✨ feat(manga,device): 漫画免签名兜底与 phone 注册身份选项

- /manga 设备池签名路线全部失败时自动降级 wkbrowser 免签名路线（materials/skeleton 三段策略收集原图直链，无需设备/签名/解密），响应附 source: wkbrowser - 新增配置 fanqie.device_identity（默认 pad 保持现状；phone 对齐书源真机注册 payload：host_bit 64 + 补 hardware/device_device/fingerprint 字段） - README 与 openapi.yaml 同步文档，新增 manga_fallback 解析测试
✨ feat(api): 新增榜单书架历史与搜索续传并重构正文缓存链路
2026-10-12 14:53:08 +08:00 · 2026-10-12 14:49:14 +08:00 · 2026-10-10 13:36:25 +08:00 · 2026-10-10 12:58:25 +08:00
38 changed files with 1770 additions and 381 deletions
@@ -32,8 +32,10 @@ SRC_DIR="/root/code/novel/fanqie-go"
 DOMAIN="fq.20010117.xyz"
 # ---------- 证书 ----------
-# Certbot 申请证书的邮箱，用于接收过期提醒
+# acme.sh 申请证书的邮箱，用于账户注册与过期提醒
 CERT_EMAIL="you@example.com"
 # 证书安装目录（acme.sh --install-cert 目标，nginx 配置从此处读取）
 # CERT_DIR="/etc/nginx/ssl"
 # ---------- 编译 ----------
 # BUILD_TARGET="./cmd/fanqie"
@@ -57,3 +57,4 @@ fanqie-*
 .env
 .finesse/
 fanqie.log
 .codegraph/
@@ -30,6 +30,12 @@ Go 语言实现的番茄小说非官方 API 服务，完整实现了签名算法
 - 分类首页数据
 - 分类详情页（支持筛选、排序、分页）
 ### 榜单与个人数据
 - 🏆 榜单（推荐 / 完结 / 热搜 / 黑马 / 爆更）
 - 📚 我的书架（按分组浏览，需 sessionid）
 - 🕐 阅读历史（含阅读进度，需 sessionid）
 ### 技术特性
 - **签名算法** - Gorgon、8404、Argus、Ladon、Khronos
@@ -171,7 +177,23 @@ go run ./cmd/fanqie
 | 接口 | 方法 | 说明 | 参数 |
 | ------ | ------ | ------ | ------ |
 | `/category-front` | GET | 分类首页 | - |
-| `/category-landing` | GET | 分类详情 | `category_id`, `word_number`, `book_status`, `sort_by`, `offset` |
+| `/category-landing` | GET | 分类详情 | `category_id`, `word_number`, `book_status`, `sort_by`, `offset`, `selected_items`（新版分类标签自带筛选项，透传上游） |
 ### 榜单与个人数据
 | 接口 | 方法 | 说明 | 参数 |
 | ------ | ------ | ------ | ------ |
 | `/rank` | GET | 榜单（免签名接口） | `side_type`（`10` 推荐 / `11` 完结 / `12` 热搜 / `13` 黑马 / `15` 爆更，默认推荐）, `page`（默认 1，每页 20 本） |
 | `/bookshelf` | GET | 我的书架（需登录） | `sessionid`（必填，见下方「获取 sessionid」）, `group`（可选，按分组过滤；响应含每本书的 `group_name`） |
 | `/read-history` | GET | 阅读历史（需登录） | `sessionid`（必填）, `offset`（默认 0，每页 20 条；响应每项的 `history` 字段保留上游原始记录，含阅读进度） |
 #### 获取 sessionid
 番茄网页版登录由字节风控（滑块验证）保护，无法通过 API 直接登录（书源实现同样依赖浏览器人工登录）。获取方式：
 1. 浏览器打开 [fanqienovel.com](https://fanqienovel.com/) 并登录（手机验证码或密码）
 2. F12 → Application → Cookies → 复制 `sessionid` 的值（有效期较长，可长期使用）
 3. 直接传入 `/bookshelf?sessionid=...` 或 `/read-history?sessionid=...` 使用（书架为空多半是 sessionid 无效或已过期）
 ### 媒体相关
@@ -256,6 +278,11 @@ fanqie:
  # - "sign4": X-Gorgon + X-Ladon + X-Argus + X-Khronos
  signature: "sign4"
  # 注册设备身份:
  # - "pad": 平板身份（默认，历史行为）
  # - "phone": 手机身份（对齐书源真机注册 payload，可对照测试注册率/存活率）
  device_identity: "pad"
  # 内容处理
  content:
    remove_images: false        # 移除图片
@@ -288,7 +315,7 @@ cache:
    password: ""           # 密码
    database: "fanqie"     # 数据库名
    sslmode: "disable"     # SSL 模式: disable | prefer | require
-  redis:                     # Redis 缓存（正文缓存与音频代理缓存，统计数据/设备池仍在 SQL 库）
+  redis:                     # Redis 热缓存层（启用后与 SQL 双写：Redis 短 TTL + SQL 持久层 30 天）
    enabled: false
    host: "127.0.0.1"      # 主机
    port: 6379             # 端口
@@ -337,6 +364,7 @@ cp .env.example .env
 | `server.protect_all` | 业务接口是否也要求访问密钥 | `false` |
 | `fanqie.algorithm` | 签名算法 `gorgon` / `8404` | `gorgon` |
 | `fanqie.signature` | 签名方法 `sign2` / `sign4` | `sign4` |
 | `fanqie.device_identity` | 注册设备身份 `phone` / `pad` | `pad` |
 | `fanqie.content.*` | 正文内容处理选项 | 见上文 |
 | `fanqie.manga.output_mode` | 漫画输出模式 | `proxy` |
 | `cache.enabled` | 是否启用正文内容缓存 | `true` |
@@ -388,15 +416,19 @@ cp .env.example .env
 ## 缓存机制
- **接口缓存** - 命中直接返回，不打上游（Redis 启用时优先走 Redis）
+Redis 启用时为**双层架构**：Redis 热层（TTL=业务缓存时长）+ SQL 持久层（不可变内容固定保留 30 天）。
- **自动清理** - 定期清理过期缓存（Redis 由 TTL 自动过期）
+
 - **读**：Redis → 未命中查 SQL（命中自动回填 Redis）→ 回源上游
 - **写**：Redis + SQL 双写；搜索/分类等动态短缓存只走 Redis 不落库
 - **防击穿**：singleflight 合并并发未命中请求，同一 key 只有一个请求打上游
 - **自动清理** - 定期按行内过期时间清理 SQL 过期缓存（Redis 由 TTL 自动过期）
 - **数据库压缩** - 支持 VACUUM 操作优化数据库
 | 接口 | 缓存时长 |
 | ------ | ------ |
-| `/content` | 24 小时 |
+| `/content`、`/multi-content` | 24 小时（SQL 层 30 天） |
-| `/detail`、`/info`、`/catalog`、`/multi-detail`、`/item-summary` | 24 小时 |
+| `/detail`、`/info`、`/catalog`、`/multi-detail`、`/item-summary` | 24 小时（SQL 层 30 天） |
-| `/search`、`/category-front`、`/category-landing` | 10 分钟 |
+| `/search`、`/category-front`、`/category-landing` | 10 分钟（仅 Redis） |
 | `/audio` 代理链接 | 2 小时（匹配 CDN 有效期） |
 失败响应不缓存；评论类接口为动态数据，不缓存。
@@ -1,61 +1,63 @@
 [
  {
-    "bookSourceComment": "fanqie-go 聚合书源：一个源覆盖 小说/听书/漫画/短剧。\n\n【设置入口】发现页顶部选择按钮：模式/音色/清晰度/段评/图片/频道，点选即生效；",
+    "bookSourceName": "🔅番茄聚合",
-    "bookSourceGroup": "🍅 番茄",
+    "bookSourceGroup": "📖番茄小说",
    "bookSourceName": "🍅 番茄聚合",
    "bookSourceType": 0,
    "bookSourceUrl": "https://fq.20010117.xyz",
-    "customOrder": 0,
+    "bookSourceComment": "番茄聚合书源（fanqie-go 后端）\n\n基础使用教程：\n1、搜索：\n    1) 基础搜索：直接搜，默认搜索当前模式内容\n    2) 快捷搜索：使用特殊关键词(x,t,m,d)搜索\n         x:书名  如(x:十日终焉)  搜小说（可省略）\n         t:书名  如(t:十日终焉)  搜听书\n         m:书名  如(m:十日终焉)  搜漫画\n         d:书名  如(d:十日终焉)  搜短剧\n         冒号支持中文英文，也可在书源设置中切换默认搜索模式\n2、发现页：\n    顶部筛选选择 类型/频道/字数/更新/排序 后，点击分类宫格浏览，下拉翻页\n3、段评：\n    阅读正文时点击段尾气泡即可查看段落评论，样式可在段评设置中调整\n4、听书/漫画/短剧：\n    听书自动用播放器播放；漫画走服务端解密代理；短剧打开章节自动唤起播放\n\n更换服务器地址：只需修改源URL(bookSourceUrl)，全部规则使用相对地址自动跟随",
-    "enabled": true,
+    "bookSourceEnabled": true,
    "enabledCookieJar": false,
    "enabledExplore": true,
-    "header": "{\"User-Agent\": \"Mozilla/5.0 (Linux; Android 13; 2210132C Build/TP1A.220624.014; wv) AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/116.0.5845.92 Mobile Safari/537.36\"}",
+    "enabledCookieJar": true,
-    "jsLib": "// ============ 源变量配置（照光遇模式：defaultConfig 兜底，全 try-catch） ============\nconst defaultConfig = {\n    模式: \"小说\",\n    音色: \"0\",\n    清晰度: \"max\",\n    段评: \"开\",\n    图片: \"保留\"\n};\n\nconst fqVarsKey = 'fq_vars';\n\nfunction getVariable(k) {\n    if (k == undefined) k = \"\";\n    var parsed = {};\n    try {\n        var raw = String(cache.get(fqVarsKey) || '');\n        if (raw) parsed = JSON.parse(raw);\n    } catch (e) {}\n    if (k == \"\") return parsed;\n    var value = parsed[k];\n    if (value == undefined) value = defaultConfig[k];\n    return value != undefined ? value : \"\";\n}\n\nfunction setVariable(k, v, t) {\n    if (t == undefined) t = true;\n    var parsed = {};\n    try {\n        var raw = String(cache.get(fqVarsKey) || '');\n        if (raw) parsed = JSON.parse(raw);\n    } catch (e) {}\n    parsed[k] = v;\n    try { cache.put(fqVarsKey, JSON.stringify(parsed)); } catch (e) {}\n    if (t) java.toast('\\n设置 ' + k + ' 为 ' + v);\n}\n\n// 发现页筛选按钮回调：存变量并刷新发现页\nfunction show(m, t) {\n    setVariable(t, m, false);\n    java.toast('\\n已设置' + t + '：' + m + '\\n已在书架的书切换模式后需刷新目录');\n    try { java.refreshExplore(); } catch (e) {}\n}\n\nfunction createFilter(title, chars, defaultVal, paramKey, size, viewName) {\n    if (viewName == undefined) viewName = \"'请选择类型'\";\n    return {\n        title: title,\n        type: \"select\",\n        chars: chars,\n        default: defaultVal,\n        \"viewName\": viewName,\n        action: `show(infoMap['${title}'],'${paramKey}')`,\n        style: {\n            layout_flexGrow: 1,\n            layout_flexBasisPercent: size\n        }\n    };\n}\n\n// ============ 基础工具 ============\n// 服务地址：source 可用时取书源地址，否则用常量兜底（部分阅读版本正文环境不注入 source）\nvar FQ_BASE = 'https://fq.20010117.xyz';\n\nfunction BaseUrl() {\n    try {\n        if (typeof source != 'undefined' && source && source.getKey) {\n            var u = String(source.getKey()).replace(/\\/+$/, '');\n            if (u.indexOf('http') == 0) return u;\n        }\n    } catch (e) {}\n    return FQ_BASE;\n}\n\nfunction request(url) {\n    let urla = url;\n    if (!url.includes('http')) urla = this.BaseUrl() + url;\n    return java.ajax(urla);\n}\n\nfunction parseJsonSafely(str, defaultValue) {\n    if (defaultValue == null) defaultValue = null;\n    try {\n        return JSON.parse(str);\n    } catch (e) {\n        return defaultValue;\n    }\n}\n\nfunction isIOS() {\n    try {\n        java.deviceID();\n        return true;\n    } catch (e) {\n        return false;\n    }\n}\n\nfunction isMod() {\n    try {\n        if (typeof java.reLoginView == 'function') return true;\n        new Packages.io.legato.kazusa.utils.TimeoutCancellationException('');\n        return true;\n    } catch (e) {\n        return false;\n    }\n}\n\n// 纯 JS base64（UTF-8），规避部分阅读版本 base64 编码 API 不可用\nfunction b64encode(str) {\n    var chars = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/';\n    var bytes = [];\n    str = String(str);\n    for (var i = 0; i < str.length; i++) {\n        var c = str.charCodeAt(i);\n        if (c < 0x80) {\n            bytes.push(c);\n        } else if (c < 0x800) {\n            bytes.push(0xC0 | (c >> 6), 0x80 | (c & 63));\n        } else {\n            bytes.push(0xE0 | (c >> 12), 0x80 | ((c >> 6) & 63), 0x80 | (c & 63));\n        }\n    }\n    var out = '';\n    for (var j = 0; j < bytes.length; j += 3) {\n        var b1 = bytes[j];\n        var b2 = j + 1 < bytes.length ? bytes[j + 1] : undefined;\n        var b3 = j + 2 < bytes.length ? bytes[j + 2] : undefined;\n        out += chars[b1 >> 2];\n        out += chars[((b1 & 3) << 4) | (b2 == undefined ? 0 : b2 >> 4)];\n        out += b2 == undefined ? '=' : chars[((b2 & 15) << 2) | (b3 == undefined ? 0 : b3 >> 6)];\n        out += b3 == undefined ? '=' : chars[b3 & 63];\n    }\n    return out;\n}\n\n// ============ 段评气泡（/content?review=keep 注入 <comment> 标记，点击打开 /review_html 页面） ============\nfunction createSvg(number, url) {\n    var t = number > 99 ? '99+' : String(number);\n    var svg = `<svg xmlns='http://www.w3.org/2000/svg' width='100' height='90' viewBox='0 0 25 17' fill='none'><path stroke='currentColor' stroke-width='1.00' stroke-linecap='round' stroke-linejoin='round' stroke-miterlimit='4.00' style='stroke:#A7A7A7;stroke-opacity:0.5;stroke-width:1px;stroke-linejoin:round;fill:transparent;fill-opacity:0.5;' d='M 7.5 0.5 L 16.5 0.5 A 7 7 0 0 1 23.5 7.5 L 23.5 7.5 A 7 7 0 0 1 16.5 14.5 L 8.5 14.5 L 5.192 16.5 L 5.5 14.5 A 7 7 30 0 1 0.5 7.5 L 0.5 7.5 A 7 7 0 0 1 7.5 0.5 Z'/><text fill='#A7A7A7' font-family='Arial, sans-serif' font-size='10' font-weight='500' text-anchor='middle' dy='0.35em' x='12.00' y='7.50'>$Line truncated
+    "customOrder": 0,
-    "lastUpdateTime": 1791569653299,
+    "weight": 0,
    "respondTime": 180000,
-    "exploreUrl": "<js>\nvar groups = {\"🔥 男频精选\": [{\"title\": \"玄幻\", \"url\": \"/category-landing?category_id=7&offset={{(page-1)*20}}\"}, {\"title\": \"都市\", \"url\": \"/category-landing?category_id=1&offset={{(page-1)*20}}\"}, {\"title\": \"历史\", \"url\": \"/category-landing?category_id=12&offset={{(page-1)*20}}\"}, {\"title\": \"悬疑\", \"url\": \"/category-landing?category_id=10&offset={{(page-1)*20}}\"}, {\"title\": \"衍生\", \"url\": \"/category-landing?category_id=1079&offset={{(page-1)*20}}\"}, {\"title\": \"同人\", \"url\": \"/category-landing?category_id=538&offset={{(page-1)*20}}\"}, {\"title\": \"科幻末世\", \"url\": \"/category-landing?category_id=8&offset={{(page-1)*20}}\"}, {\"title\": \"奇幻仙侠\", \"url\": \"/category-landing?category_id=259&offset={{(page-1)*20}}\"}, {\"title\": \"种田\", \"url\": \"/category-landing?category_id=23&offset={{(page-1)*20}}\"}, {\"title\": \"游戏体育\", \"url\": \"/category-landing?category_id=746&offset={{(page-1)*20}}\"}, {\"title\": \"系统\", \"url\": \"/category-landing?category_id=19&offset={{(page-1)*20}}\"}, {\"title\": \"重生\", \"url\": \"/category-landing?category_id=36&offset={{(page-1)*20}}\"}, {\"title\": \"穿越\", \"url\": \"/category-landing?category_id=37&offset={{(page-1)*20}}\"}, {\"title\": \"多女主\", \"url\": \"/category-landing?category_id=91&offset={{(page-1)*20}}\"}, {\"title\": \"修仙\", \"url\": \"/category-landing?category_id=517&offset={{(page-1)*20}}\"}, {\"title\": \"二次元\", \"url\": \"/category-landing?category_id=39&offset={{(page-1)*20}}\"}, {\"title\": \"搞笑轻松\", \"url\": \"/category-landing?category_id=778&offset={{(page-1)*20}}\"}, {\"title\": \"单女主\", \"url\": \"/category-landing?category_id=389&offset={{(page-1)*20}}\"}, {\"title\": \"悬疑灵异\", \"url\": \"/category-landing?category_id=751&offset={{(page-1)*20}}\"}, {\"title\": \"末日求生\", \"url\": \"/category-landing?category_id=515&offset={{(page-1)*20}}\"}, {\"title\": \"娱乐圈\", \"url\": \"/category-landing?category_id=43&offset={{(page-1)*20}}\"}, {\"title\": \"末世\", \"url\": \"/category-landing?category_id=68&offset={{(page-1)*20}}\"}], \"🌸 女频精选\": [{\"title\": \"现代言情\", \"url\": \"/category-landing?category_id=3&offset={{(page-1)*20}}\"}, {\"title\": \"古代言情\", \"url\": \"/category-landing?category_id=5&offset={{(page-1)*20}}\"}, {\"title\": \"纯爱\", \"url\": \"/category-landing?category_id=1012&offset={{(page-1)*20}}\"}, {\"title\": \"甜宠\", \"url\": \"/category-landing?category_id=96&offset={{(page-1)*20}}\"}, {\"title\": \"总裁\", \"url\": \"/category-landing?category_id=29&offset={{(page-1)*20}}\"}, {\"title\": \"幻想言情\", \"url\": \"/category-landing?category_id=32&offset={{(page-1)*20}}\"}, {\"title\": \"双男主\", \"url\": \"/category-landing?category_id=275&offset={{(page-1)*20}}\"}, {\"title\": \"双女主\", \"url\": \"/category-landing?category_id=704&offset={{(page-1)*20}}\"}, {\"title\": \"年代\", \"url\": \"/category-landing?category_id=79&offset={{(page-1)*20}}\"}, {\"title\": \"快穿\", \"url\": \"/category-landing?category_id=24&offset={{(page-1)*20}}\"}, {\"title\": \"豪门总裁\", \"url\": \"/category-landing?category_id=748&offset={{(page-1)*20}}\"}, {\"title\": \"追妻火葬场\", \"url\": \"/category-landing?category_id=616&offset={{(page-1)*20}}\"}, {\"title\": \"无CP\", \"url\": \"/category-landing?category_id=392&offset={{(page-1)*20}}\"}, {\"title\": \"大女主\", \"url\": \"/category-landing?category_id=760&offset={{(page-1)*20}}\"}, {\"title\": \"女强\", \"url\": \"/category-landing?category_id=86&offset={{(page-1)*20}}\"}], \"🏷️ 男频标签\": [{\"title\": \"古代\", \"url\": \"/category-landing?category_id=758&offset={{(page-1)*20}}\"}, {\"title\": \"历史古代\", \"url\": \"/category-landing?category_id=273&offset={{(page-1)*20}}\"}, {\"title\": \"轻小说\", \"url\": \"/category-landing?category_id=858&offset={{(page-1)*20}}\"}, {\"title\": \"都市生活\", \"url\": \"/category-landing?category_id=2&offset={{(page-1)*20}}\"}, {\"title\": \"都市高武\", \"url\": \"/category-landing?category_id=1014&offset={{(page-1)*20}}\"}, {\"title\": \"游戏动漫\", \"url\": \"/category-landing?category_id=57&offset={{(page-1)*20}}\"}, {\"title\": \"抗战谍战\", \"url\": \"/category-landing?category_id=504&offset={{(page-1)*20}}\"}, {\"title\": \"体育\", \"url\": \"/category-landing?category_id=15&offset={{(page-1)*20}}\"}, {\"title\": \"都市异能\", \"url\": \"/category-landing?category_id=516&offset={{(page-1)*20}}\"}, {\"title\": \"武侠\", \"url\": \"/category-landing?category_id=16&offset={{(page-1)*20}}\"}, {\"title\": \"东方仙侠\", \"url\": \"/category-landing?category_id=1140&offset={{(page-1)*20}}\"}, {\"title\": \"东方玄幻\", \"url\": \"/category-landing?category_id=511&offset={{(page-1)*20}}\"}, {\"title\": \"男频衍生\", \"url\": \"/category-landing?categorLine truncated
+    "lastUpdateTime": 1791607295124,
-    "searchUrl": "<js>\nvar tab = getVariable('模式');\nvar m = String(key).match(/^([xtmdXTMD])[::：]/);\nvar q = String(key);\nif (m) {\n    tab = {x: '小说', t: '听书', m: '漫画', d: '短剧'}[m[1].toLowerCase()];\n    q = q.slice(2);\n}\njava.put('fq_search_tab', tab);\nvar typeMap = {'小说': 3, '听书': 2, '漫画': 8, '短剧': 11};\n`/search?query=${encodeURIComponent(q)}&tab_type=${typeMap[tab] || 3}&offset=${(page - 1) * 10}`;\n</js>",
+    "bookUrlPattern": "https?:\\/\\/(?:[a-zA-Z0-9-]+\\.)?[a-zA-Z0-9-]+(?::\\d+)?\\/detail\\?book_id=\\d+.*",
