tt-help-cli-ycl 1.4.150 → 1.4.152

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "tt-help-cli-ycl",
-  "version": "1.4.150",
+  "version": "1.4.152",
   "description": "TikTok user & video data scraper - extract ttSeller, verified, locationCreated from HTML source",
   "type": "module",
   "bin": {

package/scripts/test-api-interceptor-captcha.mjs

@@ -0,0 +1,134 @@
+// 临时验证：被判重试期间弹出验证码 → fetchUserVideosAPI 立即抛 "验证码:"，不再空转重试
+// 运行：RM_BLOCKED_WAIT_MS=1 node scripts/test-api-interceptor-captcha.mjs
+import { fetchUserVideosAPI } from "../src/lib/api-interceptor.js";
+
+let pass = 0;
+let fail = 0;
+
+function check(name, ok, detail = "") {
+  if (ok) {
+    pass++;
+    console.log(`✓ ${name}${detail ? `: ${detail}` : ""}`);
+  } else {
+    fail++;
+    console.error(`✗ ${name}${detail ? `: ${detail}` : ""}`);
+  }
+}
+
+// 429 + 页面有验证码容器
+function make429CaptchaPage(username) {
+  let gotoCount = 0;
+  const page = {
+    _handlers: {},
+    on(event, fn) {
+      (page._handlers[event] ||= []).push(fn);
+      return page;
+    },
+    off(event) {
+      delete page._handlers[event];
+      return page;
+    },
+    url: () => `https://www.tiktok.com/@${username}`,
+    goto: async () => {
+      gotoCount++;
+    },
+    // checkCdpHealth 用 () => document.readyState（fn.length===0）→ "complete"
+    // detectCaptcha 用 () => {...}（fn.length===0，通过 captchaVisible 开关控制返回值）
+    _captchaVisible: true,
+    evaluate: async (fn) => {
+      if (
+        page._captchaVisible &&
+        fn.toString().includes("captcha-verify-container")
+      ) {
+        return {
+          exists: true,
+          visible: true,
+          rect: { x: 0, y: 0, w: 100, h: 100 },
+        };
+      }
+      // detectPageError（fn.length===1）→ null
+      if (fn.length === 1) return null;
+      return "complete";
+    },
+    reload: async () => {},
+    waitForFunction: async () => null,
+    waitForResponse: async (pred) => {
+      const resp = {
+        url: () => "https://www.tiktok.com/api/post/item_list/?...",
+        status: () => 429,
+        headers: () => ({ "content-length": "19" }),
+        json: async () => {
+          throw new Error("Unexpected token 'r'...");
+        },
+        body: async () => Buffer.from("ratelimit triggered"),
+      };
+      if (!pred(resp)) throw new Error("response 不匹配 item_list");
+      return resp;
+    },
+    get gotoCount() {
+      return gotoCount;
+    },
+  };
+  return page;
+}
+
+// 用例1：429 + 验证码可见 → 立即抛 "验证码:"，只 goto 1 次（无重试）
+{
+  const page = make429CaptchaPage("fumolbbf08o");
+  try {
+    await fetchUserVideosAPI(page, "fumolbbf08o", 5, () => {});
+    check("429+验证码 立即转验证码", false, "未抛错");
+  } catch (e) {
+    check(
+      "429+验证码 立即转验证码",
+      e.message === "验证码: fumolbbf08o",
+      `抛错 "${e.message}"`,
+    );
+  }
+  check(
+    "429+验证码 不再重试（仅 1 次 goto）",
+    page.gotoCount === 1,
+    `goto 次数 = ${page.gotoCount}`,
+  );
+}
+
+// 用例2：429 + 验证码不可见 → 仍走正常被封重试，最终抛 "被封:"
+{
+  const page = make429CaptchaPage("fumolbbf08o");
+  page._captchaVisible = false;
+  try {
+    await fetchUserVideosAPI(page, "fumolbbf08o", 5, () => {});
+    check("429无验证码 仍按被封重试", false, "未抛错");
+  } catch (e) {
+    check(
+      "429无验证码 仍按被封重试",
+      e.message === "被封: fumolbbf08o",
+      `抛错 "${e.message}"，goto ${page.gotoCount} 次`,
+    );
+  }
+}
+
+// 用例3：429 + detectCaptcha 本身抛异常（页面上下文销毁）→ 忽略，按被封重试
+{
+  const page = make429CaptchaPage("fumolbbf08o");
+  page.evaluate = async (fn) => {
+    if (fn.toString().includes("captcha-verify-container")) {
+      throw new Error("Execution context was destroyed");
+    }
+    if (fn.length === 1) return null;
+    return "complete";
+  };
+  try {
+    await fetchUserVideosAPI(page, "fumolbbf08o", 5, () => {});
+    check("检测异常时忽略按被封重试", false, "未抛错");
+  } catch (e) {
+    check(
+      "检测异常时忽略按被封重试",
+      e.message === "被封: fumolbbf08o",
+      `抛错 "${e.message}"`,
+    );
+  }
+}
+
+console.error(`\n结果: ${pass} 通过, ${fail} 失败`);
+process.exit(fail > 0 ? 1 : 0);

