#!/bin/sh # OpenWrt Passwall 广告域名列表下载与合并脚本 LOCK_FILE="/var/lock/ad_download.lock" TMP_DIR="/tmp/ad_download" RULES_PATH="/usr/share/passwall/rules" # ========== 加锁,防止重复运行 ========== exec 200>"$LOCK_FILE" if ! flock -n 200; then echo "[ERROR] 脚本已在运行中,退出" exit 1 fi echo $$ >&200 cleanup() { rm -rf "$TMP_DIR" flock -u 200 rm -f "$LOCK_FILE" } trap cleanup EXIT INT TERM rm -rf "$TMP_DIR" mkdir -p "$TMP_DIR" # ========== 单个 URL 的下载与解析函数 ========== process_url() { local idx="$1" local url="$2" local raw="$TMP_DIR/raw_${idx}.txt" local parsed="$TMP_DIR/parsed_${idx}.txt" echo "[INFO] #${idx} 正在下载: $url" wget -q --no-check-certificate -O "$raw" "$url" -T 15 if [ ! -s "$raw" ]; then echo "[WARN] #${idx} 下载失败或文件为空,跳过: $url" return fi # 一次 awk 完成:去\r、去注释、去空行、格式检测、转换、域名校验 awk ' BEGIN { fmt = ""; sample_count = 0 } { # 去除 \r gsub(/\r/, "") # 去除首尾空格 gsub(/^[[:space:]]+|[[:space:]]+$/, "") # 跳过注释和空行 if ($0 ~ /^[#!]/ || $0 ~ /^\[Adblock/ || $0 == "") next # 前20行探测格式 if (sample_count < 20) { sample_count++ if (fmt == "" && $0 ~ /^address=\/[^\/]+\//) fmt = "dnsmasq" if (fmt == "" && $0 ~ /^DOMAIN-SUFFIX,/) fmt = "clash" if (fmt == "" && $0 ~ /^\|\|[a-zA-Z0-9]/) fmt = "adguard" if (fmt == "" && $0 ~ /^(0\.0\.0\.0|127\.0\.0\.1)[[:space:]]/) fmt = "hosts" if (fmt == "" && $0 ~ /^[a-zA-Z0-9]([a-zA-Z0-9.\-]*[a-zA-Z0-9])?\.[a-zA-Z]{2,}$/) fmt = "plain" } # 按格式提取域名 if (fmt == "dnsmasq") { # 匹配 address=/domain/ 和 address=/domain/0.0.0.0 等 n = split($0, a, "/") if (n >= 3 && a[1] == "address=") print a[2] } else if (fmt == "clash") { if (sub(/^DOMAIN-SUFFIX,/, "")) print } else if (fmt == "hosts") { # 0.0.0.0 domain.com 或 127.0.0.1 domain.com if ($0 ~ /^(0\.0\.0\.0|127\.0\.0\.1)[[:space:]]/) { domain = $2 # 去掉行尾可能的注释 gsub(/#.*/, "", domain) gsub(/[[:space:]]/, "", domain) if (domain ~ /^[a-zA-Z0-9]([a-zA-Z0-9.\-]*[a-zA-Z0-9])?\.[a-zA-Z]{2,}$/) print domain } } else if (fmt == "adguard") { if ($0 ~ /^\|\|/) { line = $0 # 去掉开头 || sub(/^\|\|/, "", line) # 去掉 ^ 及之后所有内容(^$third-party 等修饰符) sub(/\^.*/, "", line) # 去掉可能残留的 $ 修饰符 sub(/\$.*/, "", line) # 校验结果是否为合法域名 if (line ~ /^[a-zA-Z0-9]([a-zA-Z0-9.\-]*[a-zA-Z0-9])?\.[a-zA-Z]{2,}$/) print line } } else if (fmt == "plain") { if ($0 ~ /^[a-zA-Z0-9]([a-zA-Z0-9\-]*[a-zA-Z0-9])?(\.[a-zA-Z0-9]([a-zA-Z0-9\-]*[a-zA-Z0-9])?)+$/) print } } END { if (fmt == "") print "[WARN] 未知格式,已丢弃" > "/dev/stderr" } ' "$raw" > "$parsed" rm -f "$raw" # 如果解析结果为空则删除 [ ! -s "$parsed" ] && rm -f "$parsed" } # ========== 读取 UCI 配置 ========== AD_URLS=$(uci -q get passwall.@global[0].ad_url 2>/dev/null) WHITELIST="$(uci -q get passwall.@global[0].white_list 2>/dev/null) ip-api.com" if [ -z "$AD_URLS" ]; then echo "[ERROR] 未找到任何 ad_url 配置" exit 1 fi INDEX=0 for url in $AD_URLS; do [ -z "$url" ] && continue INDEX=$((INDEX + 1)) process_url "$INDEX" "$url" done # ========== 合并、去重、去空行、排除白名单 ========== [ -f $RULES_PATH/my_block_host ] || touch $RULES_PATH/my_block_host if ls "$TMP_DIR"/parsed_*.txt >/dev/null 2>&1; then NEW_FILE="$TMP_DIR/ad_domains_new.txt" awk -v wl="$WHITELIST" ' BEGIN { n = split(wl, a, " ") for (i = 1; i <= n; i++) white[a[i]] = 1 } NF && !seen[$0]++ && !($0 in white) ' "$TMP_DIR"/parsed_*.txt $RULES_PATH/my_block_host > "$NEW_FILE" # ========== 比较新旧文件 hash ========== if [ -f "$RULES_PATH/block_host" ]; then OLD_HASH=$(md5sum "$RULES_PATH/block_host" | awk '{print $1}') else OLD_HASH="" fi NEW_HASH=$(md5sum "$NEW_FILE" | awk '{print $1}') if [ "$OLD_HASH" = "$NEW_HASH" ]; then rm -f "$NEW_FILE" else [ -s $NEW_FILE ] && { rm -f $RULES_PATH/block_host mv -f "$NEW_FILE" "$RULES_PATH/block_host" /etc/init.d/passwall reload > /dev/null 2>&1 & } fi else echo "[ERROR] 没有成功解析任何文件" exit 1 fi [ "$OLD_HASH" != "$NEW_HASH" ] && exit 0 || exit 2