CODEWITHSUNDEEP
stringbashawkprintf
lugte098
lugte098

Reputation: 2309

How to use a variable in an awk expression

I'm trying to make this command:

sed bla bla filename | awk '{printf "%s %s_entry_%.3f %.3f %.3f %.3f",$1,$3,$4,$5,$6,$7}'

But the thing is, i want the %.3f part to be variable. So in one case it could be %.3f and in another it could be %.3f %.3f %.3f. So i'll just use a static one in my example code for clarity. So if i want 4 of these %.3f and put them in variable $values like so:

values="%.3f %.3f %.3f %.3f"

Then how can I put this string in the awk expression, without making awk to just put literally "${values}" in there. The following is my non-working-attempt:

sed bla bla filename | awk '{printf "%s %s_entry_${values}",$1,$3,$4,$5,$6,$7}'

Upvotes: 3

Views: 16779

Answers (3)

ghostdog74
ghostdog74

Reputation: 342313

you can use -v option of awk to pass in variable from the shell

#!/bin/bash    
awk -v values="${values}" '{gsub("blah","replace");printf "%s %s_entry_"values ....}' file

the gsub() function is to replace what you have with that sed command. So just one awk command will do. sed is redundant in this case (and most other cases where awk is used)

Upvotes: 7

Marcelo Cantos
Marcelo Cantos

Reputation: 185842

If you mean that values is a shell variable, then this will work:

sed bla bla filename | awk '{printf "%s %s_entry_"ENVIRON["values"],$1,$3,$4,$5,$6,$7}'

Upvotes: 1

Ignacio Vazquez-Abrams
Ignacio Vazquez-Abrams

Reputation: 798536

Easiest to use actual awk variables:

sed bla bla filename | awk -v values="$values" '{printf "%s %s_entry_"values,$1,$3,$4,$5,$6,$7}'

Or with bash:

awk -v values="$values" '{printf "%s %s_entry_"values,$1,$3,$4,$5,$6,$7}' <(sed bla bla filename)

Upvotes: 2

Related Questions