package/scripts/test-photo-post.mjs

@@ -0,0 +1,156 @@
+/**
+ * 图文帖（photo post）标记与国家采样过滤测试
+ *
+ * 覆盖：
+ *   1. buildVideoObject：imagePost 标记 / duration=0 兜底 / href 路径
+ *   2. 采样过滤逻辑：图文不占采样位，向后补位真视频
+ *   3. EC 筛选不受影响：图文 EC 帖照常参与
+ *
+ * 运行: node scripts/test-photo-post.mjs
+ */
+import assert from "node:assert";
+
+let passed = 0;
+function check(name, fn) {
+  try {
+    fn();
+    passed++;
+    console.log(`  ✅ ${name}`);
+  } catch (e) {
+    console.error(`  ❌ ${name}\n     ${e.message}`);
+    process.exitCode = 1;
+  }
+}
+
+// buildVideoObject 未导出，从模块内部行为验证：
+// 用真实 itemList item 结构构造，直接验证导出的 fetchUserVideosAPI 太重（需浏览器），
+// 所以这里用最小复制版判定逻辑 + 读源码断言导出行为一致性。
+// 更直接的办法：将判定逻辑抽出来测。当前先用源码 grep + 等价实现双验证。
+
+console.log("===== 图文帖（photo post）处理测试 =====");
+
+console.log("\n[1] 判定逻辑等价实现验证（imagePost / duration 兜底）");
+// 与 buildVideoObject 完全一致的判定
+const isPhotoOf = (item) =>
+  item.imagePost !== undefined || item.video?.duration === 0;
+
+const VIDEO_ITEM = {
+  id: "7684619445913865504",
+  video: { duration: 66, PlayAddrStruct: {}, codecType: "h264" },
+  stats: { playCount: 77400 },
+};
+const PHOTO_ITEM = {
+  id: "7686098281570569504",
+  imagePost: { cover: {}, images: [], shareCover: {}, title: "t" },
+  video: { duration: 0 },
+  stats: { playCount: 23000 },
+};
+const PHOTO_NO_FIELD_ITEM = {
+  // 极端兜底：imagePost 偶发缺失，仅 duration=0
+  id: "111",
+  video: { duration: 0 },
+};
+
+check("正常视频 → 不是图文", () =>
+  assert.strictEqual(isPhotoOf(VIDEO_ITEM), false),
+);
+check("有 imagePost → 是图文", () =>
+  assert.strictEqual(isPhotoOf(PHOTO_ITEM), true),
+);
+check("imagePost 缺失但 duration=0 → 兜底判为图文", () =>
+  assert.strictEqual(isPhotoOf(PHOTO_NO_FIELD_ITEM), true),
+);
+check("无 video 字段（老结构）→ 不是图文", () =>
+  assert.strictEqual(isPhotoOf({ id: "x" }), false),
+);
+
+console.log("\n[2] 采样过滤逻辑（explore-core 的 filter+slice 等价行为）");
+// 模拟混合列表：前 5 个里 2 个图文，过滤后应取到第 6、7 个真视频
+const mixed = [
+  { id: "v1", isPhoto: 0 },
+  { id: "p1", isPhoto: 1 },
+  { id: "v2", isPhoto: 0 },
+  { id: "p2", isPhoto: 1 },
+  { id: "v3", isPhoto: 0 },
+  { id: "v4", isPhoto: 0 }, // 应补位进来
+  { id: "v5", isPhoto: 0 }, // 应补位进来
+  { id: "v6", isPhoto: 0 }, // 采样 5 个，到此截止
+];
+const sampled = mixed.filter((v) => !v.isPhoto).slice(0, 5);
+check("图文不占采样位，向后补位真视频", () =>
+  assert.deepStrictEqual(
+    sampled.map((v) => v.id),
+    ["v1", "v2", "v3", "v4", "v5"],
+  ),
+);
+
+// 纯图文用户：采样 0 条 → 不发起国家查询 → 自然丢弃
+check("纯图文用户 → 采样 0 条", () =>
+  assert.strictEqual(
+    [{ id: "p1", isPhoto: 1 }].filter((v) => !v.isPhoto).length,
+    0,
+  ),
+);
+
+console.log("\n[3] EC 筛选不受影响（图文 EC 帖照常保留）");
+// explore-core 的 EC 筛选条件不含 isPhoto：
+//   v.isECVideo === 1 && 7天内 —— 图文 EC 帖（isECVideo=1 的 photo）应通过
+const ecItems = [
+  {
+    id: "pv1",
+    isECVideo: 1,
+    isPhoto: 1,
+    createTime: 1758000000,
+    playCount: 900,
+  },
+  {
+    id: "vv1",
+    isECVideo: 1,
+    isPhoto: 0,
+    createTime: 1758000000,
+    playCount: 500,
+  },
+];
+const nowSec = 1758100000;
+const recent = ecItems.filter(
+  (v) =>
+    v.isECVideo === 1 && v.createTime && nowSec - v.createTime <= 7 * 86400,
+);
+check("图文 EC 帖仍在 recentEcVideos 候选中", () =>
+  assert.ok(recent.some((v) => v.id === "pv1")),
+);
+
+console.log("\n[4] 源码一致性断言（防止后续改动漂移）");
+const { readFileSync } = await import("node:fs");
+const interceptorSrc = readFileSync(
+  new URL("../src/lib/api-interceptor.js", import.meta.url),
+  "utf-8",
+);
+const exploreSrc = readFileSync(
+  new URL("../src/scraper/explore-core.js", import.meta.url),
+  "utf-8",
+);
+check("buildVideoObject 判定：imagePost + duration 双保险", () =>
+  assert.ok(
+    interceptorSrc.includes(
+      "item.imagePost !== undefined || item.video?.duration === 0",
+    ),
+  ),
+);
+check("buildVideoObject 输出 isPhoto 字段", () =>
+  assert.ok(interceptorSrc.includes("isPhoto: isPhoto ? 1 : 0")),
+);
+check("href 按类型生成 /photo/ 或 /video/", () =>
+  assert.ok(
+    interceptorSrc.includes('${isPhoto ? "photo" : "video"}/${item.id}'),
+  ),
+);
+check("explore-core 采样前过滤图文", () =>
+  assert.ok(
+    exploreSrc.includes("videoArray\n      .filter((v) => !v.isPhoto)"),
+  ),
+);
+
+console.log(
+  `\n结果: ${passed} 通过${process.exitCode ? ", 有失败" : ", 0 失败"}`,
+);

