Print lines from a file based on the index in other file

Question

I have a file (myFile) with 20000 lines and another file (myIndex) like this containing the lines that must be extracted from myFile:

I want to extract these lines with sed:

sed '10!d' myFile

How can I put the file name instead of 4,10 like this:

sed 'myIndex!d' myFile

Or is there a similar function in awk?

line 1200, I should extract line 4, 10, 1200 and ... from myFile. — user1436187, Nov 07 '15 at 12:14
`sed 4,10!d` returns lines 4-10, it **doesn't** return lines 4 *and* 10 — Sjon, Nov 07 '15 at 12:17
Possible duplicate of [Using awk to pull specific lines from a file](http://stackoverflow.com/questions/12182910/using-awk-to-pull-specific-lines-from-a-file) — Thor, Nov 07 '15 at 14:02

hek2mgl · Answer 1 · 2015-11-07T12:30:15.163

2

You can use awk:

awk 'NR==FNR{i[$0];next}i[FNR]' index file

NR==FNR is true as long as we are reading the first while index. We are storing all that line numbers in an array i. On the remaining input we are checking if the line number is stored in i. If that is true, awk prints that line.

edited Nov 07 '15 at 12:30

answered Nov 07 '15 at 12:19

hek2mgl

152,036
28
249
266

Cyrus · Accepted Answer · 2015-11-07T12:51:04.090

1

Try this with GNU sed:

sed -nf <(sed 's/.*/&p/' myIndex) myFile

-n: suppress automatic printing of pattern space

-f: add the contents of script-file to the commands to be executed

<(...): bash's Process Substitution

edited Nov 07 '15 at 12:51

answered Nov 07 '15 at 12:17

Cyrus

84,225
14
89
153

Print lines from a file based on the index in other file

2 Answers2