tt-help-cli-ycl 1.4.142 → 1.4.145

package/src/lib/api-interceptor.js

@@ -172,6 +172,8 @@ async function attemptItemListFetch(page, username, log, apiRequestUrlRef) {
   let responsePromise = null;
   let response = null;
   let rawBodyBytes = -1; // -1 未知（未读取到 body）；0 = 服务端空响应（封禁判定依据）
+  let httpStatus = null; // 响应 HTTP 状态码
+  let bodyText = ""; // 响应体文本（识别 ratelimit triggered 等限流文案）
 
   try {
     // 响应监听必须在 goto 之前注册：domcontentloaded 返回时首个 item_list
@@ -208,6 +210,7 @@ async function attemptItemListFetch(page, username, log, apiRequestUrlRef) {
     // JSON 解析失败（如 "Unexpected end of JSON input"）时打印原始响应，
     // 便于诊断空响应/截断响应/非 JSON 响应等异常形态
     if (response) {
+      httpStatus = response.status();
       try {
         // body 读取同样可能因页面异常而挂起，加独立超时保护
         const raw = await Promise.race([
@@ -218,6 +221,7 @@ async function attemptItemListFetch(page, username, log, apiRequestUrlRef) {
         ]);
         const text = raw.toString("utf-8");
         rawBodyBytes = raw.length;
+        bodyText = text;
         // content-length 用于区分两种 0 字节形态：
         // 声明长度 > 实际字节 = 传输被截断；声明缺失/为 0 = 服务端主动空响应
         const declared = parseInt(
@@ -247,6 +251,8 @@ async function attemptItemListFetch(page, username, log, apiRequestUrlRef) {
     data,
     interceptionError,
     rawBodyBytes,
+    httpStatus,
+    bodyText,
     elapsed: Date.now() - t0,
   };
 }
@@ -343,8 +349,14 @@ async function fetchUserVideosAPI(
 
   while (true) {
     attempt++;
-    const { data, interceptionError, rawBodyBytes, elapsed } =
-      await attemptItemListFetch(page, username, log, apiRequestUrlRef);
+    const {
+      data,
+      interceptionError,
+      rawBodyBytes,
+      httpStatus,
+      bodyText,
+      elapsed,
+    } = await attemptItemListFetch(page, username, log, apiRequestUrlRef);
 
     if (data && Array.isArray(data.itemList) && data.itemList.length > 0) {
       log(`  [API拦截] 成功，耗时 ${elapsed}ms`);
@@ -359,21 +371,27 @@ async function fetchUserVideosAPI(
       );
     }
 
-    // 0 字节空响应 = 被封（content-length 缺失/为 0 或截断，统一按封禁处理）。
-    // 不再依赖页面"出错了/很抱歉"文案判定，静默型封禁（页面正常渲染但 API 空响应）也能识别。
-    // 默认 30/40/50 秒等待重试 3 次，仍封才抛错交给上层（explore/refresh/refresh-mini 均受益）
-    if (rawBodyBytes === 0) {
+    // 判定为被封的两种形态，统一按封禁处理（默认 30/40/50 秒重试 3 次，仍封才抛错交给上层）：
+    // 1. 0 字节空响应（content-length 缺失/为 0 或截断）—— 静默型封禁（页面正常渲染但 API 空响应）
+    // 2. HTTP 429 或 body 含 ratelimit —— 服务端主动限流（ratelimit triggered）
+    // 两种都不再依赖页面"出错了/很抱歉"文案判定，explore/refresh/refresh-mini 均受益
+    const rateLimited = httpStatus === 429 || /ratelimit/i.test(bodyText || "");
+    if (rawBodyBytes === 0 || rateLimited) {
+      const reason =
+        rawBodyBytes === 0
+          ? "空响应（0 字节）"
+          : `限流（HTTP ${httpStatus}, ${bodyText.slice(0, 60)}）`;
       const retryIndex = attempt - 1;
       if (retryIndex < retryWaits.length) {
         const waitMs = retryWaits[retryIndex];
         log(
-          `  [API拦截] item_list 空响应（0 字节），判定为被封，${waitMs / 1000}s 后第 ${attempt}/${retryWaits.length} 次重试...`,
+          `  [API拦截] item_list ${reason}，判定为被封，${waitMs / 1000}s 后第 ${attempt}/${retryWaits.length} 次重试...`,
         );
         await new Promise((r) => setTimeout(r, waitMs));
         continue;
       }
       log(
-        `  [API拦截] item_list 空响应（0 字节），重试 ${retryWaits.length} 次仍被封`,
+        `  [API拦截] item_list ${reason}，重试 ${retryWaits.length} 次仍被封`,
       );
       throw new Error(`被封: ${username}`);
     }

