Last gasp for me. :)

I figured out what the problem is.

SED looks for the start string more then once. Is there a way to limit SED to start looking and stop look just once?

How about beginning with the first line?

I have some text before " HETATM...............A " that i dont want to select.

The problem is after TER, there is another line with " HETATM...............A ", I dont want anything before " HETATM...............A " and after "TER". SED reads the lines after TER too. How do i avoid that?

This is my command right now

sed -n "/^ATOM\|HETA.................A/, /^TER/p"

It should read if the line starts with ATOM or HETA followed by 16 characters and then A.

The 16 characters can be anything...

Threads merged because they are essentially the same.

is there a way to read just the first thread?

There is a complication to that. I am calling the command from PHP.

The file is in the following format

ATOM[or HETA] [14 random charaters] A
100 lines like this
TER

ATOM[or HETA] [14 random charaters] B
100 lines like this
TER

ATOM[or HETA] [14 random charaters] C
100 lines like this
TER

HETA [14 random charaters] A
100 lines like this
TER
HETA [14 random charaters] B
100 lines like this
TER

So i need to get A, B and C[Starting with ATOM or HETA] into seperate files. but i dont want the lines if the whole string starts only with HETA.

So i want
ATOM[or HETA] [14 random charaters] A
100 lines like this
TER

into A.txt

and
ATOM[or HETA] [14 random charaters] B
100 lines like this
TER

into B.txt

so the command should not have '0,/^TER/ !d' as it will count only A and not B or C.