Pi-hole/pi-hole
Pi-hole/pi-hole
1
0
Fork 0
mirror of https://github.com/pi-hole/pi-hole.git synced 2024-11-15 10:43:55 +00:00
Code Issues Projects Releases Packages Wiki Activity

Merge pull request #85 from dschaper/development

Browse source 
Refactored gravity.sh
This commit is contained in:
Jacob Salmela 2015-11-25 20:14:30 -06:00
commit 8ad69cea33
1 changed files with 165 additions and 106 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

271
gravity.sh
View file

@ -1,11 +1,21 @@
#!/bin/bash
#!/usr/bin/env bash
# Pi-hole: A black hole for Internet advertisements
# (c) 2015 by Jacob Salmela GPL 2.0
# Network-wide ad blocking via your Raspberry Pi
# http://pi-hole.net
# Compiles a list of ad-serving domains by downloading them from multiple sources
# This script should only be run after you have a static IP address set on the Pi
piholeIP=$(hostname -I)
piholeIPfile=/tmp/piholeIP
if [[ -f $piholeIPfile ]];then
# If the file exists, it means it was exported from the installation script and we should use that value instead of detecting it in this script
piholeIP=$(cat $piholeIPfile)
rm $piholeIPfile
else
# Otherwise, the IP address can be taken directly from the machine, which will happen when the script is run by the user and not the installation script
piholeIP=$(ip -4 addr show | awk '{match($0,/[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+/); ip = substr($0,RSTART,RLENGTH); print ip}' | sed '/^\s*$/d' | grep -v "127.0.0.1" | (head -n1))
fi
# Ad-list sources--one per line in single quotes
# The mahakala source is commented out due to many users having issues with it blocking legitimate domains. Uncomment at your own risk
sources=('https://adaway.org/hosts.txt'
'http://adblock.gjtech.net/?format=unix-hosts'
#'http://adblock.mahakala.is/'
@ -16,139 +26,152 @@ sources=('https://adaway.org/hosts.txt'
'http://winhelp2002.mvps.org/hosts.txt')
# Variables for various stages of downloading and formatting the list
adList=/etc/pihole/gravity.list
origin=/etc/pihole
piholeDir=/etc/pihole
justDomainsExtension=domains
matter=pihole.0.matter.txt
andLight=pihole.1.andLight.txt
supernova=pihole.2.supernova.txt
eventHorizon=pihole.3.eventHorizon.txt
accretionDisc=pihole.4.accretionDisc.txt
eyeOfTheNeedle=pihole.5.wormhole.txt
basename=pihole
piholeDir=/etc/$basename
adList=$piholeDir/gravity.list
blacklist=$piholeDir/blacklist.txt
whitelist=$piholeDir/whitelist.txt
latentWhitelist=$origin/latentWhitelist.txt
latentWhitelist=$piholeDir/latentWhitelist.txt
justDomainsExtension=domains
matter=$basename.0.matter.txt
andLight=$basename.1.andLight.txt
supernova=$basename.2.supernova.txt
eventHorizon=$basename.3.eventHorizon.txt
accretionDisc=$basename.4.accretionDisc.txt
eyeOfTheNeedle=$basename.5.wormhole.txt
# After setting defaults, check if there's local overrides
if [[ -r $piholeDir/pihole.conf ]];then
echo "** Local calibration requested..."
. $piholeDir/pihole.conf
. $piholeDir/pihole.conf
fi
###########################
# collapse - begin formation of pihole
function gravity_collapse() {
echo "** Neutrino emissions detected..."
# Create the pihole resource directory if it doesn't exist. Future files will be stored here
if [[ -d $piholeDir ]];then
:
# Temporary hack to allow non-root access to pihole directory
# Will update later, needed for existing installs, new installs should
# create this directory as non-root
sudo chmod 777 $piholeDir
find "$piholeDir" -type f -exec sudo chmod 666 {} \;
else
echo "** Creating pihole directory..."
sudo mkdir $piholeDir
echo "** Creating pihole directory..."
mkdir $piholeDir
fi
}
# patternCheck - check to see if curl downloaded any new files, and then process those
# files so they are in host format.
function gravity_patternCheck() {
patternBuffer=$1
# check if the patternbuffer is a non-zero length file
if [[ -s "$patternBuffer" ]];then
# Some of the blocklists are copyright, they need to be downloaded
# and stored as is. They can be processed for content after they
# have been saved.
cp $patternBuffer $saveLocation
echo "Done."
else
# curl didn't download any host files, probably because of the date check
echo "Transporter logic detected no changes, pattern skipped..."