package/src/lib/app-config.js

@@ -0,0 +1,119 @@
+/**
+ * 通用应用配置模块（key-value，落盘到 ~/.tt-help-config.json）
+ *
+ * 设计为通用命名：任意配置项以 `key` 形式存放，后续可加入其他配置。
+ * 例如当前使用 `targetLocations`，未来可扩展 `xxx` 等。
+ *
+ * 读取规则：文件存在则用文件内容，否则回退内置默认值。
+ * 写入规则：整份 JSON 落盘（原子性不强，但配置变更频率低，可接受）。
+ */
+import { existsSync, readFileSync, writeFileSync } from "fs";
+import { homedir } from "os";
+import { join } from "path";
+import { DEFAULT_TARGET_LOCATIONS } from "./target-locations.js";
+
+// 通用配置文件路径（与 ~/.tt-help.json 隔离，避免与看门狗 clientConfig 写入冲突）
+const CONFIG_FILE = join(homedir(), ".tt-help-config.json");
+
+let _cache = null;
+
+export function getConfigFilePath() {
+  return CONFIG_FILE;
+}
+
+function readConfig() {
+  if (_cache) return _cache;
+  try {
+    if (existsSync(CONFIG_FILE)) {
+      const parsed = JSON.parse(readFileSync(CONFIG_FILE, "utf-8"));
+      _cache = parsed && typeof parsed === "object" ? parsed : {};
+    } else {
+      _cache = {};
+    }
+  } catch {
+    _cache = {};
+  }
+  return _cache;
+}
+
+/**
+ * 通用读取：getConfig() 返回整份配置；getConfig(key) 返回指定项。
+ * @param {string|null} key
+ * @param {*} fallback
+ */
+export function getConfig(key = null, fallback = null) {
+  const cfg = readConfig();
+  if (key == null) return { ...cfg };
+  return cfg[key] !== undefined ? cfg[key] : fallback;
+}
+
+/**
+ * 通用写入：设置指定 key 并落盘，返回更新后的整份配置。
+ * @param {string} key
+ * @param {*} value
+ */
+export function setConfig(key, value) {
+  const cfg = readConfig();
+  cfg[key] = value;
+  _cache = cfg;
+  try {
+    writeFileSync(CONFIG_FILE, JSON.stringify(cfg, null, 2), "utf-8");
+  } catch {
+    // 写失败不抛出，保证调用方不至于崩溃；仅保留内存值
+  }
+  return { ...cfg };
+}
+
+/**
+ * 通用删除：移除指定 key 并落盘，返回更新后的整份配置。
+ * @param {string} key
+ */
+export function removeConfig(key) {
+  const cfg = readConfig();
+  if (key in cfg) {
+    delete cfg[key];
+    _cache = cfg;
+    try {
+      writeFileSync(CONFIG_FILE, JSON.stringify(cfg, null, 2), "utf-8");
+    } catch {
+      // 写失败忽略，仅保留内存值
+    }
+  }
+  return { ...cfg };
+}
+
+function normalizeLocations(locations) {
+  if (!Array.isArray(locations)) return [];
+  return [
+    ...new Set(
+      locations.map((l) => String(l).trim().toUpperCase()).filter(Boolean),
+    ),
+  ];
+}
+
+/** 读取当前目标国家，未配置则回退 DEFAULT_TARGET_LOCATIONS */
+export function getTargetLocations() {
+  const raw = getConfig("targetLocations", []);
+  const normalized = normalizeLocations(raw);
+  return normalized.length > 0 ? normalized : [...DEFAULT_TARGET_LOCATIONS];
+}
+
+/**
+ * 保存目标国家（自动去重、大写）。
+ * 传入空数组/非法值 = 移除配置项，回退到内置默认列表。
+ * 始终返回"生效后的列表"（getTargetLocations()），保证响应与读取一致。
+ */
+export function setTargetLocations(locations) {
+  const normalized = normalizeLocations(locations);
+  if (normalized.length === 0) {
+    removeConfig("targetLocations");
+  } else {
+    setConfig("targetLocations", normalized);
+  }
+  return getTargetLocations();
+}
+
+/** 清除内存缓存，强制下次读盘（配置被外部修改时调用） */
+export function reloadConfig() {
+  _cache = null;
+}

