assign values from awk output - help

Shell Programming and Scripting
1

Dear All,

I have a command which gives the number of fields of each line of a comma-separated file.

sthng like this :
cat QDB_20071126_002.bad | awk -F"," '{ print NF }'

I need to assign the first output and the last output of the above command to variables in a script.
Need help to do that.

Thanks in advance.

2 
set -- `awk -F, 'NR==1 { print $NF; next } { last=$NF } END { print last }' QDB_20071126_002.bad`
firstvariable=$1
lastvariable=$2

You could do it with two separate awk invocations or something, but the above is more efficient (although somewhat obscure, the set -- `command` trick is a standard shell programming idiom).

3

Hi,

Thanks for your quick reply.
Your command actually prints the last field of first and last line, whereas what i need is the number of fields of first and last line, to be assigned to variables.

Lines are comma-separated.

The below command prints number of fields of each line of the comma-separated file:
cat QDB_20071126_002.bad | awk -F"," '{ print NF }'

6
7
7
7
2

I need to take '6' and '2' and assign it later.
Help me out on this..

Thanks in advance.

4

Sorry, my bad, just remove the dollar signs before NF in both places; sorry for missing that.

By the by, awk can read the file all by itself; no need to invoke a separate cat for that.

5

you can assign variables back to the shell using set as era has mentioned

set -- `awk 'NR==1{print NF}END{print NF}' file`

or just do whatever you need to do in awk itself

awk 'NR==1{first=NF}END{ 
   print NF #do something.
   print f # do something
}' file