fi
}
# transport - curl the specified url with any needed command extentions, then patternCheck
function gravity_transport() {
url=$1
cmd_ext=$2
agent=$3
# tmp file, so we don't have to store the (long!) lists in RAM
patternBuffer=$(mktemp)
heisenbergCompensator=""
if [[ -r $saveLocation ]]; then
# if domain has been saved, add file for date check to only download newer
heisenbergCompensator="-z $saveLocation"
fi
# Silently curl url
curl -s $cmd_ext $heisenbergCompensator -A "$agent" $url > $patternBuffer
gravity_patternCheck $patternBuffer
# Cleanup
rm -f $patternBuffer
}
# spinup - main gravity function
function gravity_spinup() {
# Loop through domain list. Download each one and remove commented lines (lines beginning with '# 'or '/') and blank lines
for ((i = 0; i < "${#sources[@]}"; i++))
do
url=${sources[$i]}
# Get just the domain from the URL
domain=$(echo "$url" | cut -d'/' -f3)
url=${sources[$i]}
# Get just the domain from the URL
domain=$(echo "$url" | cut -d'/' -f3)
# Save the file as list.#.domain
saveLocation=$origin/list.$i.$domain.$justDomainsExtension
# Save the file as list.#.domain
saveLocation=$piholeDir/list.$i.$domain.$justDomainsExtension
activeDomains[$i]=$saveLocation
agent="Mozilla/10.0"
agent="Mozilla/10.0"
echo -n "Getting $domain list... "
echo -n "Getting $domain list... "
# Use a case statement to download lists that need special cURL commands
# to complete properly and reset the user agent when required
case "$domain" in
"adblock.mahakala.is")
agent='Mozilla/5.0 (X11; Linux x86_64; rv:30.0) Gecko/20100101 Firefox/30.0'
cmd="curl -e http://forum.xda-developers.com/"
;;
# Use a case statement to download lists that need special cURL commands
# to complete properly and reset the user agent when required
case "$domain" in
"adblock.mahakala.is")
agent='Mozilla/5.0 (X11; Linux x86_64; rv:30.0) Gecko/20100101 Firefox/30.0'
cmd_ext="-e http://forum.xda-developers.com/"
;;
"pgl.yoyo.org")
cmd="curl -d mimetype=plaintext -d hostformat=hosts"
;;
"pgl.yoyo.org")
cmd_ext="-d mimetype=plaintext -d hostformat=hosts"
;;
# Default is a simple curl request
*) cmd="curl"
esac
# tmp file, so we don't have to store the (long!) lists in RAM
patternBuffer=$(mktemp)
heisenbergCompensator=""
if [[ -r $saveLocation ]]; then
heisenbergCompensator="-z $saveLocation"
fi
CMD="$cmd -s $heisenbergCompensator -A '$agent' $url > $patternBuffer"
$cmd -s $heisenbergCompensator -A "$agent" $url > $patternBuffer
if [[ -s "$patternBuffer" ]];then
# Remove comments and print only the domain name
# Most of the lists downloaded are already in hosts file format but the spacing/formating is not contigious
# This helps with that and makes it easier to read
# It also helps with debugging so each stage of the script can be researched more in depth
awk '($1 !~ /^#/) { if (NF>1) {print $2} else {print $1}}' $patternBuffer | \
sed -nr -e 's/\.{2,}/./g' -e '/\./p' > $saveLocation
echo "Done."
else
echo "Skipping pattern because transporter logic detected no changes..."
fi
# Cleanup
rm -f $patternBuffer
# Default is a simple request
*) cmd_ext=""
esac
gravity_transport $url $cmd_ext $agent
done
}
# Find all files with the .domains extension and compile them into one file and remove CRs
# Schwarzchild - aggregate domains to one list and add blacklisted domains
function gravity_Schwarzchild() {
# Find all active domains and compile them into one file and remove CRs
echo "** Aggregating list of domains..."
find $origin/ -type f -name "*.$justDomainsExtension" -exec cat {} \; | tr -d '\r' > $origin/$matter
truncate -s 0 $piholeDir/$matter
for i in "${activeDomains[@]}"
do
cat $i |tr -d '\r' >> $piholeDir/$matter
done
# Append blacklist entries if they exist
if [[ -r $blacklist ]];then
numberOf=$(cat $blacklist | sed '/^\s*$/d' | wc -l)
echo "** Blacklisting $numberOf domain(s)..."
cat $blacklist >> $origin/$matter
numberOf=$(cat $blacklist | sed '/^\s*$/d' | wc -l)
echo "** Blacklisting $numberOf domain(s)..."
cat $blacklist >> $piholeDir/$matter
fi
###########################
function gravity_advanced() {
numberOf=$(wc -l < $origin/$andLight)
echo "** $numberOf domains being pulled in by gravity..."
