read specified range of lines from a file

Question

I have i file that contains 100000 line how i can get the lines from line# 5555 to line# 7777 under linux.

Thanks for all.

@ibrahim, consider accepting Kyle's answer (by ticking the green checkmark on the left) if it helped you — Jonik, Jan 12 '10 at 14:17
you may precise if you want to use sed or if other tools are OK. — Manu H, Feb 06 '15 at 16:17
A little bit of research throws at least http://stackoverflow.com/questions/83329/how-can-i-extract-a-range-of-lines-from-a-text-file-on-unix — sancho.s ReinstateMonicaCellio, Dec 13 '15 at 12:40

Kyle Brandt · Answer 1 · 2010-01-12T15:28:31.027

sed '5555,7777!d' <filename>

This will print lines 5555-7777 of the file inclusively.

Dennis Posted the following which I agree should be faster:

sed '5555,7777p; 7778q' filename

The following evidence that it should be faster:

$ n=1
$ while [[ n -le 100000 ]]; do echo $n >> sedtest2; n=$((n + 1)); done
$ strace -e trace=read -o sed1 sed '5555,7777!d' sedtest2
$ strace -e trace=read -o sed2 sed '5555,7777p; 7778q' sedtest2
$ wc -l sed1
149 sed1
$ wc -l sed2
14 sed1

In Bash only (for fun):

n=1
while read line; do 
    if [[ ($n -ge 5555) && ($n -le 7777)  ]]; then 
        echo $line
    elif [[ $n -gt 7777 ]]; then
        break
    fi 
    n=$(( $n + 1 ))
done < file

I think your `$n -gt 3` should be 7777 perhaps? Also, you can do `if (( n >= 5555 ))` for more "natural" looking numeric comparison operators (and the ability to leave off the dollar sign). And you can do `((n++))`. — Dennis Williamson, Jan 12 '10 at 15:18

score 16 · Answer 2 · edited Feb 06 '15 at 16:12

16

Quitting when you're done can speed things up:

sed -n '5555,7777p; 7778q' input_file

edited Feb 06 '15 at 16:12

nwk

107
4

answered Jan 12 '10 at 14:00

Dennis Williamson

60,515
14
113
148

1

+1 , should be faster, updated my post to show why. – Kyle Brandt Jan 12 '10 at 14:19

score 7 · Answer 3 · answered Jan 12 '10 at 13:43

7

Either of these should work;

sed -n 'startnumber,endnumberp'
awk 'NR>=startnumber&&NR<=endnumber' file

Great question by the way ;)

answered Jan 12 '10 at 13:43

Chopper3

100,240
9
106
238

score 1 · Answer 4 · answered Nov 03 '14 at 17:32

I found that the sed option did not work on a mysqldump file, I'm guessing due to handling of quoted line feeds or multibyte characters. head and tail slice it using the same line numbers as grep which was what I needed. To get lines $j through $k, you need:

x=$(( $k - $j + 1 ))
tail -n +$j filename | head -${x}

read specified range of lines from a file

4 Answers4

Linked