package/src/lib/args.js

@@ -1,6 +1,8 @@
 import { readFileSync } from "fs";
 import { server as defaultServer } from "./constants.js";
-import { DEFAULT_TARGET_LOCATIONS_CSV } from "./target-locations.js";
+import { getTargetLocations } from "./app-config.js";
+
+const DEFAULT_TARGET_LOCATIONS_CSV = getTargetLocations().join(",");
 
 const PRESETS = ["fast", "normal", "slow", "stealth"];
 

package/src/lib/constants.js

@@ -1,7 +1,9 @@
 import { join, dirname } from "path";
 import { readFileSync, writeFileSync, existsSync } from "fs";
 import { fileURLToPath } from "url";
-import { DEFAULT_TARGET_LOCATIONS_CSV } from "./target-locations.js";
+import { getTargetLocations } from "./app-config.js";
+
+const DEFAULT_TARGET_LOCATIONS_CSV = getTargetLocations().join(",");
 
 const __filename = fileURLToPath(import.meta.url);
 const __dirname = dirname(__filename);
@@ -358,15 +360,16 @@ const HELP_TEXT = [
   "    选项:",
   "      --countries <CSV>         目标国家，逗号分隔（默认 13 个欧洲国家）",
   "      -s, --server <URL>        服务端地址（默认 http://127.0.0.1:3000）",
-  "      --port <N>                起始 CDP 端口，端口池默认为 N~N+9（默认 7222）",
+  "      --port <N>                固定 CDP 端口（默认 7222，全程不切换）",
   "      --discover                开启自动发现：剩余 new 标签 < 5 时后台异步生成，连续 3 轮空兜底",
   "      --discover-count <N>      每次发现每个国家生成标签数（默认 10，上限 20）",
   "    行为:",
-  "      连续 3 次 HTTP 403 → 判断账号被封 → 等 2 分钟后自动切换到池中另一端口",
+  "      错误只打印不自愈：403 连续 3 次 → 上报服务端并触发 daemon 换节点；",
+  "      网络错误/其他错误连续 11 次 → 退出；验证码连续 4 次未人工解决 → 退出",
+  "      成功一个 tag 全部计数清零；浏览器窗口被手动关闭 → 自动退出",
   "    环境变量:",
-  "      TAG_SCOREALL_PORT_POOL_SIZE=N  端口池大小（默认 10）",
-  "      TAG_SCOREALL_RECYCLE_RSS_MB=N  内存回收阈值 RSS MB（默认 2000）",
-  "      TAG_SCOREALL_RECYCLE_HEAP_MB=N 内存回收阈值 Heap MB（默认 2000）",
+  "      TAG_SCOREALL_MIN_COUNTRY_RATE=0.8  国家识别率下限（低于则重试）",
+  "      TAG_SCOREALL_COUNTRY_MAX_RETRIES=2 国家识别率不足最大重试次数",
   "    示例: tt-help tag score-all",
   "          tt-help tag score-all --countries ES -s http://127.0.0.1:3001",
   "          tt-help tag score-all --port 9222 -s http://127.0.0.1:3001 --discover",

package/src/lib/target-locations.js

@@ -12,6 +12,7 @@ const DEFAULT_TARGET_LOCATIONS = [
   "IT",
   "IE",
   "AT",
+  "TH",
   "US",
 ];
 

