Como ya está utilizando awk y buscan una solución sencilla:

awk -F/ '{n=split($3, a, "."); printf("%s.%s", a[n-1], a[n])}' <<< 'http://www.example.com/index.php'
      ^ ^   ^^^^^^^^^^^^^^^^^^                  ^^^^^^^^^^^^
      | |          |                                  |
      | |          |                            last two elements 
      | |          |
      | |          +--- Split the 3rd field (aka the part after //) into
      | |               the array 'a', using '.' as the separator for splitting.
      | |               Returns the number of created array elements in 'n'.
      | |
      | +-------------- The awk code between the '' gets run once for every
      |                 input line, with the fields split by -F/ stored in
      |                 $1, $2 etc. In our case $1 contains "http:", $2 is 
      |                 empty, $3 contains "www.example.com" and $4 etc. the
      |                 various path elements (if there are any)
      |
      +---------------- Split the input lines into fields, separated by '/'