    "jsLib": "// 当前书源版本号\nlet localVersion = '26.10.10';\n\n// 初始配置\nconst defaultConfig = {\n    发现页类型: \"小说\",\n    频道: \"男频\",\n    搜索模式: \"小说\"\n};\n\n// 获取源变量\nfunction getVariable(k) {\n    if (k == undefined) k = \"\";\n    const { source } = this;\n    let parsed = {};\n    try {\n        parsed = JSON.parse(source.getVariable());\n    } catch {}\n    if (k == \"\") {\n        return parsed;\n    }\n    let value = parsed[k];\n    if (value == undefined) {\n        value = defaultConfig[k];\n    }\n    return value != undefined ? value : \"\";\n}\n\n// 设置源变量\nfunction setVariable(k, v, t) {\n    if (t == undefined) t = true;\n    const { source, java } = this;\n    const vs = this.getVariable();\n    vs[k] = v;\n    source.setVariable(JSON.stringify(vs, null, 4));\n    if (k != '搜索关键词' && t) {\n        java.toast(`\\n设置 ${k} 为 ${v}`)\n    }\n}\n\n// 还原\nfunction deleteVariable() {\n    const { source, java } = this;\n    source.setVariable('');\n    java.longToast('\\n所有设置已还原到初始化导入状态');\n}\n\n// 从 URL 提取内容类型标记(n小说/t听书/m漫画/d短剧)。\n// 注意：jsLib 函数作用域看不到规则变量 baseUrl，必须由调用方显式传入\nfunction fqTab(url, def) {\n    if (def == undefined) def = 'n';\n    let m = String(url).match(/fqtab=([ntmd])/);\n    return m ? m[1] : def;\n}\n\n// 判断当前环境\nfunction checkEnv() {\n    const {\n        java,\n        source\n    } = this;\n\n    try {\n        java.qread();\n        return \"轻阅读\";\n    } catch (e) {}\n\n    try {\n        if (typeof java.reLoginView == 'function') {\n            return \"改版\";\n        }\n        new Packages.io.legato.kazusa.utils.TimeoutCancellationException('');\n        return \"改版\";\n    } catch (e) {}\n    try {\n        java.deviceID();\n        return \"苹果\";\n    } catch (e) {}\n\n    if (typeof source.loginUi == 'function') {\n        return \"安卓\";\n    }\n\n    return \"改版\";\n}\n\n// 备用设置方案（登录页按钮）\nfunction ste(v) {\n    const { java } = this;\n    let moreSettings = this.getVariable(\"更多设置\") || {};\n    let paraSettings = this.getVariable(\"段评设置\") || {};\n    if (v == \"段评\") {\n        if (paraSettings[\"段评开关\"] != \"false\") {\n            paraSettings[\"段评开关\"] = \"false\";\n            java.longToast(\"\\n❌段评已关闭\");\n        } else {\n            paraSettings[\"段评开关\"] = \"true\";\n            java.longToast(\"\\n✅段评已开启\");\n        }\n    }\n    if (v == \"小说\" || v == \"听书\" || v == \"漫画\" || v == \"短剧\") {\n        moreSettings[\"搜索模式\"] = v;\n        java.longToast(`\\n已设置搜索模式为 ${v}`);\n    }\n    if (v == \"图片\") {\n        if (moreSettings[\"显示图片\"] != \"false\") {\n            moreSettings[\"显示图片\"] = \"false\";\n            java.longToast(\"\\n❌已关闭图片显示\");\n        } else {\n            moreSettings[\"显示图片\"] = \"true\";\n            java.longToast(\"\\n✅已开启图片显示\");\n        }\n    }\n    if (v == \"简介\") {\n        if (moreSettings[\"完整简介\"] != \"false\") {\n            moreSettings[\"完整简介\"] = \"false\";\n            java.longToast(\"\\n❌已关闭完整简介\");\n        } else {\n            moreSettings[\"完整简介\"] = \"true\";\n            java.longToast(\"\\n✅已开启完整简介\");\n        }\n    }\n    this.setVariable(\"更多设置\", moreSettings, false);\n    this.setVariable(\"段评设置\", paraSettings, false);\n}\n\n// 按钮切换时改变源变量并刷新发现\nfunction show(m, t) {\n    const { java, source } = this;\n    let data = {};\n    try {\n        data = JSON.parse(source.getVariable())\n    } catch {}\n    data[t] = m;\n    if (t == \"发现页类型\") {\n        data[\"频道\"] = '男频';\n    }\n    source.setVariable(JSON.stringify(data, null, 2));\n    java.refreshExplore();\n}\n\nfunction createFilter(title, chars, defaultVal, paramKey, size, viewName) {\n    if (viewName == undefined) viewName = \"'请选择类型'\";\n    return {\n        title: title,\n        type: \"select\",\n        chars: chars,\n        default: defaultVal,\n        \"viewName\": viewName,\n        action: `show(infoMap['${title}'],'${paramKey}')`,\n        style: {\n            layout_flexGrow: 1,\n            layout_flexBasisPercent: size\n        }\n    };\n}\n\nfunction createText(title, action, defaultValue, size, viewName) {\n    return {\n      title: title,\n      type: 'text',\n      default: defaultValue || '',\n      action: action,\n      viewName: viewName || title,\n      style: {\n        layout_flexGrow: 1,\n        layout_flexBasisPercent: size\n      }\n    };\n}\n\nfunction createButton(title, action, size) {\n    return {\n      title: title,\n      type: 'butLine truncated
    "loginUi": "[\n    {\n        \"name\": \"⚙️ 书源设置\",\n        \"type\": \"button\",\n        \"action\": \"getHtmlSettings()\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.4 }\n    }, {\n        \"name\": \"🗨 段评设置\",\n        \"type\": \"button\",\n        \"action\": \"getSvgSettings()\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.4 }\n    }, {\n        \"name\": \"===下方为备用设置方案===\",\n        \"type\": \"button\",\n        \"action\": \"\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 1 }\n    }, {\n        \"name\": \"📖搜索小说\",\n        \"type\": \"button\",\n        \"action\": \"ste('小说')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.4 }\n    }, {\n        \"name\": \"🎧搜索听书\",\n        \"type\": \"button\",\n        \"action\": \"ste('听书')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.4 }\n    }, {\n        \"name\": \"🎥搜索短剧\",\n        \"type\": \"button\",\n        \"action\": \"ste('短剧')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.4 }\n    }, {\n        \"name\": \"🎨搜索漫画\",\n        \"type\": \"button\",\n        \"action\": \"ste('漫画')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.4 }\n    }, {\n        \"name\": \"段评开关\",\n        \"type\": \"button\",\n        \"action\": \"ste('段评')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.25 }\n    }, {\n        \"name\": \"图片开关\",\n        \"type\": \"button\",\n        \"action\": \"ste('图片')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.25 }\n    }, {\n        \"name\": \"简介开关\",\n        \"type\": \"button\",\n        \"action\": \"ste('简介')\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 0.25 }\n    }, {\n        \"name\": \"⛔‼️ 清空还原设置\",\n        \"type\": \"button\",\n        \"action\": \"deleteVariable()\",\n        \"style\": { \"layout_flexGrow\": 1, \"layout_flexBasisPercent\": 1 }\n    }\n]",
    "searchUrl": "<js>\nlet moreSettings = getVariable('更多设置');\nlet tab = moreSettings && moreSettings['搜索模式'] || '小说';\nvar k = String(key);\nlet m = k.match(/^[xtmdXTMD][:：]\\s*/);\nif (m) {\n    tab = { x: '小说', t: '听书', m: '漫画', d: '短剧' }[m[0][0].toLowerCase()];\n    k = k.slice(m[0].length);\n}\nlet tabType = { 小说: 3, 听书: 2, 漫画: 8, 短剧: 11 }[tab] || 3;\n`/search?query=${encodeURIComponent(k)}&offset=${(page - 1) * 10}&tab_type=${tabType}`;\n</js>",
    "exploreUrl": "<js>\nvar base = String(source.getKey()).replace(/\\/$/, '');\nvar tab = getVariable('发现页类型') || '小说';\nvar channel = getVariable('频道') || '男频';\nvar wordNumMap = { '不限': '0', '30万字以内': '4', '100万字以内': '1', '30万-100万': '5', '100万-300万': '2', '300万字以上': '3', '100万字以上': '9' };\nvar updateMap = { '不限': '2', '完结': '0', '连载中': '1', '今日更新': '3' };\nvar sortMap = { '综合': '0', '热门': '24', '最新': '76', '评分': '12' };\nvar wordNum = getVariable('字数') || '不限';\nvar update = getVariable('更新') || '不限';\nvar sort = getVariable('排序') || '综合';\nvar q = `&word_number=${wordNumMap[wordNum] || '0'}&book_status=${updateMap[update] || '2'}&sort_by=${sortMap[sort] || '0'}`;\nvar tabCode = { 小说: 'n', 听书: 't', 漫画: 'm', 短剧: 'd' }[tab] || 'n';\n\nvar list = [];\nlist.push(createFilter('类型', ['小说', '听书', '短剧', '漫画'], tab, '发现页类型', 0.4));\nlist.push(createFilter('频道', ['男频', '女频'], channel, '频道', 0.4));\nlist.push(createText('搜索关键词', \"setVariable('搜索关键词',(infoMap['搜索关键词'] || (infoMap.get && infoMap.get('搜索关键词'))))\", getVariable('搜索关键词') || '', 0.5, \"'搜索书名或作者'\"));\nlist.push(createButton('🔍搜索', 'exploreSearch()', 0.3));\nlist.push({ title: '⚙️书源设置', type: 'button', action: 'getHtmlSettings()', style: { layout_flexGrow: 1, layout_flexBasisPercent: 0.25 } });\nlist.push({ title: '🗨段评设置', type: 'button', action: 'getSvgSettings()', style: { layout_flexGrow: 1, layout_flexBasisPercent: 0.25 } });\nlist.push(createFilter('字数', Object.keys(wordNumMap), wordNum, '字数', 0.4));\nlist.push(createFilter('更新', Object.keys(updateMap), update, '更新', 0.4));\nlist.push(createFilter('排序', Object.keys(sortMap), sort, '排序', 0.1));\n\nif (tab == '短剧') {\n    list.push({ title: '短剧暂无分类，请使用上方搜索 🔍', url: '', style: { layout_flexGrow: 1, layout_flexBasisPercent: 1 } });\n} else {\n    var ctab = tab == '小说' ? (channel == '女频' ? 0 : 1) : (tab == '听书' ? 3 : 5);\n    try {\n        var res = JSON.parse(java.ajax(base + '/category-front?tab=' + ctab));\n        var cells = ((res.data || {}).category_tab_data || {}).cell_data || [];\n        cells.forEach(function (cell) {\n            (cell.atom_data || []).forEach(function (a) {\n                var c = a.category_data;\n                if (c && c.category_id != null) {\n                    list.push({\n                        title: c.name,\n                        url: base + '/category-landing?category_id=' + c.category_id + '&offset={{(page-1)*20}}' + q + '&fqtab=' + tabCode,\n                        style: { layout_flexGrow: 1, layout_flexBasisPercent: 0.25 }\n                    });\n                }\n            });\n        });\n    } catch (e) {\n        java.toast('发现页分类获取失败');\n    }\n}\nJSON.stringify(list);\n</js>",
    "ruleSearch": {
-      "author": "author",
+      "bookList": "<js>\nvar d = JSON.parse(result);\nvar isVideo = /tab_type=11/.test(baseUrl);\nvar arr = [];\n(d.data || []).forEach(function (b) {\n    var list = isVideo ? (b.video_data || []) : (b.book_data || []);\n    list.forEach(function (x) {\n        if (isVideo) {\n            arr.push({\n                book_id: x.series_id,\n                book_name: x.title,\n                author: '',\n                abstract: x.rec_text || '',\n                thumb_url: x.horiz_cover || x.cover,\n                category: '短剧',\n                creation_status: '1',\n                score: x.score,\n                word_number: String(x.episode_cnt || '')\n            });\n        } else {\n            arr.push(x);\n        }\n    });\n});\narr;\n</js>",
-      "bookList": "<js>\nvar tab = java.get('fq_search_tab') || getVariable('模式');\nvar d = parseJsonSafely(result, {});\nvar list = [];\n(d.data || []).forEach(function (b) {\n    if (b.book_data) {\n        b.book_data.forEach(function (x) {\n            x.fq_tab = tab;\n            list.push(x);\n        });\n    } else if (b.video_data) {\n        // 短剧字段与书籍不同，归一化为 book_data 结构（book_id 用 series_id 走通 detail/catalog）\n        b.video_data.forEach(function (v) {\n            list.push({\n                book_id: v.series_id || v.vid,\n                book_name: v.title || v.raw_book_name,\n                author: (v.video_detail && v.video_detail.actor) || '',\n                thumb_url: v.cover || '',\n                abstract: v.sub_title || '',\n                category: v.sub_title || '',\n                fq_tab: tab\n            });\n        });\n    }\n});\nJSON.stringify({data: list});\n</js>$.data",
+      "name": "$.book_name",
-      "bookUrl": "<js>\n'/detail?book_id=' + result.book_id + '&tab=' + (result.fq_tab || getVariable('模式'));\n</js>",
+      "author": "$.author",
-      "checkKeyWord": "十日终焉",
+      "kind": "{{$.creation_status=='0'?'完结':'连载'}},{{$.category}},{{$.score}}分",
-      "coverUrl": "thumb_url",
+      "wordCount": "$.word_number##^0+$",
-      "intro": "abstract",
+      "intro": "$.abstract",
-      "kind": "category",
+      "coverUrl": "$.thumb_url",
-      "name": "book_name",
+      "bookUrl": "<js>\nvar t = /tab_type=11/.test(baseUrl) ? 'd' : /tab_type=2/.test(baseUrl) ? 't' : /tab_type=8/.test(baseUrl) ? 'm' : 'n';\n`/detail?book_id=${result.book_id}&fqtab=${t}`;\n</js>",
-      "wordCount": "word_number"
+      "checkKeyWord": "斗罗大陆"
    },
    "ruleExplore": {
      "author": "author",
      "bookList": "$.data.book_info[*]",
-      "bookUrl": "<js>\n'/detail?book_id=' + result.book_id + '&tab=' + (result.fq_tab || getVariable('模式'));\n</js>",
+      "name": "$.book_name",
-      "coverUrl": "thumb_url",
+      "author": "$.author",
-      "intro": "abstract",
+      "kind": "{{$.category}},{{$.score}}分",
-      "kind": "category",
+      "wordCount": "$.word_number##^0+$",
-      "name": "book_name",
+      "intro": "$.abstract",
-      "wordCount": "word_number"
+      "coverUrl": "$.thumb_url",
      "bookUrl": "<js>\nvar m = String(baseUrl).match(/fqtab=([ntmd])/);\n`/detail?book_id=${result.book_id}&fqtab=${m ? m[1] : 'n'}`;\n</js>"
    },
    "ruleBookInfo": {
-      "author": "$.data.author",
+      "init": "$.data",
-      "canReName": "1",
+      "name": "$.book_name",
-      "coverUrl": "$.data.thumb_url",
+      "author": "$.author",
-      "init": "<js>\njava.put('fq_tab', (String(baseUrl).match(/[?&]tab=([^&]+)/) || [])[1] || getVariable('模式'));\nresult;\n</js>",
+      "kind": "{{$.creation_status=='0'?'完结':'连载'}},{{$.category}}",
-      "intro": "<js>\nvar tab = java.get('fq_tab') || getVariable('模式');\nvar x = (parseJsonSafely(result, {}) || {}).data || {};\nfunction v(t) { return (t != undefined && t != null && String(t).length > 0 && String(t) != 'null') ? String(t) : ''; }\nfunction cats() {\n    if (x.category) return x.category;\n    if (Array.isArray(x.category_v2)) return x.category_v2.map(function (c) { return c.Name || c.name || ''; }).filter(String).join('/');\n    return '';\n}\nvar statusMap = {'0': '连载中', '1': '连载中', '2': '已完结', 'creation_status_end': '已完结', 'creation_status_loading': '连载中'};\nvar st = statusMap[v(x.creation_status)] || v(x.creation_status);\nvar lines = ['📡 来源：番茄  |  🔄 模式：' + tab];\n[['🌈 分类', cats()],\n ['📚 最新章节', v(x.last_chapter_title)],\n ['📝 字数', v(x.word_number)],\n ['🚩 状态', st],\n ['⭐ 评分', v(x.score)],\n ['👀 阅读人数', v(x.read_count)]\n].forEach(function (p) { if (p[1]) lines.push('    ' + p[0] + '：' + p[1]); });\nif (v(x.abstract)) lines.push('❇️─────────\\n📖 简介：\\n' + String(x.abstract).split('\\n').map(function (l) { return '    ' + l; }).join('\\n'));\nlines.join('\\n');\n</js>",
+      "wordCount": "$.word_number##^0+$",
-      "kind": "$.data.category_v2[*].Name",
+      "lastChapter": "$.last_chapter_title",
-      "lastChapter": "$.data.last_chapter_title",
+      "coverUrl": "$.thumb_url",
-      "name": "$.data.book_name",
+      "intro": "<js>\nconst { book_id, last_chapter_title, last_chapter_update_time, word_number, score, abstract, read_count, category } = result;\nlet m0 = String(baseUrl).match(/fqtab=([ntmd])/);\nconst tab = { n: '小说', t: '听书', m: '漫画', d: '短剧' }[(m0 ? m0[1] : 'n')];\nlet moreSettings = getVariable('更多设置');\nconst lightDivider = \"❇️───────❇️───────❇️\";\nconst heavyDivider = \"&lrm;\\n&lrm;\";\nconst isValid = (v) => String(v).length > 1;\nlet ctitle = book.durChapterTitle || '未开始';\nconst two = (n) => ('0' + n).slice(-2);\nconst fmtTime = (t) => {\n    let n = parseInt(t);\n    if (!n || n <= 0) return '';\n    let d = new Date(n * 1000);\n    return `${d.getFullYear()}-${two(d.getMonth() + 1)}-${two(d.getDate())}`;\n};\nconst formatAbstract = (text) => String(text).split(\"\\n\").map(line => `    ${line}`).join(\"\\n\");\nconst buildBasicInfo = (items) => {\n    let info = \"\";\n    for (let [value, prefix, icon] of items) {\n        if (isValid(value)) info += `    ${icon} ${prefix} ${value}\\n`;\n    }\n    return info;\n};\nlet info = `    🏷 数据来源：番茄小说(fanqie-go)\\n    🔄 当前模式：${tab}\\n    📖 阅读至：${ctitle}\\n`;\nconst fullBasicInfo = buildBasicInfo([\n    [category, \"书籍分类：\", \"🌈\"],\n    [last_chapter_title, \"最新章节：\", \"📚\"],\n    [fmtTime(last_chapter_update_time), \"更新时间：\", \"⏳\"],\n    [word_number, \"书籍字数：\", \"📝\"],\n    [score, \"书籍评分：\", \"⭐\"],\n    [read_count, \"今阅读人数：\", \"📃\"]\n]);\nif (fullBasicInfo) info += `${lightDivider}\\n${fullBasicInfo}`;\nif (isValid(abstract)) info += `${heavyDivider}\\n    📖 书籍简介：\\n${formatAbstract(abstract)}\\n`;\ninfo += `${lightDivider}\\n${heavyDivider}\\n    数据更新于 ${new Date().toLocaleString()}\\n`;\nconst abstractMode = moreSettings && moreSettings['完整简介'] || 'true';\nif (abstractMode == \"false\") {\n    const simplifiedBasicInfo = buildBasicInfo([\n        [last_chapter_title, \"最新章节：\", \"📚\"],\n        [fmtTime(last_chapter_update_time), \"更新时间：\", \"⏳\"],\n        [word_number, \"书籍字数：\", \"📝\"],\n        [score, \"书籍评分：\", \"⭐\"]\n    ]);\n    info = `\\n&lrm;${isValid(abstract) ? `📖 书籍简介：\\n${formatAbstract(abstract)}\\n${heavyDivider}` : ''}🏷 数据来源：番茄小说\\n${simplifiedBasicInfo}`;\n} else {\n    info = info.split(\"\\n\").map(line => line.replace(/^ {4}/, \"\")).join(\"\\n\");\n}\ninfo;\n</js>",
-      "tocUrl": "/catalog?book_id={{$.data.book_id}}",
+      "tocUrl": "<js>\nvar m = String(baseUrl).match(/fqtab=([ntmd])/);\n`/catalog?book_id=${result.book_id}&fqtab=${m ? m[1] : 'n'}`;\n</js>",
-      "wordCount": "$.data.word_number"
+      "canReName": "1"
    },
    "ruleToc": {
-      "chapterList": "<js>\njava.put('fq_book_id', (String(baseUrl).match(/book_id=(\\d+)/) || [])[1] || '');\n// 书籍类型按模式动态设置：安卓 8文本/32音频/64图片，iOS 0/1/2\nvar tab = java.get('fq_tab') || getVariable('模式');\nvar tMap = {'小说': 8, '听书': 32, '漫画': 64, '短剧': 8};\nvar t = tMap[tab] || 8;\nif (isIOS()) t = {'小说': 0, '听书': 1, '漫画': 2, '短剧': 0}[tab] || 0;\ntry { book.type = t; } catch (e) {}\nresult;\n</js>$.data.item_data_list[*]",
+      "chapterList": "<js>\n// 类型随书切换（照光遇 typeMap：安卓用 default，苹果端用其类型体系）\nconst typeMap = {\n    \"小说\": { \"苹果\": 0, \"default\": 8 },\n    \"听书\": { \"苹果\": 1, \"default\": 32 },\n    \"漫画\": { \"苹果\": 2, \"default\": 64 },\n    \"短剧\": { \"苹果\": 3, \"default\": 8 }\n};\nconst device_type = checkEnv();\nvar tabName = { n: '小说', t: '听书', m: '漫画', d: '短剧' }[fqTab(baseUrl, 'n')];\nvar currentType = typeMap[tabName] || typeMap[\"小说\"];\nif (device_type == '苹果') {\n    book.type = currentType[device_type];\n} else {\n    book.type = currentType.default;\n}\ntry {\n    book.imageStyle = checkEnv() == '改版' ? 'FULL' : 'TEXT';\n} catch (e) {}\nresult;\n</js>$.data.item_data_list[*]",
      "chapterName": "$.title",