package/src/scraper/core.js

@@ -11,17 +11,17 @@ import {
 import { extractCommentAuthors } from "./modules/comment-extractor.js";
 import { extractGuessVideos } from "./modules/guess-extractor.js";
 import {
-  DEFAULT_TARGET_LOCATIONS_CSV,
   isLocationInList,
   normalizeLocationList,
 } from "../lib/target-locations.js";
+import { getTargetLocations } from "../lib/app-config.js";
 
 async function scrapeSingleVideo(
   page,
   maxComments,
   maxGuess,
   log,
-  location = DEFAULT_TARGET_LOCATIONS_CSV,
+  location = getTargetLocations().join(","),
 ) {
   const config = getDelayConfig();
 

package/src/scraper/explore-core.js

@@ -5,12 +5,12 @@ import { getUserInfo, collectVideos } from "../videos/core.js";
 import { extractFollowAndFollowers } from "./modules/follow-extractor.js";
 import { extractVideoLocation, setScraperProxy } from "../lib/scrape.js";
 import {
-  DEFAULT_TARGET_LOCATIONS_CSV,
   findBestMatchingLocation,
   isLocationInList,
   normalizeLocation,
   normalizeLocationList,
 } from "../lib/target-locations.js";
+import { getTargetLocations } from "../lib/app-config.js";
 import {
   maxFollowing as globalMaxFollowing,
   maxFollowers as globalMaxFollowers,
@@ -23,7 +23,7 @@ async function processExplore(page, username, options, log) {
     loggedIn = false, // 由外部传入登录状态，避免每次调用 isLoggedIn(page)
     maxFollowing = 50,
     maxFollowers = 50,
-    location = DEFAULT_TARGET_LOCATIONS_CSV,
+    location = getTargetLocations().join(","),
     locationMode = "explore", // "explore" | "refresh"
     proxyServer = null,
     webserverUrl = null,

package/src/watch/data-store.js

@@ -19,10 +19,8 @@
 import fs from "fs";
 import path from "path";
 import Database from "better-sqlite3";
-import {
-  isLocationInList,
-  DEFAULT_TARGET_LOCATIONS,
-} from "../lib/target-locations.js";
+import { isLocationInList } from "../lib/target-locations.js";
+import { getTargetLocations } from "../lib/app-config.js";
 
 // Schema 与连接管理
 import {
@@ -610,6 +608,92 @@ function getTargetUsersByCountryFromDb(targetLocations = [], options = {}) {
   };
 }
 
