Removing SOME line breaks from srt/txt file - awk

Question

I have a text file which has numbered entries, a timecode and a transcript. I am trying to remove the line breaks in the transcript and leave the others. I'm trying to use grep or awk.
File is like
1
00:00:27,160 --> 00:00:29,054
Sometimes there's not much dialogue.
2
00:00:30,100 --> 00:00:31,090
But other times there is quite a bit,
and it's formatted into two lines
3
00:00:31,500 --> 00:00:33,700
I want to remove the line breaks only on
these long lines, leaving all other formatting.
4
00:00:33,805 --> 00:00:37,285
So that all dialogue ends up being on a single
line no matter how long that line.
Output would look like:
1
00:00:27,160 --> 00:00:29,054
Sometimes there's not much dialogue.
2
00:00:30,100 --> 00:00:31,090
But other times there is quite a bit,
and it's formatted into two lines
3
00:00:31,500 --> 00:00:33,700
I want to remove the line breaks only on
these long lines, leaving all other formatting.
4
00:00:33,805 --> 00:00:37,285
So that all dialogue ends up being on a single line no matter how long that line.
thanks to all who have provided help

Answer 1

Don't rely on lines starting (or not) with any specific characters - just attach the 4th and subsequent lines in each record to the end of the 3rd line of that record:
$ awk '
BEGIN { RS=ORS=""; FS=OFS="\n" }
{
print $1,$2,$3
for (i=4;i<=NF;i++)
printf " %s", $i
print "\n\n"
}
' file
1
00:00:27,160 --> 00:00:29,054
Sometimes there's not much dialogue.
2
00:00:30,100 --> 00:00:31,090
But other times there is quite a bit, and it's formatted into two lines
3
00:00:31,500 --> 00:00:33,700
I want to remove the line breaks only on these long lines, leaving all other formatting.
4
00:00:33,805 --> 00:00:37,285
So that all dialogue ends up being on a single line no matter how long that line.

Answer 2

I think you need something like
awk '/[0-9]+/,/^$/{ if(NR<3) print $0; else {while($0!=""){ printf $0;next; }}}' file
It's not working, but you may get the idea.

Answer 3

You can try something like this with awk:
awk '!NF{print}/[a-z]/{printf "%s ", $0;next}1' file
$ cat file
1
00:00:27,160 --> 00:00:29,054
Sometimes there's not much dialogue.
2
00:00:30,100 --> 00:00:31,090
But other times there is quite a bit,
and it's formatted into two lines
3
00:00:31,500 --> 00:00:33,700
I want to remove the line breaks only on
these long lines, leaving all other formatting.
4
00:00:33,805 --> 00:00:37,285
So that all dialogue ends up being on a single
line no matter how long that line.
$ awk '!NF{print}/[a-z]/{printf "%s ", $0;next}1' file
1
00:00:27,160 --> 00:00:29,054
Sometimes there's not much dialogue.
2
00:00:30,100 --> 00:00:31,090
But other times there is quite a bit, and it's formatted into two lines
3
00:00:31,500 --> 00:00:33,700
I want to remove the line breaks only on these long lines, leaving all other formatting.
4
00:00:33,805 --> 00:00:37,285
So that all dialogue ends up being on a single line no matter how long that line.

Answer 4

Delete all new lines that are preceded by a letter or a space or tab:
perl -pe 's/([a-zA-Z \t])\n$/$1/'

Answer 5

I had the same problem and wrote this little code, which solved my problem:
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
int main(int argc, char *argv[]) {
FILE *quelle,*ziel;
int i;
long maxsub,count,tmp,sub;
char puffer[10][200], *ptr,line[400];
if(argc != 3)
{
printf("Usage: srtlinejoin Filename CountOfSubtitles\n");
return EXIT_FAILURE;
}
maxsub = strtol( argv[2], &ptr, 10);
if( (quelle=fopen(argv[1],"r")) == NULL) {
fprintf(stderr, "Can't open %s\n", argv[1]);
return EXIT_FAILURE;
}
if( (ziel=fopen("out.srt","w")) == NULL) {
fprintf(stderr, "Can't open out.srt\n");
fclose(quelle);
return EXIT_FAILURE;
}
//read and write first line
fgets(puffer[0], 200, quelle);
fputs(puffer[0], ziel);
for(count=1; count < maxsub;count++)
//for(count=1; count <= 3;count++)
{
//printf("Processing subtitle %d\n",count);
tmp=0;
//Read and write time
fgets(puffer[0], 200, quelle);
fputs(puffer[0], ziel);
do {
fgets(puffer[tmp], 200, quelle);
//Scan for next Subtitle
sub = strtol( puffer[tmp], &ptr, 10);
tmp++;
}
while(sub != (count+1));
//Der Untertitel hat nur eine Zeile
if (strlen(puffer[1]) == 2)
{
fputs(puffer[0], ziel); //New Subtitle
fputs(puffer[1], ziel); //Next empty line
fputs(puffer[2], ziel); //Next number
}
//Der Untertitel hat zwei Zeile
if ((strlen(puffer[1]) > 2) && (strlen(puffer[2]) == 2))
{
for(i=0;i<400;i++)
line[i]=0;
strncpy(line,puffer[0],(strlen(puffer[0])-2));
strcat(line," ");
strcat(line,puffer[1]);
fputs(line, ziel); //New Subtitle
fputs(puffer[2], ziel); //Next empty line
fputs(puffer[3], ziel); //Next number
}
//Der Untertitel hat mehr als zwei Zeile
if ((strlen(puffer[1]) == 2) && (strlen(puffer[2]) == 2))
{
printf("Attention: The subtitles has more than two lines\n");
}
}
printf("Check last subtitle!\n");
fclose(quelle);
fclose(ziel);
return EXIT_SUCCESS;
}