-      "chapterUrl": "<js>\nvar tab = java.get('fq_tab') || getVariable('模式');\nvar id = result.item_id;\nvar url;\nif (tab == '听书') {\n    url = '/audio?item_ids=' + id + '&tone_id=' + getVariable('音色') + '&output=url';\n} else if (tab == '短剧') {\n    url = '/video?item_ids=' + id + '&output=proxy&quality=' + getVariable('清晰度');\n} else {\n    var review = getVariable('段评') != '关' ? 'keep' : 'remove';\n    var img = getVariable('图片') == '移除' ? 'remove' : 'keep';\n    url = '/content?item_id=' + id + '&text_mode=' + (review == 'keep' ? 'html' : 'plain') + '&review=' + review + '&image_mode=' + img;\n}\nurl;\n</js>"
+      "chapterUrl": "<js>\nvar t = fqTab(baseUrl, 'n');\nvar id = result.item_id;\nvar u = {\n    n: `/content?item_id=${id}&text_mode=html&review=keep&image_mode=keep&fqtab=n`,\n    t: `/audio?item_ids=${id}&fqtab=t`,\n    m: `/manga?item_ids=${id}&output=proxy&fqtab=m`,\n    d: `/video?item_ids=${id}&fqtab=d`\n}[t];\nu;\n</js>"
    },
    "ruleContent": {
-      "content": "<js>\nvar base = BaseUrl();\nvar u = String(baseUrl);\nvar item = (u.match(/item_ids?=(\\d+)/) || [])[1] || '';\nvar tab = java.get('fq_tab') || getVariable('模式');\nfunction absUrl(s) { return (s && String(s).indexOf('http') != 0) ? base + s : s; }\nfunction videoUrl() {\n    var v = parseJsonSafely(request('/video?item_ids=' + item + '&output=proxy&quality=' + getVariable('清晰度')), null);\n    var vl = v && v.data && v.data.video_list;\n    if (vl) { var ks = Object.keys(vl); if (ks.length) return absUrl(vl[ks[0]].main_url); }\n    return '';\n}\nvar out = '';\nif (u.indexOf('/video') > -1) {\n    // 短剧：解密代理直链，当前章节自动打开播放\n    var d = parseJsonSafely(result, {});\n    var vl = d.data && d.data.video_list;\n    var vu = '';\n    if (vl) { var ks = Object.keys(vl); if (ks.length) vu = absUrl(vl[ks[0]].main_url); }\n    if (vu && book.durChapterIndex == chapter.index) {\n        try { java.startBrowser(vu, title || ''); } catch (e) {}\n    }\n    out = vu ? `▶️ 正在加载视频…\\n若未自动播放，请复制以下直链到播放器打开：\\n${vu}` : '❌ 视频获取失败，请刷新重试';\n} else if (u.indexOf('/audio') > -1) {\n    // 听书：短剧视频章优先原声音轨，其余回退 TTS 音色\n    out = videoUrl();\n    if (!out) {\n        var a = parseJsonSafely(result, {});\n        if (a.data && a.data[0] && a.data[0].main_url) out = absUrl(a.data[0].main_url);\n    }\n} else {\n    // 小说/漫画：段评气泡 + 空正文回退图片列表\n    var review = getVariable('段评') != '关';\n    var c = '';\n    var d = parseJsonSafely(result, {});\n    c = (d.data && d.data.content) || '';\n    if (c && review) {\n        // 段评 ident 域名由服务端按请求 Host 拼接，强制替换为书源地址\n        c = c.replace(/(<comment ident=\")[^\"]*?\\/review_html\\?/g, '$1' + base + '/review_html?');\n        c = isIOS() ? paraForiOS(c) : paraForAndroid(c);\n    }\n    if (c) {\n        out = c;\n    } else {\n        var m = parseJsonSafely(request('/manga?item_ids=' + item + '&output=proxy'), {});\n        // 兼容两种响应：data 为图片数组，或 data.images 为图片数组\n        var imgs = (m && m.data && Array.isArray(m.data.images)) ? m.data.images : (m && Array.isArray(m.data) ? m.data : null);\n        out = imgs ? imgs.map(function (i) { return `<img src=\"${absUrl(i.data)}\">`; }).join('\\n') : ((m && m.data && m.data.content) || '❌ 内容获取失败，请刷新重试');\n    }\n}\nout;\n</js>"
+      "content": "<js>\nvar t = fqTab(baseUrl, 'n');\nlet moreSettings = getVariable('更多设置');\nlet dpSettings = getVariable('段评设置');\nlet islyc = checkEnv();\n\nif (t == 't') {\n    // 听书：直接返回音频地址\n    JSON.parse(result).data[0].main_url;\n} else if (t == 'm') {\n    // 漫画：拼装图片\n    var d = JSON.parse(result);\n    var org = String(baseUrl).match(/^https?:\\/\\/[^\\/]+/)[0];\n    var imgs = (d.data || {}).images || [];\n    var out = [];\n    for (var i = 0; i < imgs.length; i++) {\n        out.push('<img src=\"' + org + imgs[i].data + '\">');\n    }\n    out.join('\\n');\n} else if (t == 'd') {\n    // 短剧：选最高清晰度，当前章节自动唤起播放\n    var d = JSON.parse(result).data || {};\n    var vl = d.video_list || {};\n    var list = [];\n    for (var k in vl) { list.push(vl[k]); }\n    list.sort(function (a, b) { return (b.vheight || 0) - (a.vheight || 0); });\n    var org = String(baseUrl).match(/^https?:\\/\\/[^\\/]+/)[0];\n    var url = list.length ? String(list[0].main_url) : '';\n    if (url && url.indexOf('http') != 0) { url = org + url; }\n    if (url && book.durChapterIndex == chapter.index) {\n        java.startBrowser(url, chapter.title);\n    }\n    '▶️ 本集已唤起播放（菜单-刷新可重新打开）：\\n' + url;\n} else {\n    // 小说：正文 + 段评气泡\n    var d = JSON.parse(result);\n    var content = String((d.data || {}).content || '');\n    let show_img = moreSettings && moreSettings['显示图片'] || 'true';\n    if (show_img == 'false') {\n        content = removeAllImgTags(content);\n    }\n    let para = dpSettings && dpSettings['段评开关'] || 'true';\n    if (para == 'true' && content.indexOf('<comment') != -1) {\n        content = islyc == '苹果' ? paraForiOS(content, '番茄') : paraForAndroid(content, '番茄');\n    }\n    content;\n}\n</js>",
-    },
+      "title": "$.data.title"
-    "weight": 0
+    }
  }
 ]
@@ -76,6 +76,7 @@ func main() {
 		logger.Error("设备密钥池初始化失败: " + err.Error())
 		os.Exit(1)
 	}
 	defer fqapi.FlushDevicePool() // 程序退出时落盘设备使用记录
 	// 设置路由
 	setupRoutes()
@@ -164,6 +165,9 @@ func setupRoutes() {
 	http.HandleFunc("/catalog", wrap(handlers.HandleCatalog))
 	http.HandleFunc("/directory/all_items", wrap(handlers.HandleCatalog))
 	http.HandleFunc("/multi-detail", wrap(handlers.HandleMultiDetail))
 	http.HandleFunc("/rank", wrap(handlers.HandleRank))
 	http.HandleFunc("/bookshelf", wrap(handlers.HandleBookshelf))
 	http.HandleFunc("/read-history", wrap(handlers.HandleReadHistory))
 	http.HandleFunc("/multi-content", wrap(handlers.HandleMultiContent))
 	http.HandleFunc("/content", wrap(handlers.HandleContent))
 	http.HandleFunc("/item-summary", wrap(handlers.HandleItemSummary))
@@ -7,6 +7,11 @@
 #   ./deploy.sh v1.2.0                # 编译 + 部署（指定版本）
 #   ./deploy.sh v1.2.0 --init         # 首次初始化（写 systemd/nginx 配置）
 #   ./deploy.sh v1.2.0 --no-pull      # 跳过 git pull
 #   ./deploy.sh v1.2.0 --no-autostart # 部署后不开机自启
 #   ./deploy.sh v1.2.0 --no-restart   # 只部署，不重启服务
 #   ./deploy.sh --stop                # 停止服务
 #   ./deploy.sh --start               # 启动服务
 #   ./deploy.sh --clean               # 清除所有部署内容（服务/nginx/证书），重新 --init 前用
 #   ./deploy.sh --build-only          # 只编译，不部署
 #   ./deploy.sh --rollback            # 交互式回滚到指定版本
 #   ./deploy.sh --rollback v1.0.0     # 直接回滚到指定版本
@@ -36,9 +41,11 @@ NGINX_CONF="${NGINX_CONF:-/etc/nginx/conf.d/${APP_NAME}.conf}"
 APP_PORT="${APP_PORT:-40180}"              # 程序监听端口
 DOMAIN="${DOMAIN:-fq.20010117.xyz}"        # 域名
 CERT_EMAIL="${CERT_EMAIL:-admin@${DOMAIN}}"  # 证书申请邮箱（过期提醒）
 CERT_DIR="${CERT_DIR:-/etc/nginx/ssl}"       # 证书目录（acme.sh --install-cert 安装位置）
 BUILD_TARGET="${BUILD_TARGET:-./cmd/fanqie}"  # go build 的包路径
 ENV_FILE="${ENV_FILE:-${SRC_DIR}/.env}"    # 应用环境变量（.env.example 为模板）
 KEEP_VERSIONS="${KEEP_VERSIONS:-5}"        # 保留最近几个版本
 AUTO_START="${AUTO_START:-true}"           # 服务是否开机自启（--no-autostart 覆盖）
 # ==================== 颜色与日志 ====================
 C_GREEN='\033[1;32m'; C_YELLOW='\033[1;33m'; C_RED='\033[1;31m'; C_BLUE='\033[1;34m'; C_RESET='\033[0m'
@@ -63,6 +70,11 @@ ROLLBACK_VERSION=""
 LIST=false
 FORCE=false
 SYNC_NGINX=false
 AUTO_START="${AUTO_START:-true}"
 NO_RESTART=false
 STOP=false
 START=false
 CLEAN=false
 print_help() {
  cat <<EOF
@@ -71,6 +83,12 @@ print_help() {
 选项:
  --init                首次初始化，写入 systemd 和 nginx 配置
  --no-pull             跳过 git pull
  --no-autostart        服务不设置开机自启（默认自启）
  --no-restart          只部署不重启服务（自启设置仍会应用）
  --stop                停止服务后退出
  --start               启动服务后退出
  --clean               清除本脚本部署的全部内容（服务/nginx/二进制/证书，含 certbot 残留；
                        保留 SRC_DIR 下的 .env/配置/数据库），之后重新 --init 即可
  --build-only          只编译，不部署、不重启服务
  --force, -f           允许覆盖已存在的同名版本（默认报错退出）
  --rollback [版本]     回滚到指定版本（不带版本号则交互式选择）
@@ -96,6 +114,11 @@ while [ $i -lt ${#ARGS[@]} ]; do
  case "$arg" in
    --init)        INIT=true ;;
    --no-pull)     NO_PULL=true ;;
    --no-autostart) AUTO_START=false ;;
    --no-restart)  NO_RESTART=true ;;
    --stop)        STOP=true ;;
    --start)       START=true ;;
    --clean)       CLEAN=true ;;
    --build-only)  BUILD_ONLY=true ;;
    --rollback)
      ROLLBACK=true
@@ -171,28 +194,72 @@ cleanup_old_versions() {
  done
 }
-# 重启服务 + 验证
+# 定位 acme.sh：PATH → 当前用户目录 → root 目录（sudo 后 HOME 可能被重置）
-restart_and_verify() {
+find_acme_sh() {
-  log "重启 ${APP_NAME} 服务..."
+  local acme
-  # 注意：不要用 "list-unit-files | grep -q"——grep -q 提前退出会让 systemctl 收到
+  acme="$(command -v acme.sh 2>/dev/null || true)"
-  # SIGPIPE，在 pipefail 下整条管道判假（unit 多的机器必踩），误报"服务不存在"
+  for c in "${HOME}/.acme.sh/acme.sh" "/root/.acme.sh/acme.sh"; do
-  if systemctl cat "${APP_NAME}" >/dev/null 2>&1; then
+    if [ -z "$acme" ] && [ -x "$c" ]; then acme="$c"; fi
-    $SUDO systemctl restart "${APP_NAME}"
+  done
  echo "$acme"
 }
 # 删除已存在的路径并逐项记录（文件/目录/glob 展开结果均可，不存在则跳过）
 rm_logged() {
  local p
  for p in "$@"; do
    if [ -e "$p" ] || [ -L "$p" ]; then
      $SUDO rm -rf "$p"
      log "  已删除: $p"
    fi
  done
 }
 # 应用开机自启设置（AUTO_START 由 --no-autostart / .deploy.env 决定）
 apply_autostart() {
  if [ "$AUTO_START" = true ]; then
    $SUDO systemctl enable "${APP_NAME}" >/dev/null 2>&1 || true
  else
    $SUDO systemctl disable "${APP_NAME}" >/dev/null 2>&1 || true
    info "开机自启已禁用（--no-autostart）"
  fi
 }
 # 重启服务 + 验证
 restart_and_verify() {
  # 注意：不要用 "list-unit-files | grep -q"——grep -q 提前退出会让 systemctl 收到
  # SIGPIPE，在 pipefail 下整条管道判假（unit 多的机器必踩），误报"服务不存在"
  if ! systemctl cat "${APP_NAME}" >/dev/null 2>&1; then
    warn "服务 ${APP_NAME}.service 不存在，请先执行 $0 --init"
    return 0
  fi
  apply_autostart
  if [ "$NO_RESTART" = true ]; then
    info "跳过服务重启（--no-restart），需要时执行: sudo systemctl restart ${APP_NAME}"
    return 0
  fi
  log "重启 ${APP_NAME} 服务..."
  $SUDO systemctl restart "${APP_NAME}"
  if command -v nginx >/dev/null 2>&1; then
    $SUDO systemctl reload nginx 2>/dev/null || $SUDO nginx -s reload 2>/dev/null || true
  fi
  sleep 2
  # 健康自检：确认服务真正响应，而非仅进程存活
  # 应用启动时初始化设备池（逐个注册设备）较慢，轮询等待就绪，最多 60s
  if command -v curl >/dev/null 2>&1; then
-    if curl -fsS --max-time 5 "http://127.0.0.1:${APP_PORT}/healthz" >/dev/null 2>&1; then
+    healthy=false
    for _ in $(seq 1 30); do
      if curl -fsS --max-time 2 "http://127.0.0.1:${APP_PORT}/healthz" >/dev/null 2>&1; then
        healthy=true
        break
      fi
      sleep 2
    done
    if [ "$healthy" = true ]; then
      log "健康检查通过: /healthz"
    else
      err "健康检查失败: http://127.0.0.1:${APP_PORT}/healthz"
@@ -212,8 +279,8 @@ restart_and_verify() {
 # 写入 nginx 配置（--init 与 --sync-nginx 共用的唯一模板来源）
-# 模板自带完整 443/证书配置（证书存在时启用），不依赖 certbot --nginx 注入——
+# 模板自带完整 443/证书配置（证书存在时启用），不依赖签发工具注入——
-# 这样 --sync-nginx 覆盖配置时不会丢失 HTTPS 段（certbot 注入式配置会被覆盖回 80-only）
+# 这样 --sync-nginx 覆盖配置时不会丢失 HTTPS 段（注入式配置会被覆盖回 80-only）
 write_nginx_conf() {
  # 反代 location（80 无证书时与 443 共用）
  local proxy_locations
@@ -248,7 +315,7 @@ LOC_EOF
 )
  local ssl_block=""
-  if [ -f "/etc/letsencrypt/live/${DOMAIN}/fullchain.pem" ]; then
+  if [ -f "${CERT_DIR}/${DOMAIN}/fullchain.pem" ]; then
    log "检测到 ${DOMAIN} 证书，配置启用 443"
    ssl_block=$(cat <<SSL_EOF
@@ -256,8 +323,8 @@ server {
    listen 443 ssl http2;
    server_name ${DOMAIN};
-    ssl_certificate     /etc/letsencrypt/live/${DOMAIN}/fullchain.pem;
+    ssl_certificate     ${CERT_DIR}/${DOMAIN}/fullchain.pem;
-    ssl_certificate_key /etc/letsencrypt/live/${DOMAIN}/privkey.pem;
+    ssl_certificate_key ${CERT_DIR}/${DOMAIN}/privkey.pem;
    ssl_protocols TLSv1.2 TLSv1.3;
    access_log /var/log/nginx/${APP_NAME}.access.log;
@@ -267,7 +334,7 @@ ${proxy_locations}
 SSL_EOF
 )
  else
-    warn "未找到 /etc/letsencrypt/live/${DOMAIN}/ 证书，本次仅写 80 配置（不跳转 HTTPS）；证书签发后重跑: $0 --sync-nginx"
+    warn "未找到 ${CERT_DIR}/${DOMAIN}/ 证书，本次仅写 80 配置（不跳转 HTTPS）；证书签发后重跑: $0 --sync-nginx"
  fi
  # 80 端口：有证书时仅 ACME 验证 + 301 跳转；无证书时直接反代
@@ -288,7 +355,7 @@ server {
    access_log /var/log/nginx/${APP_NAME}.access.log;
    error_log  /var/log/nginx/${APP_NAME}.error.log;
-    # ACME http-01 验证（certbot webroot 模式依赖）
+    # ACME http-01 验证（acme.sh webroot 模式依赖）
    location /.well-known/acme-challenge/ {
        root /var/www/html;
    }
@@ -296,6 +363,12 @@ server {
 ${root_location}
 }${ssl_block}
 NGINX_EOF
  # 写入后立即校验并重载：ACME 验证必须打到最新配置，
  # 否则签发时线上仍是内存里的旧配置（80→301→443→反代→404），验证必败
  if $SUDO nginx -t 2>/dev/null; then
    $SUDO systemctl reload nginx 2>/dev/null || $SUDO nginx -s reload 2>/dev/null || true
  fi
 }
 # 更新已部署服务器的 nginx 配置：备份 → 重写 → 校验 → 失败回滚
@@ -342,6 +415,84 @@ if [ "$SYNC_NGINX" = true ]; then
  exit 0
 fi
 # ==================== --stop / --start ====================
 if [ "$STOP" = true ] || [ "$START" = true ]; then
  systemctl cat "${APP_NAME}" >/dev/null 2>&1 || { err "服务 ${APP_NAME}.service 不存在，请先执行 $0 --init"; exit 1; }
  if [ "$STOP" = true ]; then
    $SUDO systemctl stop "${APP_NAME}"
    log "✅ 已停止 ${APP_NAME}（开机自启设置不变）"
  else
    $SUDO systemctl start "${APP_NAME}"
    sleep 1
    $SUDO systemctl status "${APP_NAME}" --no-pager -l | head -n 15 || true
  fi
  exit 0
 fi
 # ==================== --clean ====================
 # 清除本脚本部署的全部内容，便于重新 --init 初次部署
 if [ "$CLEAN" = true ]; then
  warn "将删除以下内容："
  echo "  - systemd 服务: ${SERVICE_FILE}"
  echo "  - nginx 配置:   ${NGINX_CONF}（含 --sync-nginx 的 .bak.* 备份）"
  echo "  - 二进制目录:   ${BIN_DIR}"
  echo "  - 证书:         ${CERT_DIR}/${DOMAIN}、acme.sh 中该域名的全部数据、certbot 残留"
  echo "  （保留 ${SRC_DIR} 下的 .env / 配置 / 数据库；acme.sh 本体保留供重新部署使用）"
  read -rp "确认清理？输入 yes 继续: " REPLY
  [ "$REPLY" = "yes" ] || { info "已取消"; exit 0; }
  echo ""
  log "开始清理，实际删除项："
  # 服务
  if systemctl cat "${APP_NAME}" >/dev/null 2>&1; then
    $SUDO systemctl disable --now "${APP_NAME}" || true
    $SUDO rm -f "${SERVICE_FILE}"
    $SUDO systemctl daemon-reload
    log "  已删除: ${SERVICE_FILE}（服务已停止并取消自启）"
  fi
  # nginx 配置（含 --sync-nginx 产生的 .bak.* 备份）
  if [ -f "${NGINX_CONF}" ] || ls "${NGINX_CONF}".bak.* >/dev/null 2>&1; then
    rm_logged "${NGINX_CONF}" "${NGINX_CONF}".bak.*
    if $SUDO nginx -t 2>/dev/null; then
      $SUDO systemctl reload nginx 2>/dev/null || $SUDO nginx -s reload 2>/dev/null || true
      log "  nginx 已重载"
    fi
  fi
  # 二进制与版本目录
  rm_logged "${BIN_DIR}"
  # 证书：移除 acme.sh 管理（--remove 不吊销，LE 证书自然过期），删安装目录
  # 及 acme.sh 域名数据（含验证失败残留的 domain key，否则重签会要求 --force）
  ACME_SH="$(find_acme_sh)"
  if [ -n "$ACME_SH" ]; then
    $SUDO "$ACME_SH" --remove -d "${DOMAIN}" >/dev/null 2>&1 || true
  fi
  rm_logged "${CERT_DIR}/${DOMAIN}" \
            "${HOME}/.acme.sh/${DOMAIN}_ecc" "${HOME}/.acme.sh/${DOMAIN}" \
            "/root/.acme.sh/${DOMAIN}_ecc" "/root/.acme.sh/${DOMAIN}"
  # certbot 残留（历史部署用 certbot 时遗留）
  if command -v certbot &>/dev/null; then
    if systemctl is-enabled certbot.timer >/dev/null 2>&1; then
      $SUDO systemctl disable --now certbot.timer
      log "  已停用: certbot.timer"
    fi
    if [ -d "/etc/letsencrypt/live/${DOMAIN}" ]; then
      $SUDO certbot delete --cert-name "${DOMAIN}" --non-interactive >/dev/null 2>&1 || true
    fi
  fi
  rm_logged "/etc/letsencrypt/live/${DOMAIN}" "/etc/letsencrypt/archive/${DOMAIN}" \
            "/etc/letsencrypt/renewal/${DOMAIN}.conf" \
            "/etc/letsencrypt/renewal-hooks/deploy/reload-nginx.sh"
  echo ""
  log "✅ 清理完成。重新部署: $0 --init"
  exit 0
 fi
 # ==================== --rollback ====================
 if [ "$ROLLBACK" = true ]; then
  if [ ! -d "${BIN_DIR}" ]; then
@@ -459,36 +610,28 @@ EOF
  write_nginx_conf
-  # ---- Certbot 续期后自动重载 Nginx（deploy-hook 只在续期成功时执行）----
+  # ---- 首次初始化自动申请 HTTPS 证书（一次性；续期由 acme.sh cron 自动处理）----
-  HOOK_FILE="/etc/letsencrypt/renewal-hooks/deploy/reload-nginx.sh"
+  if [ -f "${CERT_DIR}/${DOMAIN}/fullchain.pem" ]; then
-  if [ ! -f "${HOOK_FILE}" ]; then
+    log "证书已存在，跳过申请"
-    log "创建 Certbot 续期重载 Hook: ${HOOK_FILE}"
+  else
-    $SUDO mkdir -p /etc/letsencrypt/renewal-hooks/deploy/
+    ACME_SH="$(find_acme_sh)"
-    $SUDO tee "${HOOK_FILE}" >/dev/null <<'HOOK'
+    if [ -z "$ACME_SH" ]; then
-#!/bin/bash
+      err "未找到 acme.sh，请先安装: curl https://get.acme.sh | sh -s email=${CERT_EMAIL}"
-if /usr/sbin/nginx -t 2>/dev/null; then
+      exit 1
    systemctl reload nginx
    fi
-HOOK
+    log "未检测到证书，使用 acme.sh 申请: ${DOMAIN}"
-    $SUDO chmod +x "${HOOK_FILE}"
+    $SUDO mkdir -p /var/www/html "${CERT_DIR}/${DOMAIN}"
-  fi
+    # --force：上次验证失败会残留 domain key，重试时 acme.sh 拒绝覆盖要求 --force
-
+    if $SUDO "$ACME_SH" --issue -d "${DOMAIN}" -w /var/www/html --server letsencrypt --keylength ec-256 -m "${CERT_EMAIL}" --force \
-  # ---- 首次初始化自动申请 HTTPS 证书（一次性；续期由 certbot.timer 自动处理）----
+       && $SUDO "$ACME_SH" --install-cert -d "${DOMAIN}" --ecc \
-  if [ ! -d "/etc/letsencrypt/live/${DOMAIN}" ]; then
+            --fullchain-file "${CERT_DIR}/${DOMAIN}/fullchain.pem" \
-    if command -v certbot >/dev/null 2>&1; then
+            --key-file "${CERT_DIR}/${DOMAIN}/privkey.pem" \
-      log "未检测到证书，尝试自动申请: ${DOMAIN}"
+            --reloadcmd "systemctl reload nginx"; then
      $SUDO mkdir -p /var/www/html
      if $SUDO certbot certonly --webroot -w /var/www/html -d "${DOMAIN}" --non-interactive --agree-tos -m "${CERT_EMAIL}"; then
      log "证书申请成功，写入 443 配置..."
