Fix HTML parsing when setting TAG with wget

Now the parsing part is the same as in the curl scenario

scripts/get

@@ -80,7 +80,7 @@ checkDesiredVersion() {
   if type "curl" > /dev/null; then
     TAG=$(curl -SsL $release_url | awk '/\/tag\//' | grep -v no-underline | cut -d '"' -f 2 | awk '{n=split($NF,a,"/");print a[n]}' | awk 'a !~ $0{print}; {a=$0}')
   elif type "wget" > /dev/null; then
-    TAG=$(wget -q -O - $release_url | awk '/\/tag\//' | cut -d '"' -f 2 | awk '{n=split($NF,a,"/");print a[n]}' | awk 'a !~ $0{print}; {a=$0}')
+    TAG=$(wget -q -O - $release_url | awk '/\/tag\//' | grep -v no-underline | cut -d '"' -f 2 | awk '{n=split($NF,a,"/");print a[n]}' | awk 'a !~ $0{print}; {a=$0}')
   fi
   if [ "x$TAG" == "x" ]; then
     echo "Cannot determine ${DESIRED_VERSION} tag."