Awk substring a single character

Question

Here is columns.txt

aaa bbb 3
ccc ddd 2
eee fff 1
3   3   g
3   hhh i
jjj 3   kkk
ll  3   mm
nn  oo  3

I can find the line where second column starts with "b":

awk '{if(substr($2,1,1)=="b") {print $0}}' columns.txt

I can find the line where second column starts with "bb":

awk '{if(substr($2,1,2)=="bb") {print $0}}' columns.txt

Why oh why can't I find the line where the second character in the second column is "b"?:

awk '{if(substr($2,2,2)=="b") {print $0}}' columns.txt

awk -W version == GNU Awk 3.1.8

because you're telling `substr()` to return 2 characters (the last arg) and then comparing the resulting 2-char string (e.g. `bb`) to the 1-char string `b` to see if they are identical, which of course they never can be. — Ed Morton, Sep 16 '14 at 17:06
I have been using this function as substr(field,start,end) where end is the Nth character of string. It had worked so far because I always searched from the first character. — AWE, Sep 18 '14 at 06:18

anubhava · Accepted Answer · 2014-09-16T17:10:47.660

5

You can use:

awk 'substr($2,2,1) == "b"' columns.txt
aaa bbb 3

edited Sep 16 '14 at 17:10

answered Sep 16 '14 at 12:30

anubhava

1

Yes I meant 3rd argument only. Thanks @EdMorton. – anubhava Sep 16 '14 at 17:10

score 3 · Answer 2 · answered Sep 16 '14 at 12:33

3

awk '$2 ~ /^.b/' columns.txt will also work.

answered Sep 16 '14 at 12:33

Etan Reisner

2 Answers2