awk different behaviour in command line and in shell script

Question

I have this script which works fine

#!/bin/bash
if [ $# -lt 1 ]
then
  echo "USAGE: `basename $0` username"
  exit 1
fi

logins=`last | grep $1`
times=`echo "$logins" | awk '{print $10}'`
timesCleared=`echo "$times" | sed -e 's/(//' -e 's/)//'`
minutes=`echo "$timesCleared" | awk -F: '{ print $1*60+$2}'`
total=0
for m in $minutes
do
  total=$(( $total + $m ))
done

echo $total > out.txt
cat out.txt

but, for an example, when i try this command in the interactive shell it doesn't work ( echo $minutes gives a different output):

echo "in 02:16 00:28 00:16 00:25 02:44 00:09" | awk -F: '{ print $1*60+$2;}'

it prints 16

if i write echo $minutes in the scripts it gives me the right output:

0 136 28 16 25 164 9

Does anyone know why is this?

Looks like there oughta be some escaping.... If it's not with the dollar signs, then it's with the math operators, + - * /. — yftse
– yftse, Commented Jan 17, 2016 at 17:07
in does not make sense as a number, so is evaluated as zero. (Yes, there is a numeric suffix, but this is what's going on here.) — tripleee
– tripleee, Commented Jan 17, 2016 at 17:19
Also, some of the lines I'm getting contain strings like 3+23:27. — choroba
– choroba, Commented Jan 17, 2016 at 17:21

rici · Accepted Answer · 2016-01-17 17:35:14Z

2

The variables in your script contain newlines, which are important. What is being sent to awk in the line which assigns to the variable minutes is not

in 02:16 00:28 00:16 00:25 02:44 00:09

as in your command-line, but rather

in
02:16
00:28
00:16
00:25
02:44
00:09

In the first case, awk sees $1 as in 02 and $2 as 16 00. In the second case, it works on each line individually. When awk converts a string to a number, it uses only the first characters of the string; if it does not find a number at the beginning (as with the string in 02), it converts it to 0 without any error message.

You can use the RS awk pseudo-variable to set the record separator to a space:

echo "in 02:16 00:28 00:16 00:25 02:44 00:09" |
    awk -F: -v RS=' ' '{ print $1*60+$2;}'

Alternatively, you could echo the string with newlines:

echo "in
02:16
00:28
00:16
00:25
02:44
00:09" | awk -F: '{ print $1*60+$2;}'

edited Jan 17, 2016 at 17:35

answered Jan 17, 2016 at 17:19

rici

243k30 gold badges263 silver badges364 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

tripleee · Accepted Answer · 2016-01-17 17:25:33Z

2

The string you echo in the script is newline-separated, not space-separated; so the command you tried interactively is not doing the same thing.

To reproduce what's going on in the script, try

echo "in 02:16 00:28 00:16 00:25 02:44 00:09" |
tr ' ' '\n' |
awk -F: '{ print $1*60+$2;}'

answered Jan 17, 2016 at 17:25

tripleee

192k37 gold badges318 silver badges369 bronze badges

1 Comment

Abstrac Tor Over a year ago

Yes, i suspected it might be the formatting. This works :)

Collectives™ on Stack Overflow

awk different behaviour in command line and in shell script

2 Answers 2

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related