#!/bin/bash
echo "Slurping URLs ..."
echo
sitedomain=$1
wget --spider --recursive --level=inf --no-verbose --output-file=get_urls.txt $sitedomain
grep -i URL get_urls.txt | awk -F 'URL:' '{print $2}' | awk '{$1=$1};1' | awk '{print $1}' | sort -u | sed '/^$/d' > urls_$sitedomain.txt
rm get_urls.txt
rm -r $sitedomain
echo "Done !"
echo
ls -lArth | grep urls_$sitedomain.txt