Scripted update from someonewhocares.org preserving local changes.

update_sane_hosts.sh

#!/bin/bash -ex

# Keep /etc/hosts up to date with ad blocker site:
# http://someonewhocares.org/hosts/hosts

# Depends on my endecode.py gist or equiv for cleaning up html text.

# Excluding personal allow list
# eg: googleadservices, so google image searches are clickable.
allow=${hosts}.allow

# Optional files 
#   /etc/hosts.local for local additions
#   /etc/hosts.allow for ones to prune from source

hosts="/etc/hosts"
backup="/tmp/$(basename ${hosts}).$(date +%x | tr '/' '_')"
allow=${hosts}.allow
local=${hosts}.local
tmp=$(mktemp)

sudo cp $hosts $backup

if [ -e ${local} ] ; then
    cp $local $tmp
fi

curl -s http://someonewhocares.org/hosts/hosts | endecode.py -a html_decode >> $tmp
if [ -e ${allow} ] ; then
    mv $tmp $tmp.1
    grep -v -f $allow $tmp.1 > $tmp
    rm -rf $tmp.1
fi


sudo mv $tmp /etc/hosts

[ "$DEBUG" ] && diff $backup /etc/hosts
rm -f $tmp