Print from Nth match to Mth match

Question

How do I print from the nth match to the mth match in a file?

For example

>1
dfgsdrhsdtrh
>2
zdfrgsdh
>3
zdsfgadrh
>4
sdtrh
>5
drh

If I want the 2nd to the 4th match to the regexp '^>', including the Nth match and up to but not including the M+1th match. Yielding, for example:

>2
zdfrgsdh
>3
zdsfgadrh
>4
sdtrh

See http://stackoverflow.com/a/17914105/258523 for the answer to this and many other similar requests. — Etan Reisner, Nov 10 '14 at 18:59

anubhava · Accepted Answer · 2014-11-10T19:58:41.833

1

Using gnu-awk you can use this awk:

awk -v RS='>' -v ORS='>' 'BEGIN{printf ">"} NR>5{exit} NR>2 && NR<=5' file
>2
zdfrgsdh
>3
zdsfgadrh
>4
sdtrh

edited Nov 10 '14 at 19:58

answered Nov 10 '14 at 19:48

anubhava

761,203
64
569
643

1

This works! Thanks. Is there a way to get it to quit after the last match instead of continuing to search through the file? – stuppie Nov 10 '14 at 19:57
Yes sure that can be done. See edited answer for exiting once record is passed. – anubhava Nov 10 '14 at 19:59
It still prints the last field separator '>' at the end. I can just pipe it through 'head -n-1' though... – stuppie Nov 10 '14 at 20:07
Actually with your input data it doesn't print `>` at the end for me. – anubhava Nov 10 '14 at 20:12

score 0 · Answer 2 · answered Nov 10 '14 at 19:38

You could use the \K construct if supported. If not, just remove the \K and put capture
parenthesis around what comes after it.

 #  (?m)(?:^>.*\r?\n(?:^(?!>).*\r?\n)*){1}\K(?:^>.*\r?\n(?:^(?!>).*\r?\n)*){3}

 (?m)                    # MULTI_LINE mode

 (?:                     # 0 to N-1
      ^ > .* \r? \n 
      (?:
           ^ 
           (?! > )
           .* \r? \n 
      )*
 ){1}                    # Do N-1 times

 \K                      # Disgard previous from match

 (?:                     # N to M+1
      ^ > .* \r? \n 
      (?:
           ^ 
           (?! > )
           .* \r? \n 
      )*
 ){3}                    # Do M+1-N times

Output:

 **  Grp 0 -  ( pos 18 , len 40 ) 
>2
zdfrgsdh
>3
zdsfgadrh
>4
sdtrh

score 0 · Answer 3 · answered Nov 10 '14 at 19:45

0

Here is an awk solution:

awk -v RS=">" '$1>=2 && $1<=4 {$0=RS$0;gsub(/\n$/,"");print}' file
>2
zdfrgsdh
>3
zdsfgadrh
>4
sdtrh

It prints record from >2 to >4

answered Nov 10 '14 at 19:45

Jotne

40,548
12
51
55

Print from Nth match to Mth match

3 Answers3