# Remove carriage returns and preceding whitespace
# not really needed anymore?
cp $origin/$andLight $origin/$supernova
# Sort and remove duplicates
sort -u $origin/$supernova > $origin/$eventHorizon
numberOf=$(wc -l < $origin/$eventHorizon)
echo "** $numberOf unique domains trapped in the event horizon."
# Format domain list as "192.168.x.x domain.com"
echo "** Formatting domains into a HOSTS file..."
awk '{print "'"$piholeIP"'" $1}' $origin/$eventHorizon > $origin/$accretionDisc
# Copy the file over as /etc/pihole/gravity.list so dnsmasq can use it
sudo cp $origin/$accretionDisc $adList
kill -HUP $(pidof dnsmasq)
}
# Whitelist (if applicable) then remove duplicates and format for dnsmasq
if [[ -r $whitelist ]];then
# Remove whitelist entries
numberOf=$(cat $whitelist | sed '/^\s*$/d' | wc -l)
plural=; [[ "$numberOf" != "1" ]] && plural=s
echo "** Whitelisting $numberOf domain${plural}..."
function gravity_pulsar() {
# Append a "$" to the end, prepend a "^" to the beginning, and
# replace "." with "\." of each line to turn each entry into a
# regexp so it can be parsed out with grep -x
awk -F '[# \t]' 'NF>0&&$1!="" {print "^"$1"$"}' $whitelist | sed 's/\./\\./g' > $latentWhitelist
# Whitelist (if applicable) domains
if [[ -r $whitelist ]];then
# Remove whitelist entries
numberOf=$(cat $whitelist | sed '/^\s*$/d' | wc -l)
plural=; [[ "$numberOf" != "1" ]] && plural=s
echo "** Whitelisting $numberOf domain${plural}..."
# Append a "$" to the end, prepend a "^" to the beginning, and
# replace "." with "\." of each line to turn each entry into a
# regexp so it can be parsed out with grep -x
awk -F '[# \t]' 'NF>0&&$1!="" {print "^"$1"$"}' $whitelist | sed 's/\./\\./g' > $latentWhitelist
else
rm $latentWhitelist
rm $latentWhitelist
fi
# Prevent our sources from being pulled into the hole
@ -156,10 +179,46 @@ plural=; [[ "${#sources[@]}" != "1" ]] && plural=s
echo "** Whitelisting ${#sources[@]} ad list source${plural}..."
for url in ${sources[@]}
do
echo "$url" | awk -F '/' '{print "^"$3"$"}' | sed 's/\./\\./g' >> $latentWhitelist
echo "$url" | awk -F '/' '{print "^"$3"$"}' | sed 's/\./\\./g' >> $latentWhitelist
done
# Remove whitelist entries from deduped list
grep -vxf $latentWhitelist $origin/$matter > $origin/$andLight
# Remove whitelist entries from list
grep -vxf $latentWhitelist $piholeDir/$matter > $piholeDir/$andLight
}
function gravity_unique() {
# Sort and remove duplicates
sort -u $piholeDir/$supernova > $piholeDir/$eventHorizon
numberOf=$(wc -l < $piholeDir/$eventHorizon)
echo "** $numberOf unique domains trapped in the event horizon."
}
function gravity_hostFormat() {
# Format domain list as "192.168.x.x domain.com"
echo "** Formatting domains into a HOSTS file..."
cat $piholeDir/$eventHorizon | awk '{sub(/\r$/,""); print "'"$piholeIP"' " $0}' > $piholeDir/$accretionDisc
# Copy the file over as /etc/pihole/gravity.list so dnsmasq can use it
cp $piholeDir/$accretionDisc $adList
}
function gravity_advanced() {
# Remove comments and print only the domain name
# Most of the lists downloaded are already in hosts file format but the spacing/formating is not contigious
# This helps with that and makes it easier to read
# It also helps with debugging so each stage of the script can be researched more in depth
awk '($1 !~ /^#/) { if (NF>1) {print $2} else {print $1}}' $piholeDir/$andLight | \
sed -nr -e 's/\.{2,}/./g' -e '/\./p' > $piholeDir/$supernova
numberOf=$(wc -l < $piholeDir/$supernova)
echo "** $numberOf domains being pulled in by gravity..."
gravity_unique
sudo kill -HUP $(pidof dnsmasq)
}
gravity_collapse
gravity_spinup
gravity_Schwarzchild
gravity_pulsar
gravity_hostFormat
gravity_advanced