package/scripts/test-precheck-perf.mjs

@@ -0,0 +1,270 @@
+#!/usr/bin/env node
+/**
+ * 回归测试：getPrecheckCandidates 查询改写（2026-09-18）
+ *
+ * 背景：原 SQL 的 WHERE/ORDER BY 用 COALESCE(refresh_time, ?) 表达式，
+ * 挡住索引第 4 列 + 触发 TEMP B-TREE 全排序（76 万行 ~1s）。
+ * 改写为 (refresh_time IS NULL OR refresh_time < ?) + ORDER BY refresh_time。
+ *
+ * 验证点：
+ * 1. 语义等价：新旧 SQL 返回相同的候选集与顺序（NULL 排最前 = 视作 2016 默认值）
+ * 2. 索引生效：建 idx_jobs_precheck 后查询计划不再出现 TEMP B-TREE FOR ORDER BY
+ * 3. 走真实代码路径：createStore(db).getPrecheckCandidates() 返回结构正确
+ *    （uniqueId/nickname/videoCount/refreshTime + 领取后 refresh_time 被置为 now）
+ */
+
+import Database from "better-sqlite3";
+import path from "path";
+import fs from "fs";
+import { fileURLToPath } from "url";
+
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const dbPath = path.join(__dirname, "../test-precheck-perf.db");
+
+// ── 准备库 ──
+if (fs.existsSync(dbPath)) fs.unlinkSync(dbPath);
+const db = new Database(dbPath);
+
+db.exec(`
+  CREATE TABLE jobs (
+    unique_id TEXT PRIMARY KEY,
+    nickname TEXT,
+    status TEXT DEFAULT 'pending',
+    video_count INTEGER,
+    tt_seller INTEGER DEFAULT 0,
+    verified INTEGER DEFAULT 0,
+    location_created TEXT,
+    pending_refresh INTEGER,
+    refresh_time INTEGER,
+    pinned INTEGER DEFAULT 0,
+    guessed_location TEXT,
+    sources TEXT,
+    follower_count INTEGER DEFAULT 0,
+    claimed_at INTEGER,
+    claimed_by TEXT,
+    processed_at INTEGER,
+    user_update_count INTEGER,
+    no_video INTEGER,
+    sec_uid TEXT,
+    top_video_seo_url TEXT,
+    product_name TEXT,
+    product_rating REAL,
+    product_sold_count INTEGER,
+    created_at INTEGER,
+    updated_at INTEGER
+  );
+`);
+
+const NOW = Date.now();
+const DAY = 86400000;
+const OLD = 1452000000000; // 2016-01-05，模拟历史旧值（大于代码 defaultTime）
+// 代码内 defaultTime = new Date("2016-01-01T00:00:00Z")，NULL refresh_time 折算成它
+const DEFAULT_TIME = new Date("2016-01-01T00:00:00Z").getTime();
+
+// 覆盖各种边界：NULL refresh_time、极旧、刚刷新、目标国/非目标国、
+// pending_refresh 0/1/NULL、非商家、已验证
+const rows = [
+  ["u_null_de", "DE", null, 0, 1, 0], // NULL refresh_time → 候选（最优先）
+  ["u_old_pl", "PL", OLD + 1000, 0, 1, 0], // 2016 年代值 → 候选
+  ["u_2d_fr", "FR", NOW - 2 * DAY, 0, 1, 0], // 2 天前 → 候选（maxAge=1d 时）
+  ["u_5h_es", "ES", NOW - 5 * 3600 * 1000, 0, 1, 0], // 5h 前 → 非候选（maxAge=1d）
+  ["u_pr1_it", "IT", NOW - 2 * DAY, 1, 1, 0], // pending_refresh=1 → 非候选
+  ["u_prnull_be", "BE", NOW - 2 * DAY, null, 1, 0], // pending_refresh NULL → 候选
+  ["u_nonseller", "DE", NOW - 2 * DAY, 0, 0, 0], // 非商家 → 非候选
+  ["u_verified", "DE", NOW - 2 * DAY, 0, 1, 1], // 已验证 → 非候选
+  ["u_us", "US", NOW - 3 * DAY, 0, 1, 0], // US 目标国 → 候选
+  ["u_cn", "CN", NOW - 3 * DAY, 0, 1, 0], // 非目标国 → 非候选
+];
+const insert = db.prepare(`
+  INSERT INTO jobs (unique_id, nickname, status, video_count, tt_seller, verified,
+    location_created, pending_refresh, refresh_time, created_at, updated_at)
