I want to create an awk script that, passed a URL, trims off the final section (everything after the last '/'

I've created the following in a file trim.awk:

---------------------------

BEGIN { FS="/"
new_var = "" }

{ for (count_var=1; count_var<NF; count_var++) {
new_var = $new_var$count_var
print count_var " - " $count_var " - " $new_var }
}


---------------------------

I'm calling it from the command line with:

echo "www.helloworld.com/this/is/a/test" | awk -f trim.awk

The results I get are far from what I expected:

1 - www.helloworld.com - www.helloworld.com/this/is/a/test
2 - this - www.helloworld.com/this/is/a/test
3 - is - www.helloworld.com/this/is/a/test
4 - a - www.helloworld.com/this/is/a/test

In other words, my $new_var is returning the whole input string every time, regardless of what it's assigned to...

Any ideas? I presume I'm doing something basic wrong, but I've been looking at this for a whole day...

Dave

After a flash of inspiration and some more googling, I came up with the following solution. The problem was with using $ with variable names...

---------------------------------

BEGIN { FS="/" }

{ for (count_var=1; count_var<NF; count_var++) {
new_var = new_var$count_var"/"
}
print new_var
}

---------------------------------

I know the following command was never designed for this but did you consider using:

dirname http:www.linuxquestions.org/billgates.html

However, if you are doing what you are doing as an excercise in awk then fair enough