Extract lines between two patterns when there are multiple matches in the file

Question

I want to extract lines between two patterns (say, pattern1 and pattern2). The structure of the file is the following:

a random number of lines containing other stuff
pattern1
some lines to be extracted (finding out the number of lines possible, if necessary)
pattern2
a random number of lines containing other stuff
pattern1
some lines to be extracted
pattern2
a random number of lines containing other stuff

This repeats for a large number of times (i.e. there is a large number of matching pattern1-pattern2 pairs). I want to extract the lines between the patterns for all matches, effectively discarding the random stuff.

How can I do this?

This answer might also be applicable: https://stackoverflow.com/a/48022994/2026975 — imriss, Dec 29 '17 at 17:14

score 2 · Answer 1 · answered Nov 08 '13 at 12:26

Using awk

awk '/pattern1/,/pattern2/'
pattern1
some lines to be extracted (finding out the number of lines possible, if necessary)
pattern2
pattern1
some lines to be extracted
pattern2

Only lines between pattern

awk '/pattern2/ {f=0;next} f; /pattern1/ {f=1}'
some lines to be extracted (finding out the number of lines possible, if necessary)
some lines to be extracted

score 2 · Answer 2 · answered Nov 11 '13 at 09:30

2

sed -n "/pattern1/,/pattern2/ {
 /pattern1/ !{
   /pattern2/ !p
   }
 }" InputFile

printing line BETWEEN pattern, excluding the patterns itself

answered Nov 11 '13 at 09:30

NeronLeVelu

9,908
1
23
43

score 1 · Answer 3 · answered Nov 08 '13 at 12:23

1

You can use sed for this:

cat inputfile | sed -ne '/pattern1/,/pattern2/p'

answered Nov 08 '13 at 12:23

Mark Veenstra

4,691
6
35
66

[UUOC](http://partmaps.org/era/unix/award.html) – devnull Nov 08 '13 at 12:24
1

just to show you can pipe the command.. but thnx for the award – Mark Veenstra Nov 08 '13 at 12:25

score 1 · Answer 4 · answered Nov 08 '13 at 12:31

1

Here are two more in awk:

/pattern1/ {inside_block=1}
/pattern2/ {inside_block=0}
inside_block==1 {print $0}

OR

/pattern1 { print $0; while(getline > 0) {print $0;if (/pattern2/) break }}

Neither is as elegant as the posted solution, but both might be useful depending on other requirements of the program, or the complexity of the pattern.

answered Nov 08 '13 at 12:31

DavidG

399
2
10

First example prints start pattern `pattern1` line and text to extract, but not stop pattern `pattern2`. Either none, or both. Second does not work, due to missing "\" after `pattern1`. – Jotne Nov 08 '13 at 12:34

Extract lines between two patterns when there are multiple matches in the file

4 Answers4