+  VALUES (?, ?, 'done', 10, ?, ?, ?, ?, ?, ?, ?)
+`);
+for (const [uid, loc, rt, pr, seller, verified] of rows) {
+  insert.run(uid, uid, seller, verified, loc, pr, rt, NOW, NOW);
+}
+
+const TARGETS = [
+  "PL",
+  "NL",
+  "BE",
+  "DE",
+  "FR",
+  "IT",
+  "IE",
+  "ES",
+  "CZ",
+  "GR",
+  "HU",
+  "PT",
+  "AT",
+  "TH",
+  "US",
+  "MY",
+  "VN",
+];
+const placeholders = TARGETS.map(() => "?").join(",");
+
+// ── 1) 语义等价：新旧 SQL 逐行对比（含顺序） ──
+function runOld(maxAgeSec) {
+  const threshold = NOW - maxAgeSec * 1000;
+  return db
+    .prepare(
+      `SELECT unique_id, COALESCE(refresh_time, ?) AS rt FROM jobs
+       WHERE tt_seller = 1 AND verified = 0
+         AND (pending_refresh IS NULL OR pending_refresh = 0)
+         AND location_created IN (${placeholders})
+         AND COALESCE(refresh_time, ?) < ?
+       ORDER BY COALESCE(refresh_time, ?) ASC`,
+    )
+    .all(OLD, ...TARGETS, OLD, threshold, OLD)
+    .map((r) => `${r.unique_id}@${r.rt}`);
+}
+function runNew(maxAgeSec) {
+  const threshold = NOW - maxAgeSec * 1000;
+  return db
+    .prepare(
+      `SELECT unique_id, COALESCE(refresh_time, ?) AS rt FROM jobs
+       WHERE tt_seller = 1 AND verified = 0
+         AND (pending_refresh IS NULL OR pending_refresh = 0)
+         AND location_created IN (${placeholders})
+         AND (refresh_time IS NULL OR refresh_time < ?)
+       ORDER BY refresh_time ASC`,
+    )
+    .all(OLD, ...TARGETS, threshold)
+    .map((r) => `${r.unique_id}@${r.rt}`);
+}
+
+let failed = 0;
+const check = (name, cond, detail) => {
+  if (cond) {
+    console.error(`✓ ${name}`);
+  } else {
+    failed++;
+    console.error(`✗ ${name} ${detail ?? ""}`);
+  }
+};
+
+for (const maxAge of [86400, 43200]) {
+  const a = runOld(maxAge);
+  const b = runNew(maxAge);
+  check(
+    `新旧 SQL 结果一致 (maxAge=${maxAge}s)`,
+    JSON.stringify(a) === JSON.stringify(b),
+    `\n  old=${JSON.stringify(a)}\n  new=${JSON.stringify(b)}`,
+  );
+}
+// 顺带验证期望集合（maxAge=1d：NULL 和旧值排最前，非候选被排除）
+const expect1d = [
+  "u_null_de@1452000000000",
+  "u_old_pl@1452000001000",
+  "u_2d_fr",
+  "u_prnull_be",
+  "u_us",
+]
+  .map((s) => (s.includes("@") ? s : null))
+  .filter(Boolean);
+const got1d = runNew(86400);
+check(
+  "新 SQL 命中预期候选集",
+  got1d.slice(0, 2).join() === expect1d.slice(0, 2).join() &&
+    got1d.length === 5 &&
+    got1d.includes("u_prnull_be@" + (NOW - 2 * DAY)),
+  JSON.stringify(got1d),
+);
+
+// ── 2) 索引生效：建部分索引后无 TEMP B-TREE ──
+db.exec(
+  `CREATE INDEX idx_jobs_precheck ON jobs(refresh_time)
+   WHERE tt_seller = 1 AND verified = 0 AND (pending_refresh IS NULL OR pending_refresh = 0)`,
+);
+const plan = db
+  .prepare(
+    `EXPLAIN QUERY PLAN SELECT unique_id FROM jobs
+     WHERE tt_seller = 1 AND verified = 0
+       AND (pending_refresh IS NULL OR pending_refresh = 0)
+       AND location_created IN (${placeholders})
