awk - How to print the number of characters for the first n lines in a file?

Question

I have a command:

$ awk '{ print length($0); }' /etc/passwd

It prints number of characters of every line in a passwd file:

How can I print the number of characters for only the first n lines?

For example - for the first 3 lines it would give something like:

52
52
61

If you want to ignore blank like then use below - `awk '(NR<=line) && length($0) {print length($0)}' line=3 file` — VIPIN KUMAR, Mar 28 '17 at 19:26

Thor · Accepted Answer · 2017-03-28T19:00:42.343

17

Tell awk to quit when enough lines have been read:

awk '$0 = length; NR==3 { exit }' /etc/passwd

Note that this solution ignores empty lines, although not for the line count.

edited Mar 28 '17 at 19:00

answered Mar 28 '17 at 08:57

Thor

tricky and cool! (+1) – Mar 28 '17 at 10:32
2

If any or the lines is empty, the assignment evaluates to zero (a falsy value), and the length will not be printed. – ilkkachu Mar 28 '17 at 17:03
@ilkkachu: Depending on the situation, it would be reasonable to ignore empty lines. I have added a note about this. – Thor Mar 28 '17 at 19:02

score 14 · Answer 2 · answered Mar 28 '17 at 10:30

14

A direct Awk version (not so efficient as @Thor's), but slightly more clear:

awk 'NR <= 3 {print length}' /etc/passwd

answered Mar 28 '17 at 10:30

3

If this was `awk '{ print length } NR>=3 { exit }' /etc/passwd`, I'd upvote it. – Dennis Williamson Mar 28 '17 at 18:08
3

@DennisWilliamson, thank you. That is a very good suggestion, I will not include it in my answer because it is already implicit in Thor's proposal. – Mar 28 '17 at 18:28

score 5 · Answer 3 · edited Mar 29 '17 at 06:33

5

You can execute it with awk only command, as nicely described by @Thor, and @JJoao (+1 from me)

You can combine awk and head with parameter -n follows by the number of lines as described below:

Thanks for @Maerlyn suggestion to execute in this order: head | awk

e.g. You will get the first 3 lines using:

head -n3 /etc/passwd | awk '{ print length($0); }'

-n, --lines=[-]K
    print the first K lines instead of the first 10; with the leading '-', print all but the last K lines of each file

edited Mar 29 '17 at 06:33

muru

answered Mar 28 '17 at 08:51

Yaron

@msjavx86 - great :) please [accept](http://askubuntu.com/help/accepted-answer) the answer, when you'll be able to do it – Yaron Mar 28 '17 at 08:53
of course I will - in about 10 minutes ;) thanks again! – bambosze_babuni Mar 28 '17 at 08:54
8

Maybe even reverse, `head -n3 /etc/passwd | awk` so awk doesn't have to process the whole file. – Maerlyn Mar 28 '17 at 13:02
1

AWK can count lines, no need for `head`. – Dennis Williamson Mar 28 '17 at 18:09
@DennisWilliamson - thanks for the comment, using `head` was the first working solution which came to my mind. I've upvoted @Thor, @JJoao answer which uses `awk` only solution. – Yaron Mar 29 '17 at 06:29

3 Answers3