Printing $ Lines After Finding Specific Pattern

pravin.bhande · 04-11-2012, 01:48 AM

Hi,

Following code block will help you to print lines/values after locating specific pattern in the file

i :- Key/search pattern
X :- any number e.g. 1 or 2
Y :- any number

1. Printing Xth line after finding specific pattern

sed -n '/'${i}'/,$p' <File> | sed -n 'Xp'

2. Printing X lines after finding specific pattern

sed -n '/'${i}'/,$p' <File> | sed -n 'X,Yp'

here X and Y are start to end number

Regards,
Pravin B

grail · 04-11-2012, 04:28 AM

So is there a question here or you are just showing your skills?

danielbmartin · 04-11-2012, 09:41 AM

Quote:

Originally Posted by pravin.bhande

Following code block will help you to print lines/values after locating specific pattern in the file

i :- Key/search pattern
X :- any number e.g. 1 or 2
Y :- any number

1. Printing Xth file after finding specific pattern

sed -n '/'${i}'/,$p' <File> | sed -n 'Xp'

2. Printing X files after finding specific pattern

sed -n '/'${i}'/,$p' <File> | sed -n 'X,Yp'

here X and Y are start to end number

Thank you for this contribution. A gentle word of constructive criticism regarding terminology.

Code:

1. Printing Xth file after finding specific pattern
 .. should be ...
1. Printing Xth line after finding specific pattern

... and ...

Code:

2. Printing X files after finding specific pattern
 .. should be ...
2. Printing lines X through Y after finding specific pattern

Daniel B. Martin

David the H. · 04-12-2012, 08:03 AM

Good work. Figuring out how to do such things on your own is always a good thing to do.

That said, however, with gnu sed (the Linux standard) and a few others, it's possible to do the same things in a single command. It includes additional address range forms for matching numbers of lines after the initial match.

To print line $i, plus $n lines after it:

Code:

i=pattern
n=4

sed -n "/$i/,+$n p" file.txt

#or just use grep instead:
grep -A "$n" "$i" file.txt

(BTW, as you can see, to use a variable in the expression you can usually just enclose the whole thing in double quotes.)

To print only the $n'th line after the pattern, use a nested expression to delete all the lines from the first match that you don't want:

Code:

i=pattern
n=4
m=$(( n - 1 ))

sed -n "/$i/,+$n { /$i/,+$m d; p }" file.txt

danielbmartin · 04-12-2012, 08:49 AM

Quote:

Originally Posted by David the H.

To print line $i, plus $n lines after it ...

OP showed how to not print line $i, and print only $n lines after it. Please post a version of your excellent code which does this.

Daniel B. Martin

David the H. · 04-12-2012, 10:00 AM

Easy. Just use the same technique as in my second example, but delete only the line with the pattern. Or invert the match and print everything except it.

Code:

sed -n "/$i/,+$n { /$i/d; p }" file.txt
sed -n "/$i/,+$n { /$i/!p }" file.txt

danielbmartin · 04-12-2012, 10:20 AM

Quote:

Originally Posted by David the H.

Code:

sed -n "/$i/,+$n { /$i/d; p }" file.txt
sed -n "/$i/,+$n { /$i/!p }" file.txt

Works like a charm. Thank you.

Daniel B. Martin