      write_nginx_conf
    else
      warn "证书申请失败，请检查域名解析与 80 端口；成功后重跑: $0 --sync-nginx"
    fi
    else
      warn "未安装 certbot，跳过证书申请。可手动执行: dnf install certbot python3-certbot-nginx"
    fi
  else
    log "证书已存在，跳过申请"
  fi
  if command -v getenforce >/dev/null 2>&1 && [ "$(getenforce)" != "Disabled" ]; then
@@ -10,6 +10,7 @@ require (
 	github.com/mattn/go-sqlite3 v1.14.18
 	github.com/redis/go-redis/v9 v9.22.0
 	github.com/spf13/viper v1.18.2
 	golang.org/x/sync v0.17.0
 	golang.org/x/time v0.5.0
 )
@@ -35,7 +36,6 @@ require (
 	go.uber.org/multierr v1.9.0 // indirect
 	golang.org/x/exp v0.0.0-20230905200255-921286631fa9 // indirect
 	golang.org/x/net v0.46.0 // indirect
 	golang.org/x/sync v0.17.0 // indirect
 	golang.org/x/sys v0.37.0 // indirect
 	golang.org/x/text v0.30.0 // indirect
 	gopkg.in/ini.v1 v1.67.0 // indirect
@@ -74,6 +74,7 @@ func Sign4(
 	// 添加 Ladon 和 Argus 签名
 	ladon := &Ladon{}
 	result["x-ladon"] = ladon.Encrypt(timestamp, licenseID, aid)
 	result["x-helios"] = result["x-ladon"] // X-Helios 与 X-Ladon 同值（对齐书源实现）
 	argus := &Argus{}
 	result["x-argus"] = argus.GetSign(
@@ -39,6 +39,8 @@ type RequestLogConfig struct {
 type FanqieConfig struct {
 	Algorithm string        `mapstructure:"algorithm"`
 	Signature string        `mapstructure:"signature"` // 签名方法: sign2 / sign4
 	// 注册设备身份: pad（默认，历史行为）/ phone（对齐手机 App 书源注册 payload）
 	DeviceIdentity string   `mapstructure:"device_identity"`
 	Content   ContentConfig `mapstructure:"content"`
 	Manga     MangaConfig   `mapstructure:"manga"`
 	Video     VideoConfig   `mapstructure:"video"`
@@ -192,6 +194,7 @@ func setDefaults() {
 	// 番茄小说配置
 	viper.SetDefault("fanqie.algorithm", "gorgon")
 	viper.SetDefault("fanqie.signature", "sign2")
 	viper.SetDefault("fanqie.device_identity", "pad")
 	// 内容处理配置
 	viper.SetDefault("fanqie.content.remove_images", false)
@@ -16,8 +16,31 @@ import (
 	_ "github.com/jackc/pgx/v5/stdlib"
 	_ "github.com/mattn/go-sqlite3"
 	"github.com/redis/go-redis/v9"
 	"golang.org/x/sync/singleflight"
 )
 const (
 	// sqlCacheTTLSeconds SQL 持久层保留时长：不可变内容（正文/详情/目录）与业务 TTL 解耦，
 	// Redis 过期或重启后由 SQL 回填，避免重新打上游
 	sqlCacheTTLSeconds = 30 * 86400
 )
 // volatileCachePrefixes 动态短缓存（值会变）只走 Redis/短 TTL，不进 SQL 长期层
 var volatileCachePrefixes = []string{"search_", "category_", "rank_"}
 // isVolatileCacheKey 判断是否为动态短缓存 key
 func isVolatileCacheKey(key string) bool {
 	for _, p := range volatileCachePrefixes {
 		if strings.HasPrefix(key, p) {
 			return true
 		}
 	}
 	return false
 }
 // cacheSF 合并并发 miss 的回源请求，防止缓存击穿
 var cacheSF singleflight.Group
 var (
 	cacheDB             *sql.DB
 	dbDriver            string // "sqlite3" | "pgx"
@@ -56,6 +79,11 @@ func InitRedis() error {
 		Addr:     fmt.Sprintf("%s:%d", cfg.Cache.Redis.Host, cfg.Cache.Redis.Port),
 		Password: cfg.Cache.Redis.Password,
 		DB:       cfg.Cache.Redis.DB,
 		// 缓存场景失败不重试：miss 直接降级 SQL，重试只会放大网络慢时的请求延迟
 		MaxRetries:      0,
 		ReadTimeout:     3 * time.Second,
 		WriteTimeout:    3 * time.Second,
 		MinRetryBackoff: 0,
 	})
 	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
@@ -149,12 +177,14 @@ func createTables() error {
 			`CREATE TABLE IF NOT EXISTS cache (
 			key TEXT PRIMARY KEY,
 			value TEXT,
 			expi INTEGER DEFAULT 0,
 			created_at INTEGER DEFAULT (strftime('%s', 'now')),
 			updated_at INTEGER DEFAULT (strftime('%s', 'now'))
 		)`,
 			`CREATE TABLE IF NOT EXISTS cache (
 			key TEXT PRIMARY KEY,
 			value TEXT,
 			expi BIGINT DEFAULT 0,
 			created_at BIGINT DEFAULT extract(epoch FROM now())::bigint,
 			updated_at BIGINT DEFAULT extract(epoch FROM now())::bigint
 		)`},
@@ -259,8 +289,10 @@ func createTables() error {
 	// 旧表补 created_at/updated_at 列（新表 DDL 自带，此处仅老库生效）
 	for _, m := range []struct{ table, col, ddl string }{
 		{"cache", "updated_at", "INTEGER DEFAULT 0"}, // SQLite ALTER 不允许非常量默认值，先补 0 再回填
 		{"cache", "expi", "INTEGER DEFAULT 0"},
 		{"statistics", "created_at", "INTEGER DEFAULT 0"},
 		{"device_pool", "updated_at", "INTEGER DEFAULT 0"},
 		{"device_pool", "key_ts", "INTEGER DEFAULT 0"},
 		{"audio_cache", "updated_at", "INTEGER DEFAULT 0"},
 	} {
 		if dbDriver == "pgx" {
@@ -278,6 +310,7 @@ func createTables() error {
 	// 旧行回填（幂等：仅 0 值行受影响）
 	for _, backfill := range []string{
 		`UPDATE cache SET updated_at = created_at WHERE updated_at = 0`,
 		`UPDATE cache SET expi = created_at + 86400 WHERE expi = 0`,
 		`UPDATE statistics SET created_at = updated_at WHERE created_at = 0`,
 		`UPDATE device_pool SET updated_at = last_used_at WHERE updated_at = 0`,
 		`UPDATE audio_cache SET updated_at = created_at WHERE updated_at = 0`,
@@ -299,7 +332,8 @@ func isCacheEnabled() bool {
 	return cfg.Cache.Enabled
 }
-// withCache 只读接口的缓存包装：命中直接返回，miss 回源后写入（仅缓存成功响应）
+// withCache 只读接口的缓存包装：命中直接返回，miss 回源后写入（仅缓存成功响应）。
 // singleflight 合并并发 miss，同一 key 只有一个请求打上游
 // ponytail: 空参等防御分支若漏到此处，其错误 JSON 也会被缓存到 TTL 到期——handler 层 validateParams 已拦截
 func withCache(key string, ttl int, fetch func() (string, error)) (string, error) {
 	if c := GetCache(key); c != "" {
@@ -307,21 +341,36 @@ func withCache(key string, ttl int, fetch func() (string, error)) (string, error
 		// （对新写入的已解包数据无影响）
 		return reencodeUnwrapped(c), nil
 	}
 	v, err, _ := cacheSF.Do(key, func() (interface{}, error) {
 		// 双重检查：等锁期间可能已被并发请求填充
 		if c := GetCache(key); c != "" {
 			return c, nil
 		}
 		result, err := fetch()
 		if err == nil {
 			UpdateCache(key, result, ttl)
 		}
 		return result, err
 	})
 	if err != nil {
 		return "", err
 	}
 	return v.(string), nil
 }
-// GetCache 获取正文内容缓存
+// GetCache 获取缓存：Redis 热层 → SQL 持久层（命中回填 Redis）→ 未命中
 func GetCache(key string) string {
 	if !isCacheEnabled() {
 		return ""
 	}
 	if redisCache != nil {
-		return redisGetCache(key)
+		if v := redisGetCache(key); v != "" {
 			return v
 		}
 		if isVolatileCacheKey(key) {
 			return "" // 搜索/分类等动态数据不落库，Redis 未命中即回源
 		}
 	}
 	if cacheDB == nil {
@@ -329,7 +378,8 @@ func GetCache(key string) string {
 	}
 	var value string
-	err := cacheDB.QueryRow(q("SELECT value FROM cache WHERE key = ?"), key).Scan(&value)
+	var expi int64
 	err := cacheDB.QueryRow(q("SELECT value, expi FROM cache WHERE key = ?"), key).Scan(&value, &expi)
 	if err != nil {
 		if err != sql.ErrNoRows {
 			cacheLogger.Error("查询缓存失败: " + err.Error())
@@ -337,33 +387,46 @@ func GetCache(key string) string {
 		return ""
 	}
 	// 检查是否过期（expi 为 0 的旧行由启动回填兜底，此处防御）
 	if time.Now().Unix() >= expi {
 		cacheLogger.Debug("缓存已过期(SQL): " + key)
 		return ""
 	}
 	var cacheData CacheData
-	err = json.Unmarshal([]byte(value), &cacheData)
+	if err := json.Unmarshal([]byte(value), &cacheData); err != nil {
 	if err != nil {
 		cacheLogger.Error("解析缓存数据失败: " + err.Error())
 		return ""
 	}
-	cacheLogger.Debug("缓存命中: " + key)
+	cacheLogger.Debug("缓存命中(SQL): " + key)
-	// 检查是否过期
+	// SQL 命中而 Redis 未命中（过期/重启）：异步回填热层，剩余时长为 TTL，
-	if time.Now().Unix() < int64(cacheData.Expi) {
+	// 回填（尤其大 value）不阻塞本次请求
-		cacheLogger.Debug("缓存有效！")
+	if redisCache != nil {
 		remain := int(expi - time.Now().Unix())
 		if remain < 60 {
 			remain = 60
 		}
 		go redisUpdateCache(key, cacheData.Data, remain)
 	}
 	return cacheData.Data
 }
-	cacheLogger.Debug("缓存已过期！")
+// UpdateCache 写入缓存：Redis 热层（TTL=业务过期时间）+ SQL 持久层双写。
-	return ""
+// 不可变内容在 SQL 固定保留 30 天，动态短缓存（search/category）只走 Redis。
-}
+// Redis 写失败只降级不阻断：持久层照常落库，热层下次请求回填
 // UpdateCache 更新或插入正文内容缓存
 func UpdateCache(key string, value string, expi int) error {
 	if !isCacheEnabled() {
 		return nil
 	}
 	if redisCache != nil {
-		return redisUpdateCache(key, value, expi)
+		if err := redisUpdateCache(key, value, expi); err != nil {
 			cacheLogger.Warning(fmt.Sprintf("Redis 写入失败（降级为仅 SQL 落库）: %v", err))
 		} else if isVolatileCacheKey(key) {
 			return nil // 动态短缓存不落库（Redis 成功即止）
 		}
 	}
 	if cacheDB == nil {
@@ -380,10 +443,17 @@ func UpdateCache(key string, value string, expi int) error {
 		return err
 	}
 	// SQL 保留时长：不可变内容固定 30 天（Redis 未启用时沿用业务 TTL，保持纯 SQL 模式原行为）
 	sqlTTL := int64(sqlCacheTTLSeconds)
 	if redisCache == nil || isVolatileCacheKey(key) {
 		sqlTTL = int64(expi)
 	}
 	// UPSERT：ON CONFLICT 语法 SQLite 3.24+ 与 PostgreSQL 通用
-	_, err = cacheDB.Exec(q(`INSERT INTO cache (key, value, created_at, updated_at) VALUES (?, ?, ?, ?)
+	now := time.Now().Unix()
-		ON CONFLICT(key) DO UPDATE SET value = excluded.value, updated_at = excluded.updated_at`),
+	_, err = cacheDB.Exec(q(`INSERT INTO cache (key, value, expi, created_at, updated_at) VALUES (?, ?, ?, ?, ?)
-		key, string(jsonData), time.Now().Unix(), time.Now().Unix())
+		ON CONFLICT(key) DO UPDATE SET value = excluded.value, expi = excluded.expi, updated_at = excluded.updated_at`),
 		key, string(jsonData), now+sqlTTL, now, now)
 	if err != nil {
 		cacheLogger.Error("更新缓存失败: " + err.Error())
 		return err
@@ -465,11 +535,8 @@ func CleanupExpiredCache() error {
 		return fmt.Errorf("数据库未初始化")
 	}
-	// 计算过期时间点（当前时间 - cleanup_seconds）
+	// 删除已过期的缓存（按行内过期时间 expi，而非创建时间）
-	expiredTime := time.Now().Unix() - int64(cfg.Cache.CleanupSeconds)
+	result, err := cacheDB.Exec(q("DELETE FROM cache WHERE expi < ?"), time.Now().Unix())
 	// 删除过期的缓存（created_at < expiredTime）
 	result, err := cacheDB.Exec(q("DELETE FROM cache WHERE created_at < ?"), expiredTime)
 	if err != nil {
 		cacheLogger.Error("清理过期缓存失败: " + err.Error())
 		return err
@@ -480,7 +547,7 @@ func CleanupExpiredCache() error {
 	if err != nil {
 		cacheLogger.Warning("获取删除行数失败: " + err.Error())
 	} else if rowsAffected > 0 {
-		cacheLogger.Info(fmt.Sprintf("已清理 %d 条过期缓存（创建时间早于 %d 秒前）", rowsAffected, cfg.Cache.CleanupSeconds))
+		cacheLogger.Info(fmt.Sprintf("已清理 %d 条过期缓存（SQL 持久层）", rowsAffected))
 	}
 	// 顺带清理磁盘视频缓存（7 天未访问的解密文件）
@@ -30,6 +30,7 @@ func TestConcurrentCacheAccess(t *testing.T) {
 	if _, err := db.Exec(`CREATE TABLE IF NOT EXISTS cache (
 		key TEXT PRIMARY KEY,
 		value TEXT,
 		expi INTEGER DEFAULT 0,
 		created_at INTEGER DEFAULT (strftime('%s', 'now')),
 		updated_at INTEGER DEFAULT (strftime('%s', 'now'))
 	)`); err != nil {
@@ -74,6 +75,7 @@ func TestWithCache(t *testing.T) {
 	}()
 	if _, err := db.Exec(`CREATE TABLE IF NOT EXISTS cache (
 		key TEXT PRIMARY KEY, value TEXT,
 		expi INTEGER DEFAULT 0,
 		created_at INTEGER DEFAULT (strftime('%s', 'now')),
 		updated_at INTEGER DEFAULT (strftime('%s', 'now'))
 	)`); err != nil {
@@ -17,10 +17,13 @@ func HandleCategoryFront(tab interface{}) (string, error) {
 func fetchCategoryFront(tab interface{}) (string, error) {
 	tabInt := ParseIntParam(tab, 0)
 	// 新版分类接口参数（category_new_page_715=1 + new_category_type，
 	// 旧参数 new_category_tab 已失效——对齐 2026-10 可用的书源实现）
 	reqURL := BuildURL(
 		"https://reading.snssdk.com/reading/bookapi/new_category/front/v/",
-		fmt.Sprintf("new_category_tab=%d", tabInt),
+		"category_new_page_715=1",
-		fmt.Sprintf("update_version_code=%s", "58932"),
+		fmt.Sprintf("new_category_type=%d", tabInt),
 		"update_version_code=" + ClientVersionCode,
 		"distinct_style=1",
 	)
@@ -8,32 +8,54 @@ import (
 var categoryLandingLogger = common.NewLogger("category_landing", "INFO", "")
 // HandleCategoryLanding 处理获取分类详情,分类详情短缓存 10 分钟
-func HandleCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy interface{}) (string, error) {
+// selectedItems：新版分类标签自带的筛选项（selected_items，透传上游）
-	return withCache(fmt.Sprintf("category_landing_%v_%v_%v_%v_%v", categoryID, offset, wordNumber, bookStatus, sortBy), 600, func() (string, error) {
+func HandleCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy, selectedItems interface{}) (string, error) {
-		return fetchCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy)
+	return withCache(fmt.Sprintf("category_landing_%v_%v_%v_%v_%v_%v", categoryID, offset, wordNumber, bookStatus, sortBy, selectedItems), 600, func() (string, error) {
 		return fetchCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy, selectedItems)
 	})
 }
-func fetchCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy interface{}) (string, error) {
+func fetchCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy, selectedItems interface{}) (string, error) {
 	// 参数转换
 	categoryIDInt := ParseIntParam(categoryID, 0)
 	offsetInt := ParseIntParam(offset, 0)
 	wordNumberInt := ParseIntParam(wordNumber, 0)
 	bookStatusInt := ParseIntParam(bookStatus, 0)
 	sortByInt := ParseIntParam(sortBy, 0)
 	selectedItemsStr := ParseStringParam(selectedItems, "")
-	if categoryIDInt == 0 {
+	if categoryIDInt == 0 && selectedItemsStr == "" {
 		return BuildErrorResponseWithCode(100144, "参数错误"), nil
 	}
 	params := []string{
 		fmt.Sprintf("limit=%d", 20),
 		fmt.Sprintf("offset=%d", offsetInt),
 	}
 	if categoryIDInt != 0 {
 		params = append(params, fmt.Sprintf("category_id=%d", categoryIDInt))
 	}
 	if wordNumberInt != 0 {
 		params = append(params, fmt.Sprintf("word_number=%d", wordNumberInt))
 	}
 	if bookStatusInt != 0 {
 		params = append(params, fmt.Sprintf("book_status=%d", bookStatusInt))
 	}
 	if sortByInt != 0 {
 		params = append(params, fmt.Sprintf("sort_by=%d", sortByInt))
 	}
 	// 新版分类标签：selected_items 由 front 页标签自带，需与其 from_front_page 配对透传
 	if selectedItemsStr != "" {
 		params = append(params,
 			"selected_items="+selectedItemsStr,
 			"selected_item_from_front_page="+selectedItemsStr,
 			"source=front_category",
 		)
 	}
 	reqURL := BuildURL(
 		"https://reading.snssdk.com/reading/bookapi/new_category/landing/v/",
-		fmt.Sprintf("category_id=%d", categoryIDInt),
+		params...,
 		fmt.Sprintf("limit=%d", 20),
 		fmt.Sprintf("offset=%d", offsetInt),
 		fmt.Sprintf("word_number=%d", wordNumberInt),
 		fmt.Sprintf("book_status=%d", bookStatusInt),
 		fmt.Sprintf("sort_by=%d", sortByInt),
 	)
 	resp, err := GorgonRequest(reqURL, map[string]interface{}{
@@ -5,12 +5,14 @@ import (
 	"encoding/json"
 	"fanqie/internal/common"
 	"fmt"
 	"sort"
 	"strings"
 	"golang.org/x/sync/singleflight"
 )
 var (
 	contentLogger = common.NewLogger("content", "INFO", "")
 	contentSF     singleflight.Group // 合并同一章节的并发回源请求，防缓存击穿
 )
 // ContentFetcher 内容获取器
@@ -26,19 +28,27 @@ func HandleGetContent(itemID string, textMode, imageMode interface{}) (string, e
 	textModeBool := parseBoolParamDefault(textMode, getContentConfig().RemoveHTMLTags)
 	imageModeBool := parseBoolParamDefault(imageMode, !getContentConfig().RemoveImages)
 	// contentLogger.Info(fmt.Sprintf("获取内容 item_id=%s text_mode=%v image_mode=%v", itemID, textModeBool, imageModeBool))
 	// 检查缓存
 	cache := GetCache("content_" + itemID)
-	if cache != "" {
+	if cache == "" {
 		// 同一章节并发请求只打一次上游，其余等待共享结果
 		v, err, _ := contentSF.Do("content_"+itemID, func() (interface{}, error) {
 			// 双重检查：等锁期间可能已被并发请求填充
 			if c := GetCache("content_" + itemID); c != "" {
 				return c, nil
 			}
 			return fetchContentFromAPI(itemID)
 		})
 		if err != nil {
 			return "", err
 		}
 		cache = v.(string)
 	}
 	// 出口按各自参数解析（缓存/回源共用同一路径）
 	return processCachedContent(cache, textModeBool, imageModeBool)
 }
-	// 缓存未命中，从API获取
+// processCachedContent 处理缓存内容（content 为解密后的原始正文）
 	return fetchContentFromAPI(itemID, textModeBool, imageModeBool)
 }
 // processCachedContent 处理缓存内容
 func processCachedContent(cache string, textMode, imageMode bool) (string, error) {
 	var d map[string]interface{}
 	json.Unmarshal([]byte(cache), &d)
@@ -53,18 +63,18 @@ func processCachedContent(cache string, textMode, imageMode bool) (string, error
 	return buildJSONResponse(UnwrapJSONStrings(d)), nil
 }
-// fetchContentFromAPI 从API获取内容（带设备切换重试）
+// fetchContentFromAPI 从API获取内容（带设备切换重试），返回缓存格式 JSON
-func fetchContentFromAPI(itemID string, textMode, imageMode bool) (string, error) {
+func fetchContentFromAPI(itemID string) (string, error) {
-	return fetchContentWithRetry(itemID, textMode, imageMode, 0)
+	return fetchContentWithRetry(itemID, 0)
 }
 // fetchContentWithRetry 从API获取内容，支持设备切换重试
-func fetchContentWithRetry(itemID string, textMode, imageMode bool, retryCount int) (string, error) {
+func fetchContentWithRetry(itemID string, retryCount int) (string, error) {
 	// 防止无限递归
 	const maxRetries = 5
 	if retryCount >= maxRetries {
 		contentLogger.Warning(fmt.Sprintf("已重试 %d 次，尝试备用接口", maxRetries))
-		return tryBackupAPI(itemID, textMode, imageMode)
+		return tryBackupAPI(itemID)
 	}
 	// 从设备密钥池获取设备
@@ -79,6 +89,7 @@ func fetchContentWithRetry(itemID string, textMode, imageMode bool, retryCount i
 	currentKeyInfo := &KeyInfo{
 		Key: item.Key,
 		Ver: item.KeyVer,
 		Ts:  item.KeyTs,
 	}
 	contentLogger.Debug(fmt.Sprintf("使用池中设备 ID=%d (重试次数: %d)", item.ID, retryCount))
@@ -90,16 +101,13 @@ func fetchContentWithRetry(itemID string, textMode, imageMode bool, retryCount i
 		"app_name=novelapp",
 		"channel=0",
 		"device_platform=android",
-		"version_code=66.9",
+		"version_code=" + ClientVersionCode,
 		"item_id=" + itemID,
 		"os_version=0",
 		fmt.Sprintf("key_register_ts=%d", currentKeyInfo.Ts), // 降低上游密钥不匹配概率
 		headerWithoutVersion,
 	}
-
+	sortedParams := buildSortedQuery(params...)