+       AND (refresh_time IS NULL OR refresh_time < ?)
+     ORDER BY refresh_time ASC LIMIT 20`,
+  )
+  .all(...TARGETS, NOW - 86400000)
+  .map((r) => r.detail)
+  .join(" | ");
+check(
+  "查询计划走 idx_jobs_precheck 且无 TEMP B-TREE 排序",
+  plan.includes("idx_jobs_precheck") && !plan.includes("TEMP B-TREE"),
+  plan,
+);
+
+// ── 3) 走真实代码路径 createStore().getPrecheckCandidates ──
+process.env.TT_HELP_DB_PATH = dbPath; // 防御性标记，createStore 用显式路径
+const { createStore } = await import("../src/watch/data-store.js");
+process.chdir(path.join(__dirname, "..")); // app-config 读 cwd 相对配置
+const store = createStore(dbPath);
+const candidates = store.getPrecheckCandidates({
+  limit: 3,
+  intervalSeconds: 86400,
+  locations: null,
+});
+
+check("返回数量 ≤ limit", candidates.length === 3, `got ${candidates.length}`);
+check(
+  "字段结构正确（uniqueId/nickname/videoCount/refreshTime）",
+  candidates.every(
+    (c) =>
+      typeof c.uniqueId === "string" &&
+      typeof c.videoCount === "number" &&
+      (c.refreshTime === OLD || typeof c.refreshTime === "number"),
+  ),
+  JSON.stringify(candidates[0]),
+);
+check(
+  "顺序 = refreshTime 升序（NULL 折算默认值排最前）",
+  candidates[0].uniqueId === "u_null_de" &&
+    candidates[0].refreshTime === DEFAULT_TIME &&
+    candidates[1].uniqueId === "u_old_pl" &&
+    candidates[2].uniqueId === "u_us", // 3天前 < 2天前，升序在前
+  JSON.stringify(candidates.map((c) => c.uniqueId)),
+);
+// 领取即锁定：选中的行 refresh_time 应已更新为 now（防重复领取）
+const after = db
+  .prepare(
+    `SELECT unique_id, refresh_time FROM jobs WHERE unique_id IN (${candidates.map(() => "?").join(",")})`,
+  )
+  .all(...candidates.map((c) => c.uniqueId));
+check(
+  "领取后 refresh_time 已置为当前时间（防重复领取语义保留）",
+  after.every((r) => r.refresh_time !== null && r.refresh_time >= NOW),
+  JSON.stringify(after),
+);
+// 锁定后再次领取不应重复返回这 3 个
+const second = store.getPrecheckCandidates({
+  limit: 50,
+  intervalSeconds: 86400,
+  locations: null,
+});
+check(
+  "已领取用户不会重复发放",
+  !second.some((c) => candidates.some((p) => p.uniqueId === c.uniqueId)),
+  JSON.stringify(second.map((c) => c.uniqueId)),
+);
+// markPrecheckResult 双分支（needRefresh 0/1）仍正常
+store.markPrecheckResult("u_us", { needRefresh: true, videoCount: 12 });
+store.markPrecheckResult("u_2d_fr", { needRefresh: false, videoCount: 8 });
+const marked = db
+  .prepare(
+    `SELECT unique_id, pending_refresh, video_count FROM jobs WHERE unique_id IN ('u_us','u_2d_fr')`,
+  )
+  .all();
+check(
+  "markPrecheckResult 正常写回",
+  marked.find((r) => r.unique_id === "u_us").pending_refresh === 1 &&
+    marked.find((r) => r.unique_id === "u_us").video_count === 12 &&
+    marked.find((r) => r.unique_id === "u_2d_fr").pending_refresh === 0,
+  JSON.stringify(marked),
+);
+
+db.close();
+fs.unlinkSync(dbPath);
+console.error(failed === 0 ? "\n全部通过 ✅" : `\n${failed} 项失败 ❌`);
+process.exit(failed === 0 ? 0 : 1);

