Privoxy eth0 fixes, healthcheck comparison tweak and start.sh update (#…

…2610)

* by checking for existence of default 127 address,healthcheck will never effectively run, as the non-default eth0 ip will never be matched

* the comparison was not trimming off the port, thus healthcheck was always going to fail or denote a change when there may have been none

* simplified the comparison (thanks @edgd1er )

privoxy/scripts/start.sh

@@ -27,7 +27,7 @@ set_port()
   # Set the port for the IPv4 interface
   adr=$(ip -4  a show eth0| grep -oP "(?<=inet )([^/]+)")
   adr=${adr:-"0.0.0.0"}
-  sed -i -E "s/^listen-address\s+127.*/listen-address ${adr}:$1/" "$2"
+  sed -i -E "s/^listen-address\s+.*/listen-address ${adr}:$1/" "$2"
 
   # Remove the listen-address for IPv6 for now. IPv6 compatibility should come later
   sed -i -E "s/^listen-address\s+\[\:\:1.*//" "$2"

scripts/healthcheck.sh

@@ -59,7 +59,7 @@ if [[ ${WEBPROXY_ENABLED} =~ [yY][eE]?[Ss]?|[tT][Rr][Uu][eE] ]]; then
   if [[ ${PROXY} -eq 0 ]]; then
     echo "Privoxy warning: process was stopped, restarting."
   fi
-    proxy_ip=$(grep -oP "(?<=^listen-address).*$" /etc/privoxy/config | sed 's/ //g')
+    proxy_ip=$(grep -oP "(?<=^listen-address )[0-9\.]+" /etc/privoxy/config)
     cont_ip=$(ip -j a show dev eth0 | jq -r .[].addr_info[].local)
     if [[ ${proxy_ip} != ${cont_ip} ]]; then
       echo "Privoxy error: container ip (${cont_ip} has changed: privoxy listening to ${proxy_ip}, restarting privoxy."