 	allParams := strings.Join(params, "&")
 	paramList := strings.Split(allParams, "&")
 	sort.Strings(paramList)
 	sortedParams := strings.Join(paramList, "&")
 	url := "https://reading.snssdk.com/reading/reader/full/v/?" + sortedParams
@@ -114,7 +122,7 @@ func fetchContentWithRetry(itemID string, textMode, imageMode bool, retryCount i
 		contentLogger.Warning(fmt.Sprintf("主接口请求出错，标记设备无效并切换设备: %v", err))
 		pool.MarkInvalid(header)
 		pool.RemoveInvalid()
-		return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
+		return fetchContentWithRetry(itemID, retryCount+1)
 	}
 	var body map[string]interface{}
@@ -126,7 +134,7 @@ func fetchContentWithRetry(itemID string, textMode, imageMode bool, retryCount i
 			contentLogger.Warning("响应data字段无效，切换设备重试")
 			pool.MarkInvalid(header)
 			pool.RemoveInvalid()
-			return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
+			return fetchContentWithRetry(itemID, retryCount+1)
 		}
 		contentEncoded, ok := data["content"].(string)
@@ -135,76 +143,101 @@ func fetchContentWithRetry(itemID string, textMode, imageMode bool, retryCount i
 			contentLogger.Warning(fmt.Sprintf("内容返回 Invalid，标记设备无效并切换设备 (重试: %d)", retryCount))
 			pool.MarkInvalid(header)
 			pool.RemoveInvalid()
-			return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
+			return fetchContentWithRetry(itemID, retryCount+1)
 		}
 		// 检查密钥版本
 		keyVersion, _ := data["key_version"].(float64)
-		if currentKeyInfo != nil && int(keyVersion) != currentKeyInfo.Ver {
+		if int(keyVersion) != currentKeyInfo.Ver {
 			contentLogger.Warning("密钥版本不匹配，重新获取密钥")
 			// 更新池中的密钥
 			newKeyInfo, err := GetKey(header)
 			if err == nil {
-				UpdateKeyInPool(header, newKeyInfo.Key, newKeyInfo.Ver)
+				UpdateKeyInPool(header, newKeyInfo)
 				currentKeyInfo = newKeyInfo
 			} else {
 				// 获取密钥失败，切换设备
 				contentLogger.Warning("获取密钥失败，切换设备重试")
 				pool.MarkInvalid(header)
 				pool.RemoveInvalid()
-				return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
+				return fetchContentWithRetry(itemID, retryCount+1)
 			}
 		}
-		// 解密内容
+		contentStr, err := decodeContentBody(data, currentKeyInfo)
 		contentBytes, _ := base64.StdEncoding.DecodeString(contentEncoded)
 		if len(contentBytes) < 16 {
 			contentLogger.Warning("内容长度不足，切换设备重试")
 			pool.MarkInvalid(header)
 			pool.RemoveInvalid()
 			return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
 		}
 		iv := contentBytes[0:16]
 		encryptedData := contentBytes[16:]
 		keyBytes, _ := base64.StdEncoding.DecodeString(currentKeyInfo.Key)
 		decrypted, err := De(encryptedData, keyBytes, iv)
 		if err != nil {
 			contentLogger.Warning(fmt.Sprintf("解密失败，切换设备重试: %v", err))
 			pool.MarkInvalid(header)
 			pool.RemoveInvalid()
-			return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
+			return fetchContentWithRetry(itemID, retryCount+1)
 		}
-		// 处理内容
+		contentStr = strings.ReplaceAll(contentStr, `<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>`, "")
 		contentStr := string(decrypted)
 		contentStr = strings.Replace(contentStr, `<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"no\"?>`, "", -1)
 		contentStr = strings.TrimSpace(contentStr)
 		data["content"] = contentStr
-		// 更新缓存
+		// 写缓存并返回缓存格式（按参数的解析统一放在出口 processCachedContent）
-		cacheData := buildCacheData(data)
+		cacheStr := marshalNoEscape(buildCacheData(UnwrapJSONStrings(data)))
-		saveCacheJSON("content_"+itemID, cacheData, 86400*2)
+		UpdateCache("content_"+itemID, cacheStr, 86400)
-
+		return cacheStr, nil
 		// 解析内容
 		parser := NewContentParser()
 		data["content"] = parser.Parse(contentStr, textMode, imageMode)
 		return buildJSONResponse(UnwrapJSONStrings(body)), nil
 	}
 	// 响应code不为0，切换设备重试
 	contentLogger.Warning(fmt.Sprintf("响应code异常: %.0f，切换设备重试", body["code"]))
 	pool.MarkInvalid(header)
 	pool.RemoveInvalid()
-	return fetchContentWithRetry(itemID, textMode, imageMode, retryCount+1)
+	return fetchContentWithRetry(itemID, retryCount+1)
 }
-// tryBackupAPI 尝试备用API（所有设备都失败后的最后手段）
+// decodeContentBody 按上游 crypt_status 解出正文：
-func tryBackupAPI(itemID string, textMode, imageMode bool) (string, error) {
+//   - 0：base64 密文，设备密钥 AES-CBC 解密（compress_status!=0 时解压）
 //   - 1：明文（可能压缩）
 //   - 2：密钥已失效，刷新密钥后由调用方重试
 func decodeContentBody(data map[string]interface{}, keyInfo *KeyInfo) (string, error) {
 	contentEncoded, _ := data["content"].(string)
 	cryptStatus, _ := data["crypt_status"].(float64)
 	compressStatus, _ := data["compress_status"].(float64)
 	switch int(cryptStatus) {
 	case 1:
 		if compressStatus != 0 {
 			return string(gunzipMaybe([]byte(contentEncoded))), nil
 		}
 		return contentEncoded, nil
 	case 2:
 		return "", fmt.Errorf("密钥已失效（crypt_status=2），需刷新密钥重试")
 	default:
 		// 500003 场景（设备无需解密）仍下发密文时按明文兜底
 		if keyInfo == nil || keyInfo.Key == "" {
 			return contentEncoded, nil
 		}
 		contentBytes, err := base64.StdEncoding.DecodeString(contentEncoded)
 		if err != nil || len(contentBytes) < 16 {
 			return "", fmt.Errorf("内容密文长度异常")
 		}
 		iv := contentBytes[0:16]
 		encryptedData := contentBytes[16:]
 		keyBytes, _ := base64.StdEncoding.DecodeString(keyInfo.Key)
 		decrypted, err := De(encryptedData, keyBytes, iv)
 		if err != nil {
 			return "", err
 		}
 		if compressStatus != 0 {
 			decrypted = gunzipMaybe(decrypted)
 		}
 		return string(decrypted), nil
 	}
 }
 // tryBackupAPI 尝试备用API（所有设备都失败后的最后手段），返回缓存格式 JSON
 func tryBackupAPI(itemID string) (string, error) {
 	contentLogger.Warning("尝试备用接口1（所有设备均失败）")
 	// 从设备池获取任意一个有效设备用于备用接口
@@ -217,8 +250,8 @@ func tryBackupAPI(itemID string, textMode, imageMode bool) (string, error) {
 	header := item.DeviceHeader
 	// 备用接口1：使用DH加密
-	backupAPI := fmt.Sprintf("https://novel.snssdk.com/api/novel/book/reader/content/v1?support_image=1&item_id=%s&aid=1319&app_name=novelapp&channel=0&device_platform=android&os_version=2&version_code=58932&%s",
+	backupAPI := fmt.Sprintf("https://novel.snssdk.com/api/novel/book/reader/content/v1?support_image=1&item_id=%s&aid=1319&app_name=novelapp&channel=0&device_platform=android&os_version=2&version_code=%s&%s",
-		itemID, header)
+		itemID, ClientVersionCode, header)
 	cm := NewCM()
 	resp, err := GorgonRequest(backupAPI, map[string]interface{}{
@@ -259,18 +292,14 @@ func tryBackupAPI(itemID string, textMode, imageMode bool) (string, error) {
 			contentLogger.Error("解密失败: " + err.Error())
 			return `{"code": -1, "msg": "解密失败"}`, err
 		}
 		decryptedContent = string(gunzipMaybe([]byte(decryptedContent)))
-		data["content"] = decryptedContent
+		data["content"] = strings.TrimSpace(decryptedContent)
-		// 更新缓存
+		// 写缓存并返回缓存格式
-		bodyJSONStr := buildJSONResponse(body)
+		cacheStr := marshalNoEscape(buildCacheData(UnwrapJSONStrings(data)))
-		UpdateCache("content_"+itemID, bodyJSONStr, 86400)
+		UpdateCache("content_"+itemID, cacheStr, 86400)
-
+		return cacheStr, nil
 		// 解析内容
 		parser := NewContentParser()
 		data["content"] = parser.Parse(decryptedContent, textMode, imageMode)
 		return buildJSONResponse(UnwrapJSONStrings(body)), nil
 	}
 	contentLogger.Warning("备用接口1获取正文出错！")
@@ -9,6 +9,20 @@ import (
 	"github.com/PuerkitoBio/goquery"
 )
 // imgSrcRe 匹配 img 标签的 src 属性
 var imgSrcRe = regexp.MustCompile(`(?i)(<img[^>]*\ssrc=")([^"]*)(")`)
 // permanentizeImgSrc 将正文插图链接替换为永久 CDN 链接（原链接带临时参数会过期）
 func permanentizeImgSrc(html string) string {
 	return imgSrcRe.ReplaceAllStringFunc(html, func(m string) string {
 		parts := imgSrcRe.FindStringSubmatch(m)
 		if len(parts) < 3 || parts[2] == "" {
 			return m
 		}
 		return parts[1] + NormalizeNovelImageURL(parts[2]) + parts[3]
 	})
 }
 var (
 	parserLogger = common.NewLogger("content_parser", "INFO", "")
 )
@@ -118,6 +132,8 @@ func (p *ContentParser) extractText(content string) string {
 				divHTML = decodeHTMLEntities(divHTML)
 				// 标准化img属性
 				divHTML = normalizeImgAttributes(divHTML)
 				// 插图链接永久化
 				divHTML = permanentizeImgSrc(divHTML)
 				finalResult = append(finalResult, divHTML)
 			}
 			return
@@ -173,6 +189,9 @@ func (p *ContentParser) processHTMLMode(content string) string {
 	// 解码HTML实体
 	content = decodeHTMLEntities(content)
 	// 插图链接永久化（原链接带临时参数会过期）
 	content = permanentizeImgSrc(content)
 	// 根据配置决定是否移除空行
 	if p.cfg.RemoveBlankLines {
 		// 对HTML内容移除空的<p>标签
@@ -2,12 +2,84 @@ package fqapi
 import (
 	"bytes"
 	"compress/gzip"
 	"compress/zlib"
 	"encoding/json"
 	"fanqie/internal/config"
-	"sort"
+	"io"
 	"net/url"
 	"regexp"
 	"strings"
 )
 // 客户端版本统一（与 2026-10 可用的书源实现一致；旧值 58932/5.8.9.32 混用易触发风控）
 const (
 	ClientVersionCode = "72732"
 	ClientVersionName = "7.2.7.32"
 )
 // PermanentNovelImgHost 番茄小说图片的永久 CDN 域名（原始链接带临时参数会过期失效）
 const PermanentNovelImgHost = "https://p6-novel.byteimg.com"
 // novelPicMd5Re 匹配图片路径中的 32 位十六进制资源 ID
 var novelPicMd5Re = regexp.MustCompile(`^/novel-pic/([a-f0-9]{32})`)
 // NormalizeNovelImageURL 将临时图片链接永久化：
 //   - /novel-pic/{md5} → p6-novel.byteimg.com/origin/novel-pic/{md5}（正文插图原图）
 //   - 含 ~tplv 的模板图（封面缩放）→ p6-novel.byteimg.com 保留原路径
 func NormalizeNovelImageURL(raw string) string {
 	u := strings.TrimSpace(strings.ReplaceAll(raw, "&amp;", "&"))
 	if u == "" {
 		return ""
 	}
 	if strings.HasPrefix(u, "//") {
 		u = "https:" + u
 	}
 	idx := strings.Index(u, "/novel-pic/")
 	if idx < 0 {
 		return u
 	}
 	path := u[idx:]
 	if m := novelPicMd5Re.FindStringSubmatch(path); m != nil && !strings.Contains(path, "~tplv") {
 		return PermanentNovelImgHost + "/origin/novel-pic/" + m[1]
 	}
 	return PermanentNovelImgHost + path
 }
 // gunzipMaybe 自适应解压上游正文（gzip 与 zlib 帧都可能出现），非压缩数据原样返回
 func gunzipMaybe(data []byte) []byte {
 	if r, err := gzip.NewReader(bytes.NewReader(data)); err == nil {
 		if out, err2 := io.ReadAll(r); err2 == nil {
 			r.Close()
 			return out
 		}
 	}
 	if r, err := zlib.NewReader(bytes.NewReader(data)); err == nil {
 		if out, err2 := io.ReadAll(r); err2 == nil {
 			r.Close()
 			return out
 		}
 	}
 	return data
 }
 // buildSortedQuery 将若干 "k=v" 段合并为 form 编码的查询串（按键排序）。
 // 设备机型等值含空格时必须编码，否则实际请求的 URL 与签名输入不一致，
 // registerkey 等接口会注册失败（书源 2026-10-10 修复的同类问题）
 func buildSortedQuery(parts ...string) string {
 	vals := url.Values{}
 	for _, part := range parts {
 		for _, seg := range strings.Split(part, "&") {
 			if seg == "" {
 				continue
 			}
 			k, v, _ := strings.Cut(seg, "=")
 			vals.Add(k, v)
 		}
 	}
 	return vals.Encode()
 }
 // getContentConfig 获取内容配置（带默认值）
 func getContentConfig() config.ContentConfig {
 	if config.GetAppConfig() != nil {
@@ -188,7 +260,7 @@ func getDeviceHeader() string {
 // removeVersionFromHeader 移除header中的version_name参数
 func removeVersionFromHeader(header string) string {
-	return strings.Replace(header, "&version_name=5.8.9.32", "", -1)
+	return strings.Replace(header, "&version_name="+ClientVersionName, "", -1)
 }
 // ============ API参数构建工具 ============
@@ -203,7 +275,7 @@ func BuildURL(fullURL string, customParams ...string) string {
 		"channel=0",
 		"device_platform=android",
 		"os_version=0",
-		"version_code=58932",
+		"version_code=" + ClientVersionCode,
 	}
 	// 添加设备header
@@ -215,13 +287,8 @@ func BuildURL(fullURL string, customParams ...string) string {
 	// 添加自定义参数
 	allParams = append(allParams, customParams...)
-	// 合并、拆分、排序
+	// form 编码 + 排序（机型含空格等特殊字符时保证 URL 与签名一致）
-	joined := strings.Join(allParams, "&")
+	return fullURL + "?" + buildSortedQuery(allParams...)
 	paramList := strings.Split(joined, "&")
 	sort.Strings(paramList)
 	params := strings.Join(paramList, "&")
 	return fullURL + "?" + params
 }
 // ============ 缓存处理工具函数 ============
@@ -0,0 +1,31 @@
 package fqapi
 import "testing"
 // TestBuildSortedQuery form 编码 + 排序（机型含空格时 URL 与签名一致的关键）
 func TestBuildSortedQuery(t *testing.T) {
 	got := buildSortedQuery("b=2", "a=1", "device_type=Pixel 6")
 	want := "a=1&b=2&device_type=Pixel+6"
 	if got != want {
 		t.Fatalf("buildSortedQuery = %q, want %q", got, want)
 	}
 }
 // TestNormalizeNovelImageURL 插图/封面/已永久化链接的归一化
 func TestNormalizeNovelImageURL(t *testing.T) {
 	cases := map[string]string{
 		// 正文插图（纯 md5）→ origin 原图
 		"//p9-novel.byteimg.com/img/novel-pic/0123456789abcdef0123456789abcdef": "https://p6-novel.byteimg.com/origin/novel-pic/0123456789abcdef0123456789abcdef",
 		// 已永久化的链接幂等
 		"https://p6-novel.byteimg.com/origin/novel-pic/0123456789abcdef0123456789abcdef": "https://p6-novel.byteimg.com/origin/novel-pic/0123456789abcdef0123456789abcdef",
 		// 封面模板图（~tplv 缩放）→ 保留原路径仅换域名
 		"https://p3-novel.byteimg.com/novel-pic/0123456789abcdef0123456789abcdef~tplv-novel-pic-large.webp": "https://p6-novel.byteimg.com/novel-pic/0123456789abcdef0123456789abcdef~tplv-novel-pic-large.webp",
 		// 非 novel-pic 链接原样返回
 		"https://example.com/other.jpg": "https://example.com/other.jpg",
 	}
 	for in, want := range cases {
 		if got := NormalizeNovelImageURL(in); got != want {
 			t.Errorf("NormalizeNovelImageURL(%q) = %q, want %q", in, got, want)
 		}
 	}
 }
@@ -42,9 +42,13 @@ func fetchDetail(bookID string) (string, error) {
 	}
 	if code, ok := respJSON["code"].(float64); ok && code == 0 {
 		// 删除content字段
 		if data, ok := respJSON["data"].(map[string]interface{}); ok {
 			// 删除content字段
 			delete(data, "content")
 			// 封面链接永久化（原链接带临时参数会过期）
 			if thumb, ok := data["thumb_url"].(string); ok && thumb != "" {
 				data["thumb_url"] = NormalizeNovelImageURL(thumb)
 			}
 		}
 		// 重新序列化（解包 JSON 字符串字段）
@@ -20,9 +20,14 @@ type DeviceKeyItem struct {
 	DeviceHeader string    `json:"device_header"`
 	Key          string    `json:"key"`
 	KeyVer       int       `json:"key_ver"`
 	KeyTs        int64     `json:"key_ts"` // 密钥注册时间戳（key_register_ts）
 	IsValid      bool      `json:"is_valid"`
 	CreatedAt    time.Time `json:"created_at"`
 	LastUsedAt   time.Time `json:"last_used_at"`
 	// 仅内存字段（不落库）
 	KeyFetchedAt time.Time `json:"-"` // 密钥获取时间，超 24h 异步刷新
 	KeyRefreshAt time.Time `json:"-"` // 上次刷新尝试时间，防刷新风暴
 }
 // DeviceKeyPool 设备密钥池
@@ -77,6 +82,10 @@ func (p *DeviceKeyPool) Initialize() error {
 	}
 	poolLogger.Info(fmt.Sprintf("设备密钥池初始化完成，当前有效设备数: %d", p.getValidCount()))
 	// last_used_at 批量落库（GetNext 不再同步写库）
 	p.StartDevicePoolFlusher()
 	return nil
 }
@@ -110,11 +119,22 @@ func (p *DeviceKeyPool) GetNext() (*DeviceKeyItem, error) {
 	item := validItems[p.currentIndex%len(validItems)]
 	p.currentIndex++
-	// 更新最后使用时间
+	// 只更新内存中的最后使用时间，由后台 flusher 定期批量落库
 	// （此前每次取设备同步 UPDATE，SQLite 单写锁下是吞吐瓶颈）
 	item.LastUsedAt = time.Now()
 	p.updateDeviceInDB(item)
-	// poolLogger.Debug(fmt.Sprintf("使用设备 ID=%d", item.ID))
+	// 密钥超 24h 未刷新：异步换新（本次请求仍用旧密钥，下次生效）
 	if time.Since(item.KeyFetchedAt) > 20*time.Hour && time.Since(item.KeyRefreshAt) > 10*time.Minute {
 		item.KeyRefreshAt = time.Now()
 		header := item.DeviceHeader
 		go func() {
 			if ki, err := GetKey(header); err == nil {
 				UpdateKeyInPool(header, ki)
 			} else {
 				poolLogger.Warning(fmt.Sprintf("刷新设备密钥失败: %v", err))
 			}
 		}()
 	}
 	// 异步检查是否需要补充
 	go p.checkAndRefill()
@@ -316,9 +336,11 @@ func (p *DeviceKeyPool) addNewDevice() error {
 		DeviceHeader: deviceHeader,
 		Key:          keyInfo.Key,
 		KeyVer:       keyInfo.Ver,
 		KeyTs:        keyInfo.Ts,
 		IsValid:      true,
 		CreatedAt:    time.Now(),
 		LastUsedAt:   time.Now(),
 		KeyFetchedAt: time.Now(),
 	}
 	// 保存到数据库
@@ -344,7 +366,7 @@ func (p *DeviceKeyPool) loadFromDB() error {
 	}
 	rows, err := cacheDB.Query(`
-		SELECT id, device_header, key, key_ver, is_valid, created_at, last_used_at 
+		SELECT id, device_header, key, key_ver, COALESCE(key_ts, 0), is_valid, created_at, last_used_at
 		FROM device_pool
 		ORDER BY id
 	`)
@@ -363,6 +385,7 @@ func (p *DeviceKeyPool) loadFromDB() error {
 			&item.DeviceHeader,
 			&item.Key,
 			&item.KeyVer,
 			&item.KeyTs,
 			&item.IsValid,
 			&createdAt,
 			&lastUsedAt,
@@ -374,6 +397,7 @@ func (p *DeviceKeyPool) loadFromDB() error {
 		item.CreatedAt = time.Unix(createdAt, 0)
 		item.LastUsedAt = time.Unix(lastUsedAt, 0)
 		item.KeyFetchedAt = time.Now() // 旧设备的密钥年龄未知，从加载起算
 		items = append(items, &item)
 	}
@@ -401,10 +425,10 @@ func (p *DeviceKeyPool) saveDeviceToDB(item *DeviceKeyItem) error {
 	// RETURNING id 同时兼容 SQLite(3.35+) 与 PostgreSQL，替代 LastInsertId()
 	err := cacheDB.QueryRow(q(`
-		INSERT INTO device_pool (device_header, key, key_ver, is_valid, created_at, last_used_at, updated_at)
+		INSERT INTO device_pool (device_header, key, key_ver, key_ts, is_valid, created_at, last_used_at, updated_at)
-		VALUES (?, ?, ?, ?, ?, ?, ?)
+		VALUES (?, ?, ?, ?, ?, ?, ?, ?)
 		RETURNING id
-	`), item.DeviceHeader, item.Key, item.KeyVer, boolToInt(item.IsValid),
+	`), item.DeviceHeader, item.Key, item.KeyVer, item.KeyTs, boolToInt(item.IsValid),
 		item.CreatedAt.Unix(), item.LastUsedAt.Unix(), item.CreatedAt.Unix()).Scan(&item.ID)
 	return err
@@ -418,14 +442,39 @@ func (p *DeviceKeyPool) updateDeviceInDB(item *DeviceKeyItem) error {
 	_, err := cacheDB.Exec(q(`
 		UPDATE device_pool
-		SET device_header = ?, key = ?, key_ver = ?, is_valid = ?, last_used_at = ?, updated_at = ?
+		SET device_header = ?, key = ?, key_ver = ?, key_ts = ?, is_valid = ?, last_used_at = ?, updated_at = ?
 		WHERE id = ?
