#!/bin/bash # Intellectual property of ghettosearch.net!! Patents pending!!!! set -ue query="$1" url="$2" rm -f stack visited export PATH=/pt/bin:/Users/brendano/sw/bin:$PATH # query=$(echo "$query" | sed -e 's/^ *//; s/ *$//' | tr -s ' ' ' ' | tr ' ' '|' ) query=$(echo "$query" | sed -e 's/^ *//; s/ *$//' | tr -s ' ' ' ' ) while true; do (set -x; wget -O - -q -T 1 -t 1 "$url" > page) if lynx -dump page -force_html -nolist -nonumbers | egrep -i "$query" > snippet; then cat snippet | egrep -i --color "$query" exit 0 fi cat page | sed -e 's/href=/!href=/g' | tr '!' '\n' | grep '^href=' | sed -e 's/^href=//; s/^"//; s/["<> ].*//; s/>.*//' | grep '^http:' | egrep -v 'ad\.|ads\.|rss|feed|\.(css|jpg|png|gif|jpeg|pdf|doc)$' | shuffle >> stack echo "$url" >> visited grep -vFf visited stack | tail -20000 > new_stack mv new_stack stack url="$(tail -1 stack)" if [[ ! "$url" ]]; then echo "No answer found" exit 0 fi done