64

How can I add numbers to the beginning of every line in a file?

E.g.:

This is
the text
from the file.

Becomes:

000000001 This is
000000002 the text
000000003 from the file.
zx8754
  • 52,746
  • 12
  • 114
  • 209
Village
  • 22,513
  • 46
  • 122
  • 163

7 Answers7

124

Don't use cat or any other tool which is not designed to do that. Use the program:

nl - number lines of files

Example:

$ nl --number-format=rz --number-width=9 foobar
$ nl -n rz -w 9 foobar # short-hand

Because nl is made for it ;-)

CervEd
  • 3,306
  • 28
  • 25
tamasgal
  • 24,826
  • 18
  • 96
  • 135
39

AWK's printf, NR and $0 make it easy to have precise and flexible control over the formatting:

~ $ awk '{printf("%010d %s\n", NR, $0)}' example.txt
0000000001 This is
0000000002 the text
0000000003 from the file.
Raymond Hettinger
  • 216,523
  • 63
  • 388
  • 485
  • Wonderful! I needed this to populate a huge array of static int declarations, so I had something like `public static final int CONSTANTX;` and I wanted something like `public static final int CONSTANTX;=X` where `X` is the number of the line in the source file. I rearranged your answer as `awk '{printf("%s = %d;\n", $0, NR-4)}' Constants.java` and it worked like a charm! – afe Jan 16 '19 at 13:20
32

You're looking for the nl(1) command:

$ nl -nrz -w9  /etc/passwd
000000001   root:x:0:0:root:/root:/bin/bash
000000002   daemon:x:1:1:daemon:/usr/sbin:/bin/sh
000000003   bin:x:2:2:bin:/bin:/bin/sh
...

-w9 asks for numbers nine digits long; -nrz asks for the numbers to be formatted right-justified with zero padding.

sarnold
  • 102,305
  • 22
  • 181
  • 238
  • That's the exact answer and a good one, but for a homework style question it may have been better to offer a hint, a link, or a more general purpose solution. – Raymond Hettinger Nov 21 '11 at 01:45
  • 3
    @Raymond, true enough, but I've not seen any indication that this is homework. Simple? Sure. But there's a giant pile of text-processing utilities on these things that even experienced developers never learn about. – sarnold Nov 21 '11 at 01:47
  • @Raymond: I suggest undeleting your answer, I like it. Different tools might excel for different reasons, and `awk(1)` certainly excels at tasks like this. – sarnold Nov 21 '11 at 01:48
  • I only knew because I fell for the a previous question a few minutes ago: http://stackoverflow.com/questions/8206280/delete-all-lines-beginning-with-a-from-a-file Also, the unusual number of leading zeros in the example was another hint :-) – Raymond Hettinger Nov 21 '11 at 01:49
  • 1
    This is not for homework. I just thought, it could take me 2 hours to figure out how to do that in `awk`, using some for loop, which would take another 30 minutes for the script to run through the file; or I could ask here. I've never heard of `nl` before. Am I posting the wrong kind of questions? – Village Nov 21 '11 at 02:08
  • 4
    Village: not at all -- Stack Overflow is intended to answer questions of all levels. @Raymond just noticed that you've got three questions asked in a short time frame asking for solutions to problems similar to homework a professor might assign in a "Learning Unix" course. One alone sounds like someone looking for a solution to a problem, but all three taken together makes us wonder if we're doing your homework for you. :) – sarnold Nov 21 '11 at 02:18
17

cat -n thefile will do the job, albeit with the numbers in a slightly different format.

8

Easiest, simplest option is

awk '{print NR,$0}' file

See comment above on why nl isn't really the best option.

egorulz
  • 1,455
  • 2
  • 17
  • 28
6

Here's a bash script that will do this also:

#!/bin/bash
counter=0
filename=$1
while read -r line
do
  printf "%010d %s" $counter $line
  let counter=$counter+1
done < "$filename"
slashdottir
  • 7,835
  • 7
  • 55
  • 71
4
perl -pe 'printf "%09u ", $.' -- example.txt