Re: How to pass UNIX variable into awk?

On 4/8/2008 9:57 AM, RC wrote:
> RC wrote:
> 
>>myvariable="hello"
>>
>>cat my file, if first field match "hello", then print whole line
>>
>>I tried
>>
>>cat myfile.txt | awk -F: '$1 ~ /$myvariable/ {print $0}'
>>
>>but doesn't work, I also tried double $ like $1 ~ /$$myvariable/
>>still no luck.
> 
> 
> I found the answer from
> http://www.faqs.org/faqs/unix-faq/faq/part3/section-12.html
> 
> cat myfile.txt | awk -F: '$1 ~ /'"$myvariable"'/ {print $0}'
> 
> Where '" are single quote and double quotes
>        "' are double quotes and single quote

No, never do that. Among other reasons, you just can't do it if your
script
grows moderately large and you want to save it in a file to execute using
awk
-f. Just set an awk variable based on the value of the shell variable and
use
that awk variable in your awk script:

    awk -F: -v myvar="$myvariable" '$1 ~ myvar' myfile.txt

See question 24 in the comp.unix.shell FAQ
(http://cfaj.freeshell.org/shell/cus-faq-2.html#24)
for more details.

	Ed.