package/scripts/test-product-fill-cdp.mjs

@@ -32,6 +32,10 @@ function check(name, fn) {
 const VALID_HTML =
   '<html><body><script>{"product_model":{"name":"Test Product","product_id":"123456","seller_id":"789"},"sold_count":"296","review_ratings":{"overall_score":4.7,"review_count":"3303","rating_result":[5,4,3,2,1]}}</script></body></html>';
 
+// 含多 SKU 的 PDP HTML（结构与线上实测一致：sku_id 19位 / property_pairs / sku_quantity）
+const VALID_HTML_WITH_SKUS =
+  '<html><body><script>{"product_model":{"name":"Test Product","product_id":"123456","seller_id":"789"},"sold_count":"962","review_ratings":{"overall_score":4.8,"review_count":"174"},"skus":[{"sku_id":"1731140428076257811","sku_name":"810003369352","sku_status":1,"property_pairs":[{"sku_property_id":"7003574686483515141","sku_property_value_id":"6996631003314030337","sku_property_name":"Shade","sku_property_value_name":"Wanderlust"}],"sku_quantity":{"warehouse_id":"7676336681096611597","available_quantity":106},"sku_stock_status":1},{"sku_id":"1729888889849549331","sku_name":"810003367204","sku_status":1,"property_pairs":[{"sku_property_name":"Shade","sku_property_value_name":"Ballerina"}],"sku_quantity":{"warehouse_id":"7676336681096611597","available_quantity":2017},"sku_stock_status":1}]}</script></body></html>';
+
 console.log("\n[1] isEmptyProductDetail 全空判定");
 check("全 null → true", () => {
   assert.strictEqual(
@@ -98,6 +102,46 @@ check("解析结果非全空", () =>
   assert.strictEqual(isEmptyProductDetail(detail), false),
 );
 
