adblock-generator.sh

adblock-generator.sh

#!bin/bash

SUBSCRIPTION_LIST=(
https://easylist-downloads.adblockplus.org/easyprivacy.txt
https://easylist-downloads.adblockplus.org/easylistchina.txt
https://easylist-downloads.adblockplus.org/easylist.txt
https://easylist-downloads.adblockplus.org/malwaredomains_full.txt
https://easylist-downloads.adblockplus.org/exceptionrules.txt
)

IP=$(ifconfig eth0 | awk '/inet addr/{print substr($2,6)}')
SAVE_PATH='/etc/pihole'
ADBLOCK_RULES='adblock-rules.txt'
ADBLOCK_EXCEPT_RULES='adblock-except-rules.txt'
ADBLOCK_HOST='adblock-host.list'

generate_adblock_rules() {
    TMP_DIR='/tmp/adblock-rules'
    TMP_CACHE='.cache'

    LIST=$1
    RULE_FILE=$2
    EXCEPT_RULE_FILE=$3

    if [[ -z "${LIST[@]}" || -z "${RULE_FILE}" || -z "${EXCEPT_RULE_FILE}" ]]; then
        echo "At least one input is empty!"
        exit 1
    fi

    rm -rf ${TMP_DIR} && mkdir -p ${TMP_DIR}

    # regular rules
    for URL in "${LIST[@]}"; do
        curl -s "${URL}" > "${TMP_DIR}/${TMP_CACHE}"
        # regular rules
        egrep '^\|\|[A-Za-z0-9\.-]+\^$' "${TMP_DIR}/${TMP_CACHE}" | awk '{print substr($0, 3, length($0)-3)}' >> "${TMP_DIR}/adblock.rules"
        # exception rules
        egrep '^@@\|\|[A-Za-z0-9\.-]+\^$' "${TMP_DIR}/${TMP_CACHE}" | awk '{print substr($0, 5, length($0)-5)}' >> "${TMP_DIR}/adblock.exception-rules"
    done

    # merge hosts
    awk '!a[$0]++' "${TMP_DIR}/adblock.rules" > "${RULE_FILE}"
    awk '!a[$0]++' "${TMP_DIR}/adblock.exception-rules" > "${EXCEPT_RULE_FILE}"
    
    rm -rf ${TMP_DIR}
}

generate_host() {
    awk "{print \"${IP}\", \$0}" $1 > $2
}

generate_adblock_rules ${SUBSCRIPTION_LIST} "${SAVE_PATH}/${ADBLOCK_RULES}" "${SAVE_PATH}/${ADBLOCK_EXCEPT_RULES}"

generate_host "${SAVE_PATH}/${ADBLOCK_RULES}" "${SAVE_PATH}/${ADBLOCK_HOST}"