+/**
+ * 精确商家：目标商家（tt_seller=1 且 verified=0 且 目标国家）中，
+ * 最近 7 天有 isECVideo 播放 > 1000 的视频，且有商品、商品评论数=0、销量 > 10。
+ * 支持分页与搜索（unique_id/nickname），返回 total + users。
+ */
+function getPreciseMerchantsFromDb(targetLocations = [], options = {}) {
+  if (!getDb()) return null;
+  if (!targetLocations.length) return { total: 0, users: [] };
+
+  const { limit, offset, search } = options;
+  const placeholders = targetLocations.map(() => "?").join(", ");
+
+  const where = `
+    tt_seller = 1
+    AND verified = 0
+    AND location_created IN (${placeholders})
+    AND top_video_seo_url IS NOT NULL AND top_video_seo_url != ''
+    AND product_review_count = 0
+    AND product_sold_count > 10
+    AND recent_ec_videos IS NOT NULL AND recent_ec_videos != ''
+    AND EXISTS (
+      SELECT 1 FROM json_each(recent_ec_videos)
+      WHERE CAST(json_extract(value, '$.playCount') AS INTEGER) > 1000
+    )
+  `;
+
+  let searchSql = "";
+  const searchParams = [];
+  if (search) {
+    searchSql = ` AND (unique_id LIKE ? OR nickname LIKE ?)`;
+    searchParams.push(`%${search}%`, `%${search}%`);
+  }
+
+  const totalRow = getDb()
+    .prepare(`SELECT COUNT(*) AS n FROM jobs WHERE ${where}${searchSql}`)
+    .get(...targetLocations, ...searchParams);
+  const total = totalRow?.n || 0;
+
+  const selectSql = `
+    SELECT
+      unique_id,
+      nickname,
+      sec_uid,
+      status,
+      follower_count,
+      video_count,
+      location_created,
+      latest_video_time,
+      refresh_time,
+      top_video_play_count,
+      top_video_href,
+      top_video_is_ad,
+      top_video_product_id,
+      top_video_product_title,
+      top_video_seo_url,
+      product_rating,
+      product_review_count,
+      product_sold_count,
+      product_name,
+      recent_ec_videos,
+      (SELECT COUNT(*) FROM json_each(recent_ec_videos)
+        WHERE CAST(json_extract(value, '$.playCount') AS INTEGER) > 1000) AS ecOver1k,
+      (SELECT MAX(CAST(json_extract(value, '$.playCount') AS INTEGER))
+        FROM json_each(recent_ec_videos)) AS maxEcPlay
+    FROM jobs
+    WHERE ${where}${searchSql}
+    ORDER BY product_sold_count DESC, unique_id ASC
+  `;
+
+  if (limit !== undefined) {
+    const safeLimit = Math.max(Math.min(Math.floor(limit), 1000), 1);
+    const safeOffset = Math.max(Math.floor(offset || 0), 0);
+    const rows = getDb()
+      .prepare(`${selectSql} LIMIT ? OFFSET ?`)
+      .all(...targetLocations, ...searchParams, safeLimit, safeOffset)
+      .map(mapJobRow);
+    return { total, limit: safeLimit, offset: safeOffset, users: rows };
+  }
+
+  const rows = getDb()
+    .prepare(selectSql)
+    .all(...targetLocations, ...searchParams)
+    .map(mapJobRow);
+  return { total, users: rows };
+}
+
 /**
  * 查询已修改（modified_at IS NOT NULL）且 location_created 不在 targetLocations 中的目标商家，
  * 按国家分组返回。
@@ -665,7 +749,7 @@ function getModifiedTargetUsersOutsideLocations(targetLocations = []) {
 function getTargetRefreshStats(hours = 24) {
   if (!getDb()) return { countries: [], total: 0, refreshed: 0 };
 
-  const targetLocations = DEFAULT_TARGET_LOCATIONS;
+  const targetLocations = getTargetLocations();
   const placeholders = targetLocations.map(() => "?").join(", ");
 
   // 按国家统计总数
@@ -2226,7 +2310,7 @@ export function createStore(filePath, options = {}) {
       const now = Date.now();
       const threshold = now - maxAgeSeconds * 1000;
       const defaultTime = new Date("2016-01-01T00:00:00Z").getTime();
-      const targetLocations = locations || DEFAULT_TARGET_LOCATIONS;
+      const targetLocations = locations || getTargetLocations();
       const placeholders = targetLocations.map(() => "?").join(",");
       const row = getDb()
         .prepare(
@@ -2261,7 +2345,7 @@ export function createStore(filePath, options = {}) {
     const defaultTime = new Date("2016-01-01T00:00:00Z").getTime();
 
     // 筛选目标国家用户，按 refreshTime 升序取最远的（没有则默认 2016-01-01）
-    const targetLocations = locations || DEFAULT_TARGET_LOCATIONS;
+    const targetLocations = locations || getTargetLocations();
     const targetUsers = data.filter(
       (u) =>
         u.ttSeller &&
@@ -2405,7 +2489,7 @@ export function createStore(filePath, options = {}) {
     const now = Date.now();
     const threshold = now - intervalSeconds * 1000;
     const defaultTime = new Date("2016-01-01T00:00:00Z").getTime();
-    const targetLocations = locations || DEFAULT_TARGET_LOCATIONS;
+    const targetLocations = locations || getTargetLocations();
     const placeholders = targetLocations.map(() => "?").join(",");
 
     // SELECT 候选用户
@@ -3551,6 +3635,8 @@ export function createStore(filePath, options = {}) {
     getRawJobsPage: getRawJobsPageFromDb,
     getTargetUsers: getTargetUsersFromDb,
     getTargetUsersByCountry: getTargetUsersByCountryFromDb,
+    getPreciseMerchants: (targetLocations, options) =>
+      getPreciseMerchantsFromDb(targetLocations, options),
     getModifiedTargetUsersOutsideLocations: (loc) =>
       getModifiedTargetUsersOutsideLocations(loc),
     getStats,