+console.log("\n[3b] parseProductHtml 多 SKU 提取");
+const skuDetail = parseProductHtml(VALID_HTML_WITH_SKUS);
+check("原有字段不受 skus 影响（rating）", () =>
+  assert.strictEqual(skuDetail.rating, 4.8),
+);
+check("原有字段不受 skus 影响（soldCount）", () =>
+  assert.strictEqual(skuDetail.soldCount, 962),
+);
+check(
+  "skus 数量 = 2",
+  () =>
+    assert.strictEqual(Array.isArray(skuDetail.skus), true) &&
+    assert.strictEqual(skuDetail.skus.length, 2),
+);
+check("skuId 19位字符串保精度", () =>
+  assert.strictEqual(skuDetail.skus[0].skuId, "1731140428076257811"),
+);
+check("specs 展开 property_pairs", () =>
+  assert.strictEqual(skuDetail.skus[0].specs, "Shade: Wanderlust"),
+);
+check(
+  "库存 available_quantity 提取",
+  () =>
+    assert.strictEqual(skuDetail.skus[0].stock, 106) &&
+    assert.strictEqual(skuDetail.skus[1].stock, 2017),
+);
+check("warehouseId 字符串化", () =>
+  assert.strictEqual(skuDetail.skus[0].warehouseId, "7676336681096611597"),
+);
+check("stockStatus 提取", () =>
+  assert.strictEqual(skuDetail.skus[0].stockStatus, 1),
+);
+check("无 skus 段 → skus = null", () =>
+  assert.strictEqual(parseProductHtml(VALID_HTML).skus, null),
+);
+check("skus 为 null 时不影响 isEmptyProductDetail", () => {
+  const d = parseProductHtml(VALID_HTML);
+  return assert.strictEqual(isEmptyProductDetail(d), false);
+});
+
 console.log("\n[4] parseProductHtml 验证码页抛 captcha 错误");
 check("captcha-verify-container → 抛 captcha", () => {
   assert.throws(

package/scripts/test-product-fill-db.mjs

@@ -153,6 +153,49 @@ check("sellerA 不再出现在候选", () => {
   assert(!tasksAfter.some((t) => t.uniqueId === "sellerA"));
 });
 
+console.log("\n[3b] skus 入库（product_skus JSON 列）");
+const skuRet = store.updateProductInfo("sellerB", {
+  rating: 4.8,
+  reviewCount: 174,
+  soldCount: 962,
+  productName: "Sku Product",
+  skus: [
+    {
+      skuId: "1731140428076257811",
+      specs: "Shade: Wanderlust",
+      stock: 106,
+      stockStatus: 1,
+      warehouseId: "7676336681096611597",
+    },
+    {
+      skuId: "1729888889849549331",
+      specs: "Shade: Ballerina",
+      stock: 2017,
+      stockStatus: 1,
+      warehouseId: "7676336681096611597",
+    },
+  ],
+});
+check("带 skus 回写返回 ok", () => assert(skuRet.ok === true));
+check("product_skus 列已写入 JSON（读回可解析）", () => {
+  const rowB = db
+    .prepare("SELECT product_skus FROM jobs WHERE unique_id = ?")
+    .get("sellerB");
+  const parsed = JSON.parse(rowB.product_skus);
+  assert(parsed.length === 2);
+  assert(parsed[0].skuId === "1731140428076257811");
+  assert(parsed[1].stock === 2017);
+});
+check("无 skus 回写（老客户端）不报错", () => {
+  const ret = store.updateProductInfo("sellerB", {
+    rating: 4.0,
+    soldCount: 10,
+    productName: "Old",
+    reviewCount: 5,
+  });
+  assert(ret.ok === true);
+});
+
 console.log("\n[4] 异常输入");
 check("缺 productDetail → error", () => {
   const r = store.updateProductInfo("sellerB", null);