#!/bin/sh

# (c) Dinko Korunic, InfoMAR, 2008
# SquidGuard merge blacklists

# bail out on error
set -e

# variables
TARGET=/var/lib/squidguard/tmp
TMPDIR=bl-final
SADIR=/var/lib/squidguard/db
LOCKFILE=/var/lock/squidguard-black-gen.lock

# merge lists
merge_bl() {
	SRC="$1"
	DST="$2"

	# sanity checking
	if [ -z "$SRC" -o -z "$DST" ]; then
		return
	fi

	if [ ! -r "$SRC" ]; then
		return
	fi

	cat "$SRC" >> "$DST"

	sort -u "$DST" | grep -v '^#' > "$DST.$$" && \
		mv -f "$DST.$$" "$DST"
	rm -f "$DST.$$"
}

# iterate lists, sort and merge if possible
add_bl() {
	SRCDIR="$1"
	DSTDIR="$2"

	# sanity checking
	if [ -z "$SRCDIR" -o -z "$DSTDIR" ]; then
		return
	fi
	if [ ! -d "$SRCDIR" ]; then
		return
	fi

	# possible that it is empty
	mkdir -p "$DSTDIR"

	# iterate the merge procedure
	DIRS=$(find "$SRCDIR" \( -type d -o -type l \))
	for dir in $DIRS; do
		# extract first-level name
		short=$(echo "$dir" | cut -d/ -f2)
		for file in domains urls; do
			if [ -e "$dir/$file" ]; then
				mkdir -p "$DSTDIR/$short"
				merge_bl "$dir/$file" "$DSTDIR/$short/$file"
			fi
		done
	done

	rm -rf "$SRCDIR"
}

# lock us
lockfile -r1 "$LOCKFILE"
trap "rm -f $LOCKFILE $TARGET/$TMPDIR; exit" INT TERM EXIT

mkdir -p "$TARGET"
cd "$TARGET"

# cleanup
rm -rf bl BL blacklists "$TMPDIR" *tgz *tar.gz
mkdir "$TMPDIR"

# if possible, merge the local lists
if [ -d crolist ]; then
	cp -a crolist/* "$TMPDIR"
fi
    
# mesd
wget http://squidguard.mesd.k12.or.us/blacklists.tgz
tar -xvf blacklists.tgz
rm -f blacklists.tgz
add_bl blacklists "$TMPDIR"

# shalla
wget http://squidguard.shalla.de/Downloads/shallalist.tar.gz
tar -xzf shallalist.tar.gz
rm -f shallalist.tar.gz
add_bl BL "$TMPDIR"

# touloise
wget ftp://ftp.univ-tlse1.fr/pub/reseau/cache/squidguard_contrib/blacklists.tar.gz
tar -xvf blacklists.tar.gz
rm -f blacklists.tar.gz
add_bl blacklists "$TMPDIR"

# post actions (be atomic if possible)
mv "$SADIR" "$SADIR.$$" && mv "$TARGET/$TMPDIR" "$SADIR"
/usr/bin/squidGuard -c /etc/squid/squidGuard.conf -C all || true
chown -R proxy:proxy "$SADIR"
/usr/sbin/squid -k reconfigure
rm -rf "$TARGET/$TMPDIR" "$SADIR.$$"

# unlock
rm -f "$LOCKFILE"
trap - INT TERM EXIT