-	`), item.DeviceHeader, item.Key, item.KeyVer, boolToInt(item.IsValid),
+	`), item.DeviceHeader, item.Key, item.KeyVer, item.KeyTs, boolToInt(item.IsValid),
 		item.LastUsedAt.Unix(), item.LastUsedAt.Unix(), item.ID)
 	return err
 }
 // StartDevicePoolFlusher 启动 last_used_at 批量落库任务（GetNext 已改为只写内存）
 func (p *DeviceKeyPool) StartDevicePoolFlusher() {
 	go func() {
 		for range time.Tick(time.Minute) {
 			FlushDevicePool()
 		}
 	}()
 	poolLogger.Info("设备池 last_used_at 批量落库已启动（每分钟）")
 }
 // FlushDevicePool 将内存中的设备使用时间批量刷入数据库
 // ponytail: 进程退出时最多丢 1 分钟使用记录（仅统计用途），无需精确停表
 func FlushDevicePool() {
 	p := GetDeviceKeyPool()
 	if !p.enabled || cacheDB == nil {
 		return
 	}
 	p.mu.RLock()
 	items := append([]*DeviceKeyItem(nil), p.items...)
 	p.mu.RUnlock()
 	for _, it := range items {
 		p.updateDeviceInDB(it)
 	}
 }
 // removeInvalidFromDB 从数据库移除无效设备
 func (p *DeviceKeyPool) removeInvalidFromDB() error {
 	if cacheDB == nil {
@@ -483,10 +532,10 @@ func GetKeyInfoFromPool(deviceHeader string) *KeyInfo {
 	return nil
 }
-// UpdateKeyInPool 更新池中设备的密钥
+// UpdateKeyInPool 更新池中设备的密钥（含注册时间戳）
-func UpdateKeyInPool(deviceHeader string, key string, ver int) error {
+func UpdateKeyInPool(deviceHeader string, ki *KeyInfo) error {
 	pool := GetDeviceKeyPool()
-	if !pool.IsEnabled() {
+	if !pool.IsEnabled() || ki == nil {
 		return nil
 	}
@@ -495,8 +544,10 @@ func UpdateKeyInPool(deviceHeader string, key string, ver int) error {
 	for _, item := range pool.items {
 		if item.DeviceHeader == deviceHeader {
-			item.Key = key
+			item.Key = ki.Key
-			item.KeyVer = ver
+			item.KeyVer = ki.Ver
 			item.KeyTs = ki.Ts
 			item.KeyFetchedAt = time.Now()
 			return pool.updateDeviceInDB(item)
 		}
 	}
@@ -5,6 +5,7 @@ import (
 	"encoding/hex"
 	"encoding/json"
 	"fanqie/internal/common"
 	"fanqie/internal/config"
 	"fmt"
 	"math/big"
 	"net"
@@ -186,6 +187,22 @@ func generateIPv6List() []map[string]string {
 	}
 }
 // deviceIdentity 注册设备身份：pad（默认，历史行为）/ phone（对齐手机 App 书源注册 payload）
 func deviceIdentity() string {
 	if cfg := config.GetConfig(); cfg != nil && cfg.Fanqie.DeviceIdentity == "phone" {
 		return "phone"
 	}
 	return "pad"
 }
 // deviceCustomFields 注册身份对应的 custom 字段
 func deviceCustomFields() map[string]interface{} {
 	if deviceIdentity() == "phone" {
 		return map[string]interface{}{"host_bit": 64, "dragon_device_type": "phone"}
 	}
 	return map[string]interface{}{"host_bit": 32, "dragon_device_type": "pad"}
 }
 // GenerateFullRequestBody 生成完整的请求体数据
 func GenerateFullRequestBody() map[string]interface{} {
 	rawData := map[string]interface{}{
@@ -214,7 +231,7 @@ func GenerateFullRequestBody() map[string]interface{} {
 			"sim_serial_number":     []interface{}{},
 			"oaid_may_support":      false,
 			"device_platform":       "android",
-			"custom":                map[string]interface{}{"host_bit": 32, "dragon_device_type": "pad"},
+			"custom":                deviceCustomFields(),
 			"pre_installed_channel": "",
 			"is_system_app":         0,
 			"sdk_flavor":            "china",
@@ -234,6 +251,16 @@ func GenerateFullRequestBody() map[string]interface{} {
 	}
 	header["ipv6_list"] = generateIPv6List()
 	// phone 身份补充真机指纹字段（对齐书源注册 payload，规避 pad 身份的差异化风控）
 	if deviceIdentity() == "phone" {
 		rom, _ := header["rom"].(string)
 		header["hardware"] = "qcom"
 		header["device_device"] = rom
 		header["fingerprint"] = fmt.Sprintf("%v/%v/%v:%v/%s/%s:user/release-keys",
 			header["device_brand"], header["device_model"], header["device_model"],
 			header["os_version"], rom, rom)
 	}
 	for k, v := range genTime {
 		rawData[k] = v
 	}
@@ -252,9 +279,9 @@ func tryActivatePremium(header string) {
 		"channel=0",
 		"device_platform=android",
 		"os_version=0",
-		"version_code=58932",
+		"version_code=" + ClientVersionCode,
-		"manifest_version_code=58932",
+		"manifest_version_code=" + ClientVersionCode,
-		"update_version_code=58932",
+		"update_version_code=" + ClientVersionCode,
 		header,
 	}
@@ -365,7 +392,7 @@ func Register() (string, error) {
 		}
 		deviceModel := data["header"].(map[string]interface{})["device_model"].(string)
-		header := fmt.Sprintf("device_id=%v&device_type=%s&iid=%v&version_name=5.8.9.32",
+		header := fmt.Sprintf("device_id=%v&device_type=%s&iid=%v&version_name="+ClientVersionName,
 			respData["device_id_str"],
 			deviceModel,
 			respData["install_id_str"])
@@ -14,6 +14,16 @@ func CreateMD5(s string) string {
 	return hex.EncodeToString(hasher.Sum(nil))
 }
 // setCommonHeaders 注入番茄客户端通用请求头（与签名头不冲突，对齐书源实现降低风控特征）
 func setCommonHeaders(headers map[string]string) {
 	if _, ok := headers["x-tt-store-region"]; ok {
 		return
 	}
 	headers["x-tt-store-region"] = "cn-zj"
 	headers["x-tt-store-region-src"] = "did"
 	headers["sdk-version"] = "2"
 }
 // GorgonRequest Gorgon请求（使用签名切换器）
 func GorgonRequest(urlStr string, options map[string]interface{}) (*common.ClientResponse, error) {
 	headers, ok := options["headers"].(map[string]string)
@@ -21,6 +31,7 @@ func GorgonRequest(urlStr string, options map[string]interface{}) (*common.Clien
 		headers = make(map[string]string)
 		options["headers"] = headers
 	}
 	setCommonHeaders(headers)
 	dataMD5 := ""
 	if body, ok := options["body"].(string); ok && body != "" {
@@ -53,6 +64,7 @@ func GorgonRequestSync(urlStr string, options map[string]interface{}) (*common.C
 		headers = make(map[string]string)
 		options["headers"] = headers
 	}
 	setCommonHeaders(headers)
 	dataMD5 := ""
 	if body, ok := options["body"].(string); ok && body != "" {
@@ -69,9 +81,7 @@ func GorgonRequestSync(urlStr string, options map[string]interface{}) (*common.C
 	signedHeaders := algos.GlobalSignatureSwitcher.Sign(urlStr, headers)
 	// 删除x-ss-stub（如果存在）
 	if _, exists := signedHeaders["x-ss-stub"]; exists {
 	delete(signedHeaders, "x-ss-stub")
 	}
 	// 更新headers
 	for k, v := range signedHeaders {
@@ -90,6 +100,7 @@ func ContentRequest(urlStr string, options map[string]interface{}) (*common.Clie
 		headers = make(map[string]string)
 		options["headers"] = headers
 	}
 	setCommonHeaders(headers)
 	dataMD5 := ""
 	if body, ok := options["body"].(string); ok && body != "" {
@@ -58,9 +58,14 @@ type ImageResult struct {
 	Index  int    `json:"index"`
 }
-// HandleGetManga 处理漫画请求
+// HandleGetManga 处理漫画请求；设备池签名路线全部失败时降级 wkbrowser 免签名兜底
 func HandleGetManga(itemID string, showHTML bool, urlOutputMode string) (interface{}, error) {
-	return getMangaWithRetry(itemID, showHTML, urlOutputMode, 0)
+	res, err := getMangaWithRetry(itemID, showHTML, urlOutputMode, 0)
 	if err == nil {
 		return res, nil
 	}
 	mangaLogger.Warning(fmt.Sprintf("签名路线失败，尝试 wkbrowser 免签名兜底: %v", err))
 	return fetchMangaFallback(itemID, showHTML)
 }
 // getMangaWithRetry 获取漫画内容，支持设备切换重试
@@ -87,7 +92,7 @@ func getMangaWithRetry(itemID string, showHTML bool, urlOutputMode string, retry
 	mangaLogger.Debug(fmt.Sprintf("使用池中设备 ID=%d (重试次数: %d)", item.ID, retryCount))
 	// 2. 构建URL - 使用 BuildURL 添加设备参数
-	headerWithoutVersion := strings.Replace(header, "&version_name=5.8.9.32", "", -1)
+	headerWithoutVersion := removeVersionFromHeader(header)
 	url := BuildURL("https://reading.snssdk.com/reading/reader/full/v/",
 		"aid=1967",
@@ -146,7 +151,7 @@ func getMangaWithRetry(itemID string, showHTML bool, urlOutputMode string, retry
 			mangaLogger.Warning("密钥版本不匹配，重新获取密钥")
 			newKeyInfo, err := GetKey(header)
 			if err == nil {
-				UpdateKeyInPool(header, newKeyInfo.Key, newKeyInfo.Ver)
+				UpdateKeyInPool(header, newKeyInfo)
 				keyInfo = newKeyInfo
 			} else {
 				mangaLogger.Warning("获取密钥失败，切换设备重试")
@@ -0,0 +1,212 @@
 package fqapi
 import (
 	"encoding/json"
 	"fmt"
 	"regexp"
 	"sort"
 	"strconv"
 	"strings"
 	"fanqie/internal/common"
 )
 // 漫画免签名兜底（对齐书源 20261012 的 wkbrowser 路线）：
 // 设备池签名路线全部失败时，直接从 m.wkbrowser.com 取章节图片，
 // 无需注册设备/签名/解密，返回原图直链。
 var (
 	mangaFallbackItemIDRe = regexp.MustCompile(`^\d+$`)
 	trailingNumberRe      = regexp.MustCompile(`(\d+)$`)
 )
 const mangaFallbackOrigin = "https://p6-novel.byteimg.com/origin/"
 // fetchMangaFallback 免签名获取漫画章节图片（wkbrowser 原图直链）
 func fetchMangaFallback(itemID string, showHTML bool) (interface{}, error) {
 	if !mangaFallbackItemIDRe.MatchString(itemID) {
 		return nil, fmt.Errorf("无效的 item_id: %s", itemID)
 	}
 	client := common.NewHTTPClient()
 	resp, err := client.Get("https://m.wkbrowser.com/i"+itemID+"/info/v2/", map[string]string{
 		"User-Agent": "Mozilla/5.0 (Linux; Android 14) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0 Mobile Safari/537.36",
 		"Accept":     "application/json, text/plain, */*",
 	})
 	if err != nil {
 		return nil, fmt.Errorf("wkbrowser 请求失败: %w", err)
 	}
 	var body map[string]interface{}
 	if err := json.Unmarshal(resp.Content, &body); err != nil {
 		return nil, fmt.Errorf("wkbrowser 响应解析失败: %w", err)
 	}
 	data, _ := body["data"].(map[string]interface{})
 	if data == nil {
 		return nil, fmt.Errorf("wkbrowser 响应缺少 data（code=%v）", body["code"])
 	}
 	// content 为 JSON 字符串或对象，内含 materials/skeleton 图片清单
 	payload := data["content"]
 	if s, ok := payload.(string); ok {
 		var p map[string]interface{}
 		if err := json.Unmarshal([]byte(s), &p); err != nil {
 			return nil, fmt.Errorf("wkbrowser 图片内容解析失败: %w", err)
 		}
 		payload = p
 	}
 	uris := collectMangaFallbackURIs(payload)
 	if len(uris) == 0 {
 		return nil, fmt.Errorf("wkbrowser 未找到图片")
 	}
 	images := make([]ImageResult, len(uris))
 	for i, u := range uris {
 		images[i] = ImageResult{Data: u, Format: "unknown", Index: i}
 	}
 	if showHTML {
 		var b strings.Builder
 		for _, img := range images {
 			fmt.Fprintf(&b, "<img src=\"%s\">\n", img.Data)
 		}
 		return map[string]interface{}{"content": b.String(), "source": "wkbrowser"}, nil
 	}
 	return map[string]interface{}{"images": images, "source": "wkbrowser"}, nil
 }
 // collectMangaFallbackURIs 按书源三段策略收集图片链接（保证页序）：
 // 1) skeleton.data 顺序引用 materials 的 key
 // 2) materials key 按末尾数字排序
 // 3) 递归遍历全树兜底（顺序不保证，仅防白屏）
 func collectMangaFallbackURIs(payload interface{}) []string {
 	p, _ := payload.(map[string]interface{})
 	if p == nil {
 		return nil
 	}
 	images := make([]string, 0, 32)
 	seen := map[string]bool{}
 	push := func(uri string) {
 		src := mangaFallbackImageURL(uri)
 		if src != "" && !seen[src] {
 			seen[src] = true
 			images = append(images, src)
 		}
 	}
 	materials, _ := p["materials"].(map[string]interface{})
 	skeletonWrapper, _ := p["skeleton"].(map[string]interface{})
 	skeleton, _ := skeletonWrapper["data"].([]interface{})
 	if materials != nil && skeleton != nil {
 		for _, s := range skeleton {
 			node, _ := s.(map[string]interface{})
 			if node == nil {
 				continue
 			}
 			key := ""
 			for _, k := range []string{"elementInfo", "key", "id"} {
 				if v, _ := node[k].(string); v != "" {
 					key = v
 					break
 				}
 			}
 			if m, ok := materials[key].(map[string]interface{}); ok {
 				if uri := mangaFallbackNodeURI(m); uri != "" {
 					push(uri)
 				}
 			}
 		}
 		if len(images) > 0 {
 			return images
 		}
 	}
 	if materials != nil {
 		keys := make([]string, 0, len(materials))
 		for k := range materials {
 			keys = append(keys, k)
 		}
 		sort.Slice(keys, func(i, j int) bool {
 			return trailingNumber(keys[i]) < trailingNumber(keys[j])
 		})
 		for _, k := range keys {
 			if m, ok := materials[k].(map[string]interface{}); ok {
 				if uri := mangaFallbackNodeURI(m); uri != "" {
 					push(uri)
 				}
 			}
 		}
 		if len(images) > 0 {
 			return images
 		}
 	}
 	var walk func(node interface{})
 	walk = func(node interface{}) {
 		switch t := node.(type) {
 		case []interface{}:
 			for _, item := range t {
 				walk(item)
 			}
 		case map[string]interface{}:
 			if uri := mangaFallbackNodeURI(t); uri != "" {
 				push(uri)
 				return
 			}
 			for _, v := range t {
 				switch v.(type) {
 				case map[string]interface{}, []interface{}:
 					walk(v)
 				}
 			}
 		}
 	}
 	walk(p)
 	return images
 }
 // mangaFallbackNodeURI 从 materials 节点取图片资源（web_uri 优先，其次 novel-pic/md5）
 func mangaFallbackNodeURI(node map[string]interface{}) string {
 	if d, ok := node["data"].(map[string]interface{}); ok {
 		if u, _ := d["web_uri"].(string); u != "" {
 			return u
 		}
 	}
 	if u, _ := node["web_uri"].(string); u != "" {
 		return u
 	}
 	if d, ok := node["data"].(map[string]interface{}); ok {
 		if md5, _ := d["md5"].(string); md5 != "" {
 			return "novel-pic/" + md5
 		}
 	}
 	return ""
 }
 // mangaFallbackImageURL web_uri 转可访问的原图直链
 func mangaFallbackImageURL(uri string) string {
 	u := strings.TrimPrefix(uri, "/")
 	if u == "" {
 		return ""
 	}
 	if strings.HasPrefix(u, "http://") {
 		return "https://" + strings.TrimPrefix(u, "http://")
 	}
 	if strings.HasPrefix(u, "https://") {
 		return u
 	}
 	return mangaFallbackOrigin + u
 }
 // trailingNumber 取字符串末尾数字（materials key 排序用），无数字排最后
 func trailingNumber(s string) int {
 	m := trailingNumberRe.FindStringSubmatch(s)
 	if m == nil {
 		return 1 << 30
 	}
 	n, _ := strconv.Atoi(m[1])
 	return n
 }
@@ -0,0 +1,44 @@
 package fqapi
 import "testing"
 // TestCollectMangaFallbackURIs skeleton 顺序优先、web_uri/md5 取值与原图直链拼接
 func TestCollectMangaFallbackURIs(t *testing.T) {
 	payload := map[string]interface{}{
 		"materials": map[string]interface{}{
 			"img_1": map[string]interface{}{"data": map[string]interface{}{"web_uri": "novel-pic/aaa"}},
 			"img_2": map[string]interface{}{"data": map[string]interface{}{"web_uri": "/novel-pic/bbb"}},
 			"img_3": map[string]interface{}{"data": map[string]interface{}{"md5": "c0ffee"}},
 		},
 		"skeleton": map[string]interface{}{
 			"data": []interface{}{
 				map[string]interface{}{"elementInfo": "img_2"},
 				map[string]interface{}{"key": "img_1"},
 				map[string]interface{}{"id": "img_3"},
 			},
 		},
 	}
 	// 页序以 skeleton 引用顺序为准
 	want := []string{
 		"https://p6-novel.byteimg.com/origin/novel-pic/bbb",
 		"https://p6-novel.byteimg.com/origin/novel-pic/aaa",
 		"https://p6-novel.byteimg.com/origin/novel-pic/c0ffee",
 	}
 	got := collectMangaFallbackURIs(payload)
 	if len(got) != len(want) {
 		t.Fatalf("got %d 张 want %d 张: %v", len(got), len(want), got)
 	}
 	for i := range want {
 		if got[i] != want[i] {
 			t.Errorf("[%d] got %s want %s", i, got[i], want[i])
 		}
 	}
 	// 无 skeleton 时退化为 materials key 末尾数字排序
 	delete(payload, "skeleton")
 	got = collectMangaFallbackURIs(payload)
 	if got[0] != want[1] || got[1] != want[0] {
 		t.Errorf("key 排序错误: %v", got)
 	}
 }
@@ -1,14 +1,9 @@
 package fqapi
 import (
 	"bytes"
 	"compress/zlib"
 	"encoding/base64"
 	"encoding/json"
 	"fanqie/internal/common"
 	"fmt"
 	"io"
 	"sort"
 	"strings"
 	"sync"
 )
@@ -17,57 +12,29 @@ var (
 	multiContentLogger = common.NewLogger("multi_content", "INFO", "")
 )
-// handleDecryptTask 处理解密任务
+// handleDecryptTask 处理解密任务（crypt_status 分支见 decodeContentBody）
 func handleDecryptTask(itemID string, itemData map[string]interface{}, keyInfo *KeyInfo, textMode, imageMode bool) (string, map[string]interface{}) {
 	if code, ok := itemData["code"].(float64); !ok || code != 0 {
 		return itemID, itemData
 	}
-	contentEncoded, ok := itemData["content"].(string)
+	if _, ok := itemData["content"].(string); !ok {
 	if !ok {
 		return itemID, itemData
 	}
-	// 解码content
+	// 解密（含压缩与明文分支）
-	contentBytes, err := base64.StdEncoding.DecodeString(contentEncoded)
+	contentStr, err := decodeContentBody(itemData, keyInfo)
-	if err != nil || len(contentBytes) < 16 {
+	if err != nil || contentStr == "" {
 		// 解密失败保留上游原样（ignore_failed 默认跳过失败章节）
 		return itemID, itemData
 	}
 	iv := contentBytes[0:16]
 	encryptedData := contentBytes[16:]
 	// 解密
 	if keyInfo == nil {
 		return itemID, itemData
 	}
 	keyBytes, _ := base64.StdEncoding.DecodeString(keyInfo.Key)
 	decrypted, err := De(encryptedData, keyBytes, iv)
 	if err != nil {
 		return itemID, itemData
 	}
 	// 检查是否需要解压
 	compressStatus, _ := itemData["compress_status"].(float64)
 	if compressStatus != 0 {
 		// zlib解压
 		reader, err := zlib.NewReader(bytes.NewReader(decrypted))
 		if err == nil {
 			decompressed, err := io.ReadAll(reader)
 			reader.Close()
 			if err == nil {
 				decrypted = decompressed
 			}
 		}
 	}
 	// 更新内容
 	bodyCopy := make(map[string]interface{})
 	for k, v := range itemData {
 		bodyCopy[k] = v
 	}
-	bodyCopy["content"] = string(decrypted)
+	bodyCopy["content"] = contentStr
 	// 缓存
 	cacheData := buildCacheData(bodyCopy)
@@ -75,7 +42,7 @@ func handleDecryptTask(itemID string, itemData map[string]interface{}, keyInfo *
 	// 解析内容
 	parser := NewContentParser()
-	bodyCopy["content"] = parser.Parse(string(decrypted), textMode, imageMode)
+	bodyCopy["content"] = parser.Parse(contentStr, textMode, imageMode)
 	return itemID, bodyCopy
 }
@@ -121,10 +88,11 @@ func MultiContent(itemIDs interface{}, bookID string, textMode, imageMode, ignor
 	currentKeyInfo := &KeyInfo{
 		Key: item.Key,
 		Ver: item.KeyVer,
 		Ts:  item.KeyTs,
 	}
 	multiContentLogger.Debug(fmt.Sprintf("使用池中设备 ID=%d (depth: %d)", item.ID, depth))
-	// 构建参数
+	// 构建参数（form 编码，见 buildSortedQuery 注释）
 	headerWithoutVersion := removeVersionFromHeader(header)
 	params := []string{
@@ -133,18 +101,15 @@ func MultiContent(itemIDs interface{}, bookID string, textMode, imageMode, ignor
 		"channel=0",
 		"device_platform=android",
 		"os_version=0",
-		"version_code=66.9",
+		"version_code=" + ClientVersionCode,
 		"book_id=" + bookID,
 		"item_ids=" + itemIDStr,
 		"novel_text_type=1",
 		"req_type=1",
 		fmt.Sprintf("key_register_ts=%d", currentKeyInfo.Ts),
 		headerWithoutVersion,
 	}
-
+	sortedParams := buildSortedQuery(params...)
