I want to take a shell variable and use it in an awk command

Example:

>var="Hello People"
>awk -F":" '$2 == var {print $0}' filename

Jean thanks in advance, ha, or thanks to who ever else answers these forums.

Two solutions :

Play with shell quotes :
Executes : awk -F":" '$2 == "Hello People" {print $0}' filename

The other solution (my prefered) is to use awk variables :

Hi Micster,

Just so you don't think JP is out on a limb here, I'll chip in my 2 cents...

There are 2 ways to pass a variable to awk using awk syntax (there are many others using the shell but we'll ignore that).

The original method implemented in awk was to declare a variable after the awk command.

e.g. awk '{print var}' var="hello world"

The second method was introduced a little later in the life of awk and was implemented in order that a passed variable could be used in a BEGIN block. This uses the -v flag as illustrated by JP.

Compare the following 2 examples:

awk -v var="hello world" 'BEGIN {print var}'

awk 'BEGIN {print var}' var="hello world"

The second example will not output anything because the variable would only be available in the main body (and the END block).

I guess this is why the -v method is preferred by most (including myself).

HTH

Man you guys are great! It might seem like small stuff to you guys but I beat my head on this for 6 hours lastnight the whole time cursing mentally at my instructor.

As always thanks so much!