How to use multiple, different characters as field separator in busybox awk?

Question

I'd like to split a string with BusyBox's awk using multiple and different characters. For example, spaces and slashes.

I read the answer of Can field separator in awk encompass multiple characters?. It states that you should be able to pass "even a regex" as separator. But I'm unable to find the right syntax:

It is this output line of ip which I'm working on:

inet 192.168.2.1/24 brd 192.168.2.255 scope global eth0

I'd like to exact the IPv4 address:

ip -4 addr show dev eth0 | grep inet | awk -F'[ \/]' '{print $2;}'

But it doesn't work.

Possible duplicate of [AWK multiple delimiter](http://stackoverflow.com/questions/12204192/awk-multiple-delimiter) — tripleee, Mar 14 '17 at 06:36
Works just fine over here. Are you sure you're using GNU awk? Which version? — James Brown, Mar 14 '17 at 06:42
The answer of the duplicate question, does not work for me. I need the repetition operator. — Patrick B., Mar 14 '17 at 06:42
@JamesBrown Damn it, I didn't notice BusyBox v1.22.1 (2017-02-24 07:41:34 UTC) multi-call binary. — Patrick B., Mar 14 '17 at 06:42
@PatrickB. If any of the following answer helped you in your goal please accept/upvote the answer for closure by ticking right sign besides the answer. — P...., Mar 27 '17 at 10:59

score 2 · Answer 1 · answered Mar 14 '17 at 06:34

Following code will check for the line containing a string "inet". If this matched, then the second column in split into an array named "a". and the first element of that array is printed.

ip -4 addr show dev eth0  |awk '/inet/{split($2,a,"/");print a[1]}'

Or you can try grep :

ip -4 addr show dev eth0 |grep -oP 'inet\s\K.*?[^/]+'

Patrick B. · Accepted Answer · 2017-03-14T06:48:59.617

0

When using a regex, one needs to specify how many characters to match for. My regex was missing that:

ip -4 addr show dev eth0 | awk -F"[ /]+" '/inet/{print $3}'

is working. (Note, I'm running the busybox-version of awk)

The +-operator stands for at least one of the character list needs to match.

I also could remove the grep, thanks @PS.

edited Mar 14 '17 at 06:48

answered Mar 14 '17 at 06:40

Patrick B.

11,773
8
58
101

Well, I just had to test it but on my Busybox at hand (Ubuntu 1:1.21.0-1ubuntu1) and that `+` was not needed. You (awk) code worked as-is. – James Brown Mar 14 '17 at 08:42

score 0 · Answer 3 · answered Mar 14 '17 at 06:45

0

Use split:

$ awk '{split($2,a,"/");print a[1]}' foo
192.168.2.1

answered Mar 14 '17 at 06:45

James Brown

36,089
7
43
59

How to use multiple, different characters as field separator in busybox awk?

3 Answers3