 	allParams := strings.Join(params, "&")
 	paramList := strings.Split(allParams, "&")
 	sort.Strings(paramList)
 	sortedParams := strings.Join(paramList, "&")
 	url := "https://reading.snssdk.com/reading/reader/batch_full/v/?" + sortedParams
@@ -205,7 +170,7 @@ func MultiContent(itemIDs interface{}, bookID string, textMode, imageMode, ignor
 					multiContentLogger.Warning("密钥版本不匹配，重新获取密钥")
 					newKeyInfo, err := GetKey(header)
 					if err == nil {
-						UpdateKeyInPool(header, newKeyInfo.Key, newKeyInfo.Ver)
+						UpdateKeyInPool(header, newKeyInfo)
 					} else {
 						// 获取密钥失败，切换设备
 						multiContentLogger.Warning(fmt.Sprintf("获取密钥失败，切换设备重试 (depth: %d)", depth))
@@ -3,6 +3,7 @@ package fqapi
 import (
 	"fanqie/internal/common"
 	"fmt"
 	"sort"
 	"strings"
 	"sync"
 )
@@ -108,11 +109,41 @@ func chunkArray(arr []string, size int) [][]string {
 // HandleGetMultiDetail 处理批量获取书籍详情,详情数据基本不变,缓存 24 小时
 func HandleGetMultiDetail(bookID interface{}) (string, error) {
-	return withCache(fmt.Sprintf("multi_detail_%v", bookID), 86400, func() (string, error) {
+	// 缓存 key 归一化：ID 排序去重，相同集合不同顺序共享缓存
 	ids := normalizeIDList(bookID)
 	if ids == "" {
 		return fetchMultiDetail(bookID)
 	}
 	return withCache("multi_detail_"+ids, 86400, func() (string, error) {
 		return fetchMultiDetail(bookID)
 	})
 }
 // normalizeIDList 将 book_id 参数归一化为排序去重后的逗号串
 func normalizeIDList(bookID interface{}) string {
 	var list []string
 	switch v := bookID.(type) {
 	case string:
 		list = strings.Split(v, ",")
 	case []string:
 		list = v
 	default:
 		return ""
 	}
 	seen := make(map[string]bool, len(list))
 	out := make([]string, 0, len(list))
 	for _, id := range list {
 		id = strings.TrimSpace(id)
 		if id == "" || seen[id] {
 			continue
 		}
 		seen[id] = true
 		out = append(out, id)
 	}
 	sort.Strings(out)
 	return strings.Join(out, ",")
 }
 func fetchMultiDetail(bookID interface{}) (string, error) {
 	if bookID == nil {
 		return BuildErrorResponse("需要参数 book_id"), nil
@@ -0,0 +1,77 @@
 package fqapi
 import (
 	"fanqie/internal/common"
 	"fmt"
 	"net/url"
 )
 var rankLogger = common.NewLogger("rank", "INFO", "")
 // rankSDKHosts 免签名榜单接口的可用域名（依次容错）
 var rankSDKHosts = []string{
 	"https://novel.snssdk.com",
 	"https://api.fanqiesdk.com",
 	"https://api-lf.fanqiesdk.com",
 }
 const rankSDKUA = "com.ss.android.article.news/7.0.3 (Linux; U; Android 12; zh_CN; Pixel 6)"
 // RankTypes 榜单类型
 var RankTypes = map[int]string{
 	10: "推荐榜",
 	11: "完结榜",
 	12: "热搜榜",
 	13: "黑马榜",
 	15: "爆更榜",
 }
 // HandleGetRank 处理获取榜单,榜单数据动态,短缓存 10 分钟
 // sideType 见 RankTypes（无效值回退推荐榜）；page 从 1 起，每页 20 本
 func HandleGetRank(sideType, page int) (string, error) {
 	if _, ok := RankTypes[sideType]; !ok {
 		sideType = 10
 	}
 	if page < 1 {
 		page = 1
 	}
 	return withCache(fmt.Sprintf("rank_%d_%d", sideType, page), 600, func() (string, error) {
 		return fetchRank(sideType, page)
 	})
 }
 // fetchRank 免签名 SDK 接口获取榜单（无设备/签名依赖，多域名容错）
 func fetchRank(sideType, page int) (string, error) {
 	vals := url.Values{
 		"aid":       {"13"},
 		"type":      {"1"},
 		"side_type": {fmt.Sprintf("%d", sideType)},
 		"offset":    {fmt.Sprintf("%d", page-1)},
 		"limit":     {"20"},
 	}
 	path := "/api/novel/channel/homepage/rank/rank_list/v2/?" + vals.Encode()
 	var lastErr error
 	client := common.NewHTTPClient()
 	for _, host := range rankSDKHosts {
 		resp, err := client.Get(host+path, map[string]string{
 			"User-Agent": rankSDKUA,
 			"Accept":     "application/json, text/plain, */*",
 		})
 		if err != nil {
 			lastErr = err
 			continue
 		}
 		respJSON, err := resp.JSON()
 		if err != nil {
 			lastErr = fmt.Errorf("响应解析失败: %w", err)
 			continue
 		}
 		if code, ok := respJSON["code"].(float64); ok && code == 0 {
 			return reencodeUnwrapped(resp.Text), nil
 		}
 		lastErr = fmt.Errorf("code=%.0f msg=%v", respJSON["code"], respJSON["message"])
 	}
 	rankLogger.Error(fmt.Sprintf("获取榜单失败: %v", lastErr))
 	return BuildErrorResponse("获取榜单失败"), lastErr
 }
@@ -12,15 +12,13 @@ import (
 	"fmt"
 	"math/rand"
 	"regexp"
 	"sort"
 	"strings"
 	"time"
 )
 // KeyInfo 密钥信息
 type KeyInfo struct {
 	Key string `json:"key"`
 	Ver int    `json:"ver"`
 	Ts  int64  `json:"ts"` // 密钥注册时间戳（key_register_ts），随正文请求上传可降低密钥不匹配重试
 }
 var (
@@ -154,25 +152,22 @@ func GetKey(header string) (*KeyInfo, error) {
 	keyLogger.Info("正在获取密钥...")
-	// 构建参数
+	// 构建参数（form 编码：机型含空格等字符时 URL 与签名一致才不会注册失败）
 	params := []string{
 		"aid=1967",
 		"app_name=novelapp",
 		"channel=0",
 		"device_platform=android",
 		"os_version=0",
-		"version_code=5.8.9.32",
+		"version_code=" + ClientVersionCode,
 		header,
 	}
-
+	sortedParams := buildSortedQuery(params...)
 	allParams := strings.Join(params, "&")
 	paramList := strings.Split(allParams, "&")
 	sort.Strings(paramList)
 	sortedParams := strings.Join(paramList, "&")
 	// 准备请求体
 	requestBody := map[string]string{
 		"content": content,
 		"keyver":  "1",
 	}
 	bodyJSON, _ := json.Marshal(requestBody)
@@ -203,6 +198,12 @@ func GetKey(header string) (*KeyInfo, error) {
 		var respData map[string]interface{}
 		json.Unmarshal(resp.Content, &respData)
 		// 上游判定该设备无需解密（正文明文下发），返回空密钥信息由调用方分支处理
 		if code, ok := respData["code"].(float64); ok && int(code) == 500003 {
 			keyLogger.Info("上游标记无需解密（code=500003）")
 			return &KeyInfo{}, nil
 		}
 		dataObj, ok := respData["data"].(map[string]interface{})
 		if !ok {
 			return nil, fmt.Errorf("响应格式错误")
@@ -231,12 +232,17 @@ func GetKey(header string) (*KeyInfo, error) {
 		finalKey := base64.StdEncoding.EncodeToString(decryptedKey)
 		keyVer, _ := dataObj["keyver"].(float64)
 		keyTs, _ := dataObj["key_register_ts"].(float64)
 		if keyTs == 0 {
 			keyTs, _ = dataObj["register_time"].(float64)
 		}
 		keyInfo = &KeyInfo{
 			Key: finalKey,
 			Ver: int(keyVer),
 			Ts:  int64(keyTs),
 		}
-		keyLogger.Info(fmt.Sprintf("密钥获取成功！"))
+		keyLogger.Info("密钥获取成功！")
 		return keyInfo, nil
 	}
@@ -248,7 +254,3 @@ func GetKey(header string) (*KeyInfo, error) {
 func GetKeyInfo() *KeyInfo {
 	return keyInfo
 }
 func init() {
 	rand.Seed(time.Now().UnixNano())
 }
@@ -16,7 +16,7 @@ type RequestLogEntry struct {
 	IP         string `json:"ip"`
 	Method     string `json:"method"`
 	Path       string `json:"path"`
-	Query      string `json:"query"`     // 解码后查询串（含 key，不脱敏）
+	Query      string `json:"query"`     // 解码后查询串（key 参数值已脱敏）
 	Status     int    `json:"status"`
 	DurationMs int64  `json:"duration_ms"`
 	ErrorMsg   string `json:"error_msg"`  // 仅 status>=400 时截断保留
@@ -6,12 +6,22 @@ import (
 	"fanqie/internal/common"
 	"fmt"
 	"net/url"
 	"regexp"
 	"strconv"
 	"time"
 )
 var searchLogger = common.NewLogger("search", "INFO", "")
-// GetDeviceHeader 获取设备header（需要从外部注入）
+// 搜索分页与上游续传参数（对齐书源 20261012：上游翻页依赖上一页返回的
-var GetDeviceHeader func() string
+// passback/search_id/next_offset/corrected_query 续传，仅传数字 offset 深翻页会拿到重复/错序结果）
 const (
 	searchPageSize     = 10 // 对外每页书数
 	searchMaxUpstreams = 24 // 单次搜索累计上游翻页上限（对齐书源）
 )
 // bookIDKeyRe 纯数字 15~20 位关键词按书 ID 直搜详情
 var bookIDKeyRe = regexp.MustCompile(`^\d{15,20}$`)
 // HandleBookSearch 处理书籍搜索,搜索结果短缓存 10 分钟,减轻上游压力
 func HandleBookSearch(key string, offset int, tabType int) (string, error) {
@@ -24,21 +34,111 @@ func fetchBookSearch(key string, offset int, tabType int) (string, error) {
 	if key == "" {
 		return BuildErrorResponseWithCode(1, "需要参数 key"), nil
 	}
 	if tabType == 0 {
 		tabType = 3
 	}
 	if offset < 0 {
 		offset = 0
 	}
-	// 编码查询关键字
+	// 书 ID 直搜：跳过搜索接口直接查详情，精确命中单本
-	encodedKey := url.QueryEscape(key)
+	if bookIDKeyRe.MatchString(key) {
 		return idSearchResult(key)
 	}
-	// 使用新的参数构建函数
+	// 累计翻页直到凑够本页书：上游单页 count=20，按 book_id 去重后切出 offset 起的 10 本
-	reqURL := BuildURL("https://reading.snssdk.com/reading/bookapi/search/tab/v/",
+	session := strconv.FormatInt(time.Now().UnixNano(), 10)
-		fmt.Sprintf("tab_type=%d", tabType),
+	var (
-		"update_version_code=58932",
+		rows      []interface{} // 上游搜索行（含 book_data 包装），原样透传
-		"query="+encodedKey,
+		seen      = map[string]bool{}
-		fmt.Sprintf("passback=%d", offset),
+		passback  = "0"
 		searchID  = ""
 		corrected = ""
 		nextOff   = 0
 		hasMore   = true
 	)
 	want := offset + searchPageSize
 	for i := 0; i < searchMaxUpstreams && hasMore && len(rows) < want; i++ {
 		params := []string{
 			fmt.Sprintf("tab_type=%d", tabType),
 			"update_version_code=" + ClientVersionCode,
 			"query=" + url.QueryEscape(key),
 			"passback=" + passback,
 			fmt.Sprintf("offset=%d", nextOff),
 			"count=20",
 			fmt.Sprintf("is_first_enter_search=%t", i == 0),
 			"normal_session_id=" + session,
 			"cold_start_session_id=" + session,
 		}
 		if searchID != "" {
 			params = append(params, "search_id="+searchID)
 		}
 		if corrected != "" {
 			params = append(params, "corrected_query="+corrected)
 		}
 		tab, err := fetchSearchTab(params, tabType)
 		if err != nil {
 			if i == 0 {
 				return BuildErrorResponseWithCode(1, "搜索请求失败"), err
 			}
 			break // 已有数据时后续页失败，降级返回现有结果
 		}
 		if tab == nil {
 			break
 		}
 		if data, ok := tab["data"].([]interface{}); ok {
 			for _, r := range data {
 				row, _ := r.(map[string]interface{})
 				if row == nil {
 					continue
 				}
 				// 行 ID 取 book_data[0].book_id；无包装的行按行本身的 book_id
 				bd, _ := row["book_data"].([]interface{})
 				if len(bd) > 0 {
 					if b, ok := bd[0].(map[string]interface{}); ok {
 						if id := searchRowID(b); id != "" && !seen[id] {
 							seen[id] = true
 							rows = append(rows, row)
 						}
 						continue
 					}
 				}
 				if id := searchRowID(row); id != "" && !seen[id] {
 					seen[id] = true
 					rows = append(rows, row)
 				}
 			}
 		}
 		hasMore = searchHasMore(tab["has_more"])
 		nextOff = searchInt(tab["next_offset"])
 		passback = searchStr(tab["passback"])
 		searchID = searchStr(tab["search_id"])
 		corrected = searchStr(tab["corrected_query"])
 	}
 	start, end := offset, offset+searchPageSize
 	if start > len(rows) {
 		start = len(rows)
 	}
 	if end > len(rows) {
 		end = len(rows)
 	}
 	return marshalNoEscape(UnwrapJSONStrings(map[string]interface{}{
 		"code":        0,
 		"message":     "SUCCESS",
 		"data":        rows[start:end],
 		"has_more":    hasMore || end < len(rows),
 		"next_offset": offset + searchPageSize,
 	})), nil
 }
 // fetchSearchTab 请求一页搜索并返回目标 tab_type 的 tab；nil 表示该页无匹配 tab
 func fetchSearchTab(params []string, tabType int) (map[string]interface{}, error) {
 	reqURL := BuildURL("https://reading.snssdk.com/reading/bookapi/search/tab/v/", params...)
 	resp, err := GorgonRequest(reqURL, map[string]interface{}{
 		"method": "GET",
@@ -46,35 +146,61 @@ func fetchBookSearch(key string, offset int, tabType int) (string, error) {
 			"User-Agent": "com.dragon.read",
 		},
 	})
 	if err != nil {
 		searchLogger.Error(fmt.Sprintf("搜索请求失败: %v", err))
-		return BuildErrorResponseWithCode(1, "搜索请求失败"), err
+		return nil, err
 	}
-	respJSON, err := resp.JSON()
+	// UseNumber 重新解析后按 tab_type 过滤:
 	if err != nil {
 		searchLogger.Error(fmt.Sprintf("解析响应失败: %v", err))
 		return BuildErrorResponseWithCode(1, "解析响应失败"), err
 	}
 	if code, ok := respJSON["code"].(float64); ok && code == 0 {
 		// 用 UseNumber 重新解析后按 tab_type 过滤:
 	// 上游 search/tab/v/ 固定返回所有 tab 的 search_tabs;
 	// 大整数 ID 经 float64 重组会丢精度,必须 UseNumber
 	dec := json.NewDecoder(bytes.NewReader(resp.Content))
 	dec.UseNumber()
-		var numJSON map[string]interface{}
+	var pageJSON map[string]interface{}
-		if err := dec.Decode(&numJSON); err != nil {
+	if err := dec.Decode(&pageJSON); err != nil {
-			return resp.Text, nil // 解析失败时退回原样转发
+		searchLogger.Error(fmt.Sprintf("解析响应失败: %v", err))
 		return nil, fmt.Errorf("解析失败")
 	}
-		filterSearchTabs(numJSON, tabType)
+	if code, _ := pageJSON["code"].(float64); code != 0 {
-		unwrapSearchTab(numJSON)
+		return nil, fmt.Errorf("搜索失败 code=%v", pageJSON["code"])
 		UnwrapJSONStrings(numJSON)
 		return ToJSON(numJSON)
 	}
-	return BuildErrorResponseWithCode(1, "搜索失败"), nil
+	filterSearchTabs(pageJSON, tabType)
 	tabs, _ := pageJSON["search_tabs"].([]interface{})
 	for _, t := range tabs {
 		m, ok := t.(map[string]interface{})
 		if !ok {
 			continue
 		}
 		if tt, ok := m["tab_type"].(json.Number); ok {
 			if v, err := tt.Int64(); err == nil && int(v) == tabType {
 				return m, nil
 			}
 		}
 	}
 	return nil, nil
 }
 // idSearchResult 书 ID 直搜：复用详情接口（24h 缓存），包成上游搜索行形状（book_data 包装）
 func idSearchResult(bookID string) (string, error) {
 	detailJSON, err := HandleGetDetail(bookID)
 	if err != nil {
 		return BuildErrorResponseWithCode(1, "ID 搜索失败"), err
 	}
 	var detail map[string]interface{}
 	if err := json.Unmarshal([]byte(detailJSON), &detail); err != nil {
 		return detailJSON, nil // 透传上游错误响应
 	}
 	if code, _ := detail["code"].(float64); code != 0 {
 		return detailJSON, nil
 	}
 	return marshalNoEscape(UnwrapJSONStrings(map[string]interface{}{
 		"code":        0,
 		"message":     "SUCCESS",
 		"data":        []interface{}{map[string]interface{}{"book_data": []interface{}{detail["data"]}}},
 		"has_more":    false,
 		"next_offset": 0,
 	})), nil
 }
 // filterSearchTabs 按 tab_type 只保留 search_tabs 中匹配的项;无匹配时保留原样
@@ -100,26 +226,44 @@ func filterSearchTabs(respJSON map[string]interface{}, tabType int) {
 	}
 }
-// unwrapSearchTab 将过滤后的唯一 tab 展平：去掉 search_tabs 包装，
+// searchRowID 取行的 book_id（UseNumber 下为 json.Number，也可能是字符串）
-// tab 的 data 直接作为顶层 data，并提升分页字段 has_more/next_offset
+func searchRowID(m map[string]interface{}) string {
-func unwrapSearchTab(respJSON map[string]interface{}) {
+	if m == nil {
-	tabs, ok := respJSON["search_tabs"].([]interface{})
+		return ""
 	if !ok || len(tabs) != 1 {
 		return
 	}
-	tab, ok := tabs[0].(map[string]interface{})
+	if s, _ := m["book_id"].(string); s != "" {
-	if !ok {
+		return s
 		return
 	}
-	delete(respJSON, "search_tabs")
+	if n, ok := m["book_id"].(json.Number); ok {
-	delete(respJSON, "selected_tab_idx") // tab 已展平，索引无意义
+		return n.String()
 	delete(respJSON, "server_ab_params") // 上游 AB 实验参数，透传无意义
 	if data, ok := tab["data"]; ok {
 		respJSON["data"] = data
 	}
-	for _, k := range []string{"has_more", "next_offset"} {
+	return ""
 		if v, ok := tab[k]; ok {
 			respJSON[k] = v
 }
 func searchStr(v interface{}) string {
 	s, _ := v.(string)
 	return s
 }
 func searchInt(v interface{}) int {
 	switch t := v.(type) {
 	case json.Number:
 		n, _ := t.Int64()
 		return int(n)
 	case string:
 		n, _ := strconv.Atoi(t)
 		return n
 	}
 	return 0
 }
 func searchHasMore(v interface{}) bool {
 	switch t := v.(type) {
 	case bool:
 		return t
 	case json.Number:
 		n, _ := t.Int64()
 		return n != 0
 	}
 	return false
 }
@@ -0,0 +1,194 @@
 package fqapi
 import (
 	"fanqie/internal/common"
 	"fmt"
 	"strings"
 )
 var shelfLogger = common.NewLogger("shelf", "INFO", "")
 // HandleBookshelf 获取登录用户书架。个人动态数据不缓存。
 // group 非空时只返回该分组的书；书架接口对无效/过期 sessionid 也返回空书架，无法区分——
 // 空结果时提示重新登录
 func HandleBookshelf(sessionID, group string) (string, error) {
 	if sessionID == "" {
 		return BuildErrorResponse("需要参数 sessionid（在 fanqienovel.com 登录后从浏览器 Cookie 中获取）"), nil
 	}
 	resp, err := signedBookAPIGet("/reading/bookapi/bookshelf/info/v/", nil, sessionID)
 	if err != nil {
 		shelfLogger.Error(fmt.Sprintf("获取书架失败: %v", err))
 		return BuildErrorResponse("获取书架失败"), err
 	}
 	items, err := bookAPIList(resp, "book_shelf_info")
 	if err != nil {
 		return BuildErrorResponse("获取书架失败: " + err.Error()), nil
 	}
 	// 过滤分组并收集 book_id
 	type shelfRow struct {
 		id        string
 		groupName string
 	}
 	rows := make([]shelfRow, 0, len(items))
 	for _, it := range items {
 		m, ok := it.(map[string]interface{})
 		if !ok {
 			continue
 		}
 		id := bookIDOf(m)
 		if id == "" {
 			continue
 		}
 		gn, _ := m["group_name"].(string)
 		if group != "" && gn != group {
 			continue
 		}
 		rows = append(rows, shelfRow{id: id, groupName: gn})
 	}
 	if len(rows) == 0 {
 		return BuildErrorResponse("书架为空：sessionid 可能无效或已过期，请重新获取"), nil
 	}
 	ids := make([]string, len(rows))
 	for i, row := range rows {
 		ids[i] = row.id
 	}
 	details := detailMapByIDs(ids)
 	result := make([]map[string]interface{}, 0, len(rows))
 	for _, row := range rows {
 		d := details[row.id]
 		if d == nil {
 			d = map[string]interface{}{"book_id": row.id}
 		}
 		d["group_name"] = row.groupName
 		result = append(result, d)
 	}
 	return marshalNoEscape(UnwrapJSONStrings(buildCacheData(result))), nil
 }
 // HandleReadHistory 获取登录用户阅读历史。个人动态数据不缓存。
 // offset 从 0 起，每页 20 条；每项合并书籍详情并保留原始记录（进度等）于 history 字段
 func HandleReadHistory(sessionID string, offset int) (string, error) {
 	if sessionID == "" {
 		return BuildErrorResponse("需要参数 sessionid（在 fanqienovel.com 登录后从浏览器 Cookie 中获取）"), nil
 	}
 	if offset < 0 {
 		offset = 0
 	}
 	resp, err := signedBookAPIGet("/reading/bookapi/read_history/list/v/",
 		[]string{fmt.Sprintf("offset=%d", offset), "book_type=0"}, sessionID)
 	if err != nil {
 		shelfLogger.Error(fmt.Sprintf("获取阅读历史失败: %v", err))
 		return BuildErrorResponse("获取阅读历史失败"), err
 	}
 	items, err := bookAPIList(resp, "data_list")
 	if err != nil {
 		return BuildErrorResponse("获取阅读历史失败: " + err.Error()), nil
 	}
 	type historyRow struct {
 		id  string
 		raw map[string]interface{}
 	}
 	rows := make([]historyRow, 0, len(items))
 	for _, it := range items {
 		m, ok := it.(map[string]interface{})
 		if !ok {
 			continue
 		}
 		id := bookIDOf(m)
 		if id == "" {
 			continue
 		}
 		rows = append(rows, historyRow{id: id, raw: m})
 	}
 	if len(rows) == 0 {
 		return marshalNoEscape(buildCacheData([]interface{}{})), nil
 	}
 	ids := make([]string, len(rows))
 	for i, row := range rows {
 		ids[i] = row.id
 	}
 	details := detailMapByIDs(ids)
 	result := make([]map[string]interface{}, 0, len(rows))
 	for _, row := range rows {
 		d := details[row.id]
 		if d == nil {
 			d = map[string]interface{}{"book_id": row.id}
 		}
 		// 原始记录（含阅读进度/时间）挂 history 字段保留
 		d["history"] = row.raw
 		result = append(result, d)
 	}
 	return marshalNoEscape(UnwrapJSONStrings(buildCacheData(result))), nil
 }
 // signedBookAPIGet 个人书架/历史接口：常规签名请求 + 登录态 Cookie
 func signedBookAPIGet(path string, extraParams []string, sessionID string) (*common.ClientResponse, error) {
 	reqURL := BuildURL("https://reading.snssdk.com"+path, extraParams...)
