example:

Code: Select all

2 00:00:03,560 --> 00:00:03,720 "celle qui permet de faire le bonheur autour de toi, 3 00:00:03,720 --> 00:00:07,520 "celle qui permet de faire le bonheur autour de toi, 4 00:00:08,400 --> 00:00:08,480 "la source d'or que tu as cherchée aux 4 coins de la terre, 5 00:00:08,480 --> 00:00:12,280 "la source d'or que tu as cherchée aux 4 coins de la terre,

[vankasteelj]

From a technical point, those are really easy-to-solve problems, in any language. I do not know C# so I can't do it myself, but a plugin for https://github.com/SubtitleEdit/subtitleedit would be a great idea.

[Gautier]

In the meantime, someone contacted me with the solution. Subtitle Edit has an option Tools>Merge Lines with same text/same time codes. Works a miracle!

[vankasteelj]

That's what I thought, it's so easy to make (regarding code) that it seems perfectly normal to me it already exists^^