Nawk : Matching dollar sign 1

[IMAUser]

i='$MY_TEST/my_file_name.dat '
echo $i
FILE_DIR=`echo $i | nawk '{print toupper(substr ( $1,match($1,"$&quot+1, match($1,"/" )))}'`
echo "FILE DIR $FILE_DIR "


Thats the script and what I want to do is match the occurance of the dollar sign. But it dosent match and there is no error message as well.
So I want the programme to return the string $MY_TEST

I tried to escape the dollar sign with a backslash (\) but didnt work.
Any help appreciated.
Thanks

 

[vgersh99]

echo '$MY_TEST/my_file_name.dat' | nawk 'match($0, "\\$[^/][^/]*&quot { print substr($0, RSTART, RLENGTH)}'

vlad
+----------------------------+
| #include<disclaimer.h> |
+----------------------------+

 

[IMAUser]

vlad,

I tried that as it is, but didnt see any output. Maybe it will help if you can tell me what the match is trying to do with the expression &quot;\\$[^/][^/]*&quot;

Thanks

 

[PHV]

Yet another way:

FILE_DIR=`echo $i | sed -n -e 's!.*\(\$[^/]*\).*!\1!p'`

Hope This Help
PH.

 

[vgersh99]

strange.. works with your sample from the original post.

&quot;\\$[^/][^/]*&quot;

Dollar followed by anything BUT a '/' repeated 1 or more times.

vlad
+----------------------------+
| #include<disclaimer.h> |
+----------------------------+