 	return GorgonRequest(reqURL, map[string]interface{}{
 		"method": "GET",
 		"headers": map[string]string{
 			"User-Agent": "com.dragon.read",
 			"Cookie":     "sessionid=" + sessionID,
 		},
 	})
 }
 // bookAPIList 校验上游响应并取出 data 下的指定列表字段
 func bookAPIList(resp *common.ClientResponse, field string) ([]interface{}, error) {
 	respJSON, err := resp.JSON()
 	if err != nil {
 		return nil, fmt.Errorf("响应解析失败: %w", err)
 	}
 	if code, ok := respJSON["code"].(float64); !ok || code != 0 {
 		msg, _ := respJSON["message"].(string)
 		return nil, fmt.Errorf("code=%.0f msg=%s", respJSON["code"], msg)
 	}
 	data, _ := respJSON["data"].(map[string]interface{})
 	items, _ := data[field].([]interface{})
 	return items, nil
 }
 // bookIDOf 兼容提取 book_id / book_id_str（数字或字符串）
 func bookIDOf(m map[string]interface{}) string {
 	for _, k := range []string{"book_id_str", "book_id", "bookId"} {
 		switch v := m[k].(type) {
 		case string:
 			if v != "" {
 				return v
 			}
 		case float64:
 			return fmt.Sprintf("%.0f", v)
 		}
 	}
 	return ""
 }
 // detailMapByIDs 按 book_id 建立详情映射（复用批量详情接口，分批请求）
 func detailMapByIDs(ids []string) map[string]map[string]interface{} {
 	m := make(map[string]map[string]interface{}, len(ids))
 	for _, chunk := range chunkArray(ids, 100) {
 		list, err := multiDetail(strings.Join(chunk, ","))
 		if err != nil {
 			shelfLogger.Warning(fmt.Sprintf("批量获取详情失败（该批保留 book_id）: %v", err))
 			continue
 		}
 		for _, d := range list {
 			if id, ok := d["book_id"].(string); ok {
 				m[id] = d
 			}
 		}
 	}
 	return m
 }
@@ -103,3 +103,74 @@ func HandleMultiDetail(w http.ResponseWriter, r *http.Request) {
 	fmt.Fprint(w, result)
 }
 // HandleRank 榜单处理（推荐/完结/热搜/黑马/爆更榜，免签名接口）
 func HandleRank(w http.ResponseWriter, r *http.Request) {
 	incrementStat("total", "category")
 	w.Header().Set("Content-Type", "application/json; charset=utf-8")
 	sideType, err := parseIntParam(r.URL.Query().Get("side_type"), "side_type")
 	if err != nil {
 		respondError(w, err.Error(), true)
 		return
 	}
 	page, err := parseIntParam(r.URL.Query().Get("page"), "page")
 	if err != nil {
 		respondError(w, err.Error(), true)
 		return
 	}
 	result, err := fqapi.HandleGetRank(sideType, page)
 	if err != nil {
 		respondError(w, "获取榜单失败: "+err.Error(), false)
 		return
 	}
 	fmt.Fprint(w, result)
 }
 // HandleBookshelf 书架处理（需登录 sessionid）
 func HandleBookshelf(w http.ResponseWriter, r *http.Request) {
 	incrementStat("total", "category")
 	w.Header().Set("Content-Type", "application/json; charset=utf-8")
 	params, err := validateParams(r, "sessionid")
 	if err != nil {
 		respondError(w, err.Error(), true)
 		return
 	}
 	result, err := fqapi.HandleBookshelf(params["sessionid"], r.URL.Query().Get("group"))
 	if err != nil {
 		respondError(w, "获取书架失败: "+err.Error(), false)
 		return
 	}
 	fmt.Fprint(w, result)
 }
 // HandleReadHistory 阅读历史处理（需登录 sessionid）
 func HandleReadHistory(w http.ResponseWriter, r *http.Request) {
 	incrementStat("total", "category")
 	w.Header().Set("Content-Type", "application/json; charset=utf-8")
 	params, err := validateParams(r, "sessionid")
 	if err != nil {
 		respondError(w, err.Error(), true)
 		return
 	}
 	offset, err := parseIntParam(r.URL.Query().Get("offset"), "offset")
 	if err != nil {
 		respondError(w, err.Error(), true)
 		return
 	}
 	result, err := fqapi.HandleReadHistory(params["sessionid"], offset)
 	if err != nil {
 		respondError(w, "获取阅读历史失败: "+err.Error(), false)
 		return
 	}
 	fmt.Fprint(w, result)
 }
@@ -32,8 +32,10 @@ func HandleCategoryLanding(w http.ResponseWriter, r *http.Request) {
 	wordNumber := r.URL.Query().Get("word_number")
 	bookStatus := r.URL.Query().Get("book_status")
 	sortBy := r.URL.Query().Get("sort_by")
 	// 新版分类标签自带的筛选项（见 /category-front 返回的 category_landpage_url 参数）
 	selectedItems := r.URL.Query().Get("selected_items")
-	result, err := fqapi.HandleCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy)
+	result, err := fqapi.HandleCategoryLanding(categoryID, offset, wordNumber, bookStatus, sortBy, selectedItems)
 	if err != nil {
 		respondError(w, "获取分类详情失败: "+err.Error(), false)
 		return
@@ -270,7 +270,8 @@ func LogRequestMiddleware(handler http.HandlerFunc) http.HandlerFunc {
 	}
 }
-// decodedQuery 重组解码后的查询串（keyword=%E6%B5%8B → keyword=测），日志可读
+// decodedQuery 重组解码后的查询串（keyword=%E6%B5%8B → keyword=测），日志可读；
 // key 参数值脱敏（访问密钥不落日志）
 // ponytail: map 遍历乱序、%26 解码后与分隔符同名 —— 展示用途可接受
 func decodedQuery(r *http.Request) string {
 	vals := r.URL.Query() // 已解码
@@ -280,6 +281,9 @@ func decodedQuery(r *http.Request) string {
 	parts := make([]string, 0, len(vals))
 	for k, vs := range vals {
 		for _, v := range vs {
 			if strings.EqualFold(k, "key") && v != "" {
 				v = "***"
 			}
 			parts = append(parts, k+"="+v)
 		}
 	}
@@ -124,23 +124,17 @@ func addReviewToContent(jsonResult, itemID string, r *http.Request) string {
 		}
 	}
-	// 获取基础 URL
+	// 获取基础 URL（反代场景识别 X-Forwarded-Proto，避免生成 http:// 被 301）
 	cfg := config.GetConfig()
 	baseURL := ""
-	if cfg != nil {
+	if r.Host != "" {
 		baseURL = externalBaseURL(r)
 	} else if cfg != nil {
 		host := cfg.Server.Host
 		if host == "0.0.0.0" {
 			host = "localhost"
 		}
-		port := cfg.Server.Port
+		baseURL = "http://" + host + ":" + cfg.Server.Port
 		scheme := "http"
 		// 检查是否来自外部请求
 		if r.Host != "" {
 			baseURL = scheme + "://" + r.Host
 		} else {
 			baseURL = scheme + "://" + host + ":" + port
 		}
 	}
 	// 生成带段评标记的内容
@@ -12,6 +12,16 @@ import (
 	"time"
 )
 // externalBaseURL 取外部可访问的基础 URL（nginx 反代下识别 X-Forwarded-Proto，
 // 避免生成 http:// 链接被 301，弹窗 WebView 内 fetch 对 301 处理不稳）
 func externalBaseURL(r *http.Request) string {
 	scheme := "http"
 	if p := r.Header.Get("X-Forwarded-Proto"); p != "" {
 		scheme = p
 	}
 	return scheme + "://" + r.Host
 }
 // validateParams 验证必需参数
 func validateParams(r *http.Request, required ...string) (map[string]string, error) {
 	params := make(map[string]string)
@@ -207,7 +217,7 @@ func HandleReviewHTML(w http.ResponseWriter, r *http.Request) {
 	w.Header().Set("Content-Type", "text/html; charset=utf-8")
 	w.WriteHeader(http.StatusOK)
-	html := generateReviewHTML(data, params["item_id"], bookID, para, page, pageSize, sort)
+	html := generateReviewHTML(r, data, params["item_id"], bookID, para, page, pageSize, sort)
 	fmt.Fprint(w, html)
 }
@@ -318,6 +328,7 @@ var reviewTmpl = template.Must(template.New("review").Parse(reviewHTML))
 // reviewPageData 评论页模板数据
 type reviewPageData struct {
 	ItemID, BookID         string
 	BaseURL                string
 	ParaPlus1, Para, Sort  int
 	Total, NextOffset      int
 	PageSize, CommentCount int
@@ -327,7 +338,7 @@ type reviewPageData struct {
 }
 // generateReviewHTML 生成评论 HTML 页面
-func generateReviewHTML(data map[string]interface{}, itemID, bookID string, para, page, pageSize, sort int) string {
+func generateReviewHTML(r *http.Request, data map[string]interface{}, itemID, bookID string, para, page, pageSize, sort int) string {
 	total := 0
 	hasMore := false
 	nextOffset := 0
@@ -383,6 +394,7 @@ func generateReviewHTML(data map[string]interface{}, itemID, bookID string, para
 	var b strings.Builder
 	if err := reviewTmpl.Execute(&b, reviewPageData{
 		ItemID: itemID, BookID: bookID,
 		BaseURL:   externalBaseURL(r),
 		ParaPlus1: para + 1, Para: para, Sort: sort,
 		Total: total, NextOffset: nextOffset,
 		PageSize: pageSize, CommentCount: len(pageComments),
@@ -49,25 +49,50 @@
            box-sizing: border-box;
            font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif;
        }
        html, body { height: 100%; }
        body {
            background-color: var(--bg-primary);
            color: var(--text-primary);
            padding: 15px;
            max-width: 600px;
            margin: 0 auto;
            display: flex;
            flex-direction: column;
            overflow: hidden;
            transition: background-color 0.3s ease, color 0.3s ease;
        }
        .header {
            flex-shrink: 0;
            display: flex;
            justify-content: space-between;
            align-items: center;
-            margin-bottom: 20px;
+            padding: 10px 15px 12px;
            padding: 10px 15px;
            background: var(--bg-secondary);
-            border-radius: 12px;
+            border-radius: 20px 20px 0 0;
            border-bottom: 1px solid var(--border-color);
            box-shadow: 0 2px 8px var(--shadow);
            position: relative;
            z-index: 10;
            transition: background-color 0.3s ease;
        }
        /* BottomSheet 顶部拖拽把手 */
        .header::before {
            content: '';
            position: absolute;
            top: 6px;
            left: 50%;
            transform: translateX(-50%);
            width: 36px;
            height: 4px;
            border-radius: 2px;
            background: var(--border-color);
        }
        #comment-list {
            flex: 1;
            overflow-y: auto;
            -webkit-overflow-scrolling: touch;
            padding: 15px;
            overscroll-behavior: contain;
        }
        .header h1 {
            font-size: 18px;
            color: var(--text-primary);
@@ -333,8 +358,8 @@
            <i class="fas fa-book-open" style="color: var(--text-tertiary); font-size: 16px; margin: 0 10px;"></i>
        </span>
        <div class="sort-tabs">
-            <div class="sort-tab{{if eq .Sort 1}} active{{end}}" onclick="switchSort(1)">最热</div>
+            <div class="sort-tab{{if eq .Sort 1}} active{{end}}" onclick="switchSort(1, this)">最热</div>
-            <div class="sort-tab{{if eq .Sort 3}} active{{end}}" onclick="switchSort(3)">最新</div>
+            <div class="sort-tab{{if eq .Sort 3}} active{{end}}" onclick="switchSort(3, this)">最新</div>
        </div>
    </div>
    <div id="comment-list">
@@ -370,6 +395,8 @@
        const bookID = '{{.BookID}}';
        const para = {{.Para}};
        let currentSort = {{.Sort}};
        // 服务端注入的绝对地址：弹窗 WebView(loadDataWithBaseURL) 中相对路径解析不可靠
        const API_BASE = '{{.BaseURL}}';
        const commentList = document.getElementById('comment-list');
        const imageViewer = document.getElementById('imageViewer');
@@ -380,7 +407,7 @@
        // 已加载回复的评论
        const loadedReplies = new Set();
-        // URL 构建
+        // URL 构建（绝对地址，兼容弹窗 WebView）
        function buildReviewURL(params) {
            const baseParams = {
                item_id: itemID,
@@ -388,7 +415,7 @@
                para: para
            };
            const query = new URLSearchParams({...baseParams, ...params});
-            return '/review?' + query.toString();
+            return API_BASE + '/review?' + query.toString();
        }
        // 时间格式化
@@ -429,11 +456,11 @@
            };
        }
-        // 滚动信息获取
+        // 滚动信息获取（评论区为独立滚动容器）
        function getScrollInfo() {
-            const scrollTop = window.pageYOffset || document.documentElement.scrollTop;
+            const scrollTop = commentList.scrollTop;
-            const scrollHeight = document.documentElement.scrollHeight;
+            const scrollHeight = commentList.scrollHeight;
-            const clientHeight = document.documentElement.clientHeight;
+            const clientHeight = commentList.clientHeight;
            return {
                scrollTop,
                scrollHeight,
@@ -443,11 +470,11 @@
        }
        // 切换排序
-        async function switchSort(newSort) {
+        async function switchSort(newSort, tab) {
            if (currentSort === newSort) return;
-            document.querySelectorAll('.sort-tab').forEach(tab => tab.classList.remove('active'));
+            document.querySelectorAll('.sort-tab').forEach(t => t.classList.remove('active'));
-            event.target.closest('.sort-tab').classList.add('active');
+            if (tab) tab.classList.add('active');
            commentList.innerHTML = LOADING_HTML;
@@ -479,7 +506,7 @@
                    }
                    loadedReplies.clear();
-                    window.scrollTo({ top: 0, behavior: 'smooth' });
+                    commentList.scrollTo({ top: 0, behavior: 'smooth' });
                } else {
                    commentList.innerHTML = '<div class="empty">加载失败，请重试</div>';
                    commentIndex = 0;
@@ -751,13 +778,13 @@
        // 回到顶部
        function scrollToTop() {
-            window.scrollTo({
+            commentList.scrollTo({
                top: 0,
                behavior: 'smooth'
            });
        }
-        // 滚动监听
+        // 滚动监听（挂在评论容器上，顶栏固定不参与滚动）
        const handleScroll = debounce(() => {
            const scroll = getScrollInfo();
@@ -770,6 +797,6 @@
            }
        }, 100);
-        window.addEventListener('scroll', handleScroll, { passive: true });
+        commentList.addEventListener('scroll', handleScroll, { passive: true });
    </script>
 </body>
@@ -577,6 +577,7 @@ paths:
      summary: 获取漫画内容
      description: |
        返回漫画章节内容,图片经解密处理。
        设备池签名路线全部失败时自动降级 wkbrowser 免签名兜底(直接返回原图直链,响应附 `source: wkbrowser`)。
        输出模式 `output` 优先于配置 `fanqie.manga.output_mode`:
        - `url`: 返回原始图片 URL 数组(不可用:链接 404 且内容加密,仅作上游原始信息展示)
        - `proxy`: 返回本地代理 URL(/manga/image/{item_id}/{index})——推荐,先调本接口再访问代理 URL(缓存 5 分钟)
@@ -710,6 +711,10 @@ paths:
          in: query
          description: 排序方式,透传上游(0=默认)
          schema: { type: integer, default: 0 }
        - name: selected_items
          in: query
          description: 新版分类标签自带筛选项(从 /category-front 返回的 category_landpage_url 参数中获取),透传上游
          schema: { type: string }
      responses:
        "200":
          description: 分类书籍列表(上游转发)
@@ -721,6 +726,85 @@ paths:
          $ref: "#/components/responses/BadRequest"
        "500":
          $ref: "#/components/responses/ServerError"
  /rank:
    get:
      tags: [榜单与个人数据]
      summary: 获取书籍榜单
      description: 免签名接口,支持五种榜单,每页 20 本,短缓存 10 分钟。
      operationId: getRank
      parameters:
        - name: side_type
          in: query
          description: 榜单类型
          schema:
            type: integer
            default: 10
            enum: [10, 11, 12, 13, 15]
            x-enum-varnames: [推荐榜, 完结榜, 热搜榜, 黑马榜, 爆更榜]
        - name: page
          in: query
          description: 页码,从 1 起
          schema: { type: integer, default: 1, minimum: 1 }
      responses:
        "200":
          description: 榜单书籍列表(上游转发,data.result 为书籍数组)
          content:
            application/json:
              schema: { $ref: "#/components/schemas/UpstreamResponse" }
        "500":
          $ref: "#/components/responses/ServerError"
  /bookshelf:
    get:
      tags: [榜单与个人数据]
      summary: 获取我的书架
      description: 需要番茄网页版登录态;每本书合并详情数据并附带 group_name。个人动态数据不缓存。
      operationId: getBookshelf
      parameters:
        - name: sessionid
          in: query
          required: true
          description: 番茄网页版(fanqienovel.com)登录后的 Cookie sessionid
          schema: { type: string }
        - name: group
          in: query
          description: 按分组过滤,为空返回全部
          schema: { type: string }
      responses:
        "200":
          description: 书架书籍列表(data 为数组,每项含书籍详情 + group_name)
          content:
            application/json:
              schema: { $ref: "#/components/schemas/UpstreamResponse" }
        "400":
          $ref: "#/components/responses/BadRequest"
        "500":
          $ref: "#/components/responses/ServerError"
  /read-history:
    get:
      tags: [榜单与个人数据]
      summary: 获取阅读历史
      description: 需要番茄网页版登录态;每页 20 条,每项合并书籍详情并保留原始记录(含阅读进度)于 history 字段。个人动态数据不缓存。
      operationId: getReadHistory
      parameters:
        - name: sessionid
          in: query
          required: true
          description: 番茄网页版(fanqienovel.com)登录后的 Cookie sessionid
          schema: { type: string }
        - name: offset
          in: query
          description: 偏移量,默认 0,每页 20 条
          schema: { type: integer, default: 0, minimum: 0 }
      responses:
        "200":
          description: 阅读历史列表(data 为数组)
          content:
            application/json:
              schema: { $ref: "#/components/schemas/UpstreamResponse" }
        "400":
          $ref: "#/components/responses/BadRequest"
        "500":
          $ref: "#/components/responses/ServerError"
  /emoji/{file}:
    get:
      tags: [静态资源]
Author	SHA1	Message	Date
chenxin	9aec09bb40	✨ feat(manga,device): 漫画免签名兜底与 phone 注册身份选项 - /manga 设备池签名路线全部失败时自动降级 wkbrowser 免签名路线（materials/skeleton 三段策略收集原图直链，无需设备/签名/解密），响应附 source: wkbrowser - 新增配置 fanqie.device_identity（默认 pad 保持现状；phone 对齐书源真机注册 payload：host_bit 64 + 补 hardware/device_device/fingerprint 字段） - README 与 openapi.yaml 同步文档，新增 manga_fallback 解析测试	2026-10-12 14:53:08 +08:00
chenxin	84bd3238ab	✨ feat(api): 新增榜单书架历史与搜索续传并重构正文缓存链路 - 新增 /rank 免签名榜单接口（多域名容错）、/bookshelf 与 /read-history（sessionid 登录态，书架支持分组过滤） - 搜索重写：翻页透传上游 passback/search_id/next_offset/corrected_query 并按 book_id 去重，15~20 位纯数字关键词直查详情（ID 直搜） - 删除 /login/check 登录校验端点（login.go、路由与文档同步移除，sessionid 说明改为直接供书架/历史使用） - 正文链路重构：singleflight 合并同章节并发回源防击穿，缓存与解析解耦统一在出口处理，decodeContentBody 按 crypt_status 分支并兜底 500003 明文 - 缓存层：SQL 持久层 TTL 与业务 TTL 解耦（不可变内容过期后由 SQL 回填），动态短缓存只走 Redis，withCache 合并并发 miss - 密钥池：KeyTs 随正文请求上传降低密钥不匹配重试，密钥超 24 小时异步刷新并防刷新风暴，last_used_at 批量落库 - 签名补 X-Helios（与 X-Ladon 同值，对齐书源实现）；设备注册版本号统一走 ClientVersionCode/Name - 分类 landing 透传新版标签 selected_items；multi-detail 批量详情供书架/历史复用 - README 与 openapi.yaml 补充新接口文档	2026-10-12 14:49:14 +08:00
chenxin	8ca372050b	✨ feat(bookSource): 聚合书源适配阅读 app 并完善评论体验 - 聚合书源取代早期书源版本：发现页筛选宫格/分类宫格、x:t:m:d: 前缀搜索、 fqtab 类型随书切换、段评气泡(内置六种样式可配)；showCmt 能力检测，魔改版原地弹窗(点遮罩关闭)、官方阅读降级 startBrowser；jsLib 不引用规则变量， searchUrl 修正编码与作用域 - 评论页：注入绝对 API_BASE，修复弹窗 WebView 内切换排序/懒加载失败； externalBaseURL 识别 X-Forwarded-Proto，气泡 ident 不再硬编码 http；顶栏固定(拖拽把手)+评论区独立滚动 - 部署：证书方案从 certbot 迁移到 acme.sh(CERT_DIR 可配置)；新增服务生命周期参数 --stop/--start/--no-restart/--no-autostart；--clean 一键清理全部部署内容(服务/nginx/二进制/证书，含 certbot 残留)，acme.sh 定位抽为 find_acme_sh 供 init/clean 共用 - 忽略 .codegraph 本地索引目录	2026-10-10 13:36:25 +08:00
chenxin	49d0a371fb	✨ feat(video): 新增视频 CENC 解密与流式代理播放 - 逆向 SpadeA(黑桃A) PlayAuth 混淆，从 spade_a 还原 AES-128 内容密钥 - 实现 cenc-aes-ctr MP4 解密（senc 逐样本 CTR + stsd entry 修复），实测可播放 - 新增 /video/proxy/{key} 代理：稳定 key + 磁盘缓存（二次命中提速 130x）+ http.ServeContent 支持 Range/206/seek 流式播放 - /video 接口新增 output(url/proxy，默认 proxy) 与 quality(max/720p/档位key) 参数，可由 config fanqie.video.* 配置默认值；无匹配清晰度时明确报错 - 代理映射复用 audio_cache 存储，抽取 loadMediaProxyEntry 共用；挂接 7 天缓存清理 - deploy.sh 新增 --sync-nginx：备份→重写→校验→失败自动回滚，同步视频代理段配置	2026-10-10 12:58:25 +08:00