Shell Programming and Scripting

I'm an experienced awk user, but this one has me stumped. I have an awk script which is called from a UNIX command line as you'd expect:

Code:

myscript.awk -v foo=$1 -v bar=$2 filename

My question is this: is there a mechanism for determining the names of the -v variables within a script?

ARGV doesn't include them (and quite rightly so, since they're not arguments).

I'd like my script to be able to parse a list of variables passed, so that I can print an error message along the lines of "Unknown variable name - please check and try again", or call a Usage() function, or some such.

Thanks in advance, guys 'n' gals.

Why not put this whole idea inside a ksh script and let ksh handle the error checking?

ie:

Code:

#!/bin/ksh

error_ind=0

for arg in $* ; do

  case $arg in

    frogs=*) : ;;
    dogs=*) : ;;
    *=*) print unknown variable: $arg ; error_ind=1 ;;

  esac

done

if [[ $error_ind -eq 1 ]]; then
  exit 1
fi

awk etc..... here.....

With Bell-lab's awk, you have:

Code:

bell-awk -v v11=aa -v v22=bb 'BEGIN{for (i in SYMTAB) print i}'
...
ENVIRON
v22
v11
FILENAME
...

SYMTAB used to be in gawk. There is

Code:

-W dump-variables

command line option for gawk, but not very useful in your case.

Cheers, guys.

quirkasarus: Great minds, and all that: that's exactly what I ended up doing (and then scrapped it - see below).

binlib: You've cracked it! SYMTAB holds the variable names (amongst much else). Just what I was looking for:-)

Many thanks to both of you for looking into this.