script monkeys, a hand please

pickarooney

I have a subtitle file for an AVI which is out of sync by 20 or so seconds, or 840 frames to be more precise. Therefore, I need to edit the frame numbers listed in the subtitle file (it's a textfile).
Simple enough task, or so I thought, until I got bogged down in the pleasantries of sed/awk.

The two sets of numbers in the sample below correspond to the beginning and end frames for displaying the subtitles. What I need to do is decrement all these numbers by 840.

{148360}{148390}Say it!
{148395}{148472}I won't come out|until nobody is around.
{148477}{148561}Good boy, stubborn.
{149220}{149252}Leave!
{149257}{149305}Go away, dog!
{149310}{149356}Go, dog!
{149361}{149430}Leave.
{149705}{149741}Good job, Ferruccio.
{149746}{149829}It works.
{150385}{150471}Is there anyone|called Dora here?

The first line should thus read:

{147520}{147550}Say it!

I'll probably want to be able to run the script on other files, with a different decrement/increment, so the command line will be like:

resynch_sub.sh 220 old.sub new.sub
resynch_sub.sh -840 old.sub new.sub

A bonus point for anyone who can name the film

Kev

Heres some perl to do it if thats any use.

#!/usr/local/bin/perl -Wp

s/{(\d+)}{(\d+)}/sprintf("{%d}{%d}", $1 - 840, $2 - 840)/e;

__END__


resynch_sub.pl old.sub > new.sub

pickarooney

I don't actually have perl installed - I'm trying to run this through a bash shell on windows. I can try installing a compiler though, I suppose.
Cheers.

davej

Ok off the top of my head (just built the expression as i went along so it's not efficient):


cat filename| sed -e 's/[{}]/ /g'|awk '{print "{" $1-840 "} {"$2-840"} "$0}'|awk 'sub($3 "[ ]*" $4 "[ ]*", "", $0)'

This with GNU Awk 3.1.0 and GNU sed version 3.02

davej

pickarooney

You beauty! Thanks a million, Dave

MiCr0

la vita bella?

pickarooney

Yup. You now have 1 pickarooney point

krinDar

One line awk:

awk -F} '{printf("{%s}{%s}%s\n",(substr($1,2)-840),substr($2,2)-840,$NF)}' filename

In davej's version the cat is redundant, instead of doing:
cat <filename> | sed,
you can just do
sed 'command' <filename>