1

.gitignore

@@ -1,2 +1,4 @@
 server/node_modules/*
+server/puppeteer_profile/*
 .env
+chrome-win/*

server/.env.example

@@ -11,3 +11,5 @@ CRX_SRC_PATH=d:/项目/电商抓取项目/mv2_simple_crx/src
 SERVER_PORT=38080
 ACTION_TIMEOUT_MS=300000
 PUPPETEER_HEADLESS=false
+# 可选：指定浏览器路径（不填默认用 ../chrome-win/chrome.exe）
+CHROME_EXECUTABLE_PATH=

server/README.md

@@ -0,0 +1,40 @@
+## server（Koa + Sequelize + MySQL）
+
+### 功能
+- `POST /api/crawl/run_action`：服务端调用扩展 action，返回结果，并按 `stage` 自动入库
+- **定时任务写死配置**：`config/cron_tasks.js`（不走数据库）
+
+### 运行
+1. 安装依赖
+
+```bash
+cd server
+npm install
+```
+
+2. 配置环境变量
+
+```bash
+copy .env.example .env
+```
+
+3. 初始化/更新表结构
+
+```bash
+node scripts/db_sync.js
+```
+
+4. 启动
+
+```bash
+node app.js
+```
+
+### 定时任务
+编辑 `config/cron_tasks.js` 的 `cron_task_list`，重启服务即可生效。
+
+### 落库表
+- `crawl_run_record`：所有 action 调用的原始请求/响应
+- `amazon_product`：`stage=detail` 详情
+- `amazon_search_item`：`stage=list` 列表 item
+- `amazon_review`：`stage=reviews` 评论

server/app.js

@@ -4,8 +4,7 @@ import body_parser from 'koa-bodyparser';
 
 import { sequelize } from './models/index.js';
 import { crawl_router } from './routes/crawl.js';
-import { schedule_task_router } from './routes/schedule_task.js';
-import { reload_all_schedules } from './services/schedule_loader.js';
+import { start_all_cron_tasks } from './services/schedule_loader.js';
 
 dotenv.config();
 
@@ -13,7 +12,6 @@ const app = new Koa();
 app.use(body_parser({ jsonLimit: '10mb' }));
 
 app.use(crawl_router.routes()).use(crawl_router.allowedMethods());
-app.use(schedule_task_router.routes()).use(schedule_task_router.allowedMethods());
 
 app.use(async (ctx) => {
   ctx.status = 404;
@@ -24,7 +22,7 @@ const port = Number(process.env.SERVER_PORT || 38080);
 
 await sequelize.authenticate();
 await sequelize.sync();
-await reload_all_schedules();
+start_all_cron_tasks();
 
 app.listen(port);
 // eslint-disable-next-line no-console

server/config/cron_tasks.js

@@ -0,0 +1,14 @@
+/**
+ * 写死定时任务配置（不走数据库）
+ * cron_expression 参考 node-cron
+ */
+
+export const cron_task_list = [
+  // 示例：每 6 小时跑一次列表抓取
+  // {
+  //   name: 'amazon_search_list_every_6h',
+  //   cron_expression: '0 */6 * * *',
+  //   action_name: 'amazon_search_list',
+  //   action_payload: { keyword: '午餐包', limit: 100 }
+  // }
+];

server/models/amazon_product.js

@@ -0,0 +1,39 @@
+import { DataTypes } from 'sequelize';
+
+export function define_amazon_product(sequelize) {
+  return sequelize.define(
+    'amazon_product',
+    {
+      id: { type: DataTypes.BIGINT.UNSIGNED, primaryKey: true, autoIncrement: true },
+      asin: { type: DataTypes.STRING(32), allowNull: false },
+      url: { type: DataTypes.TEXT, allowNull: false },
+      title: { type: DataTypes.TEXT, allowNull: true },
+      price: { type: DataTypes.STRING(64), allowNull: true },
+      sku: { type: DataTypes.STRING(256), allowNull: true },
+      sku_color: { type: DataTypes.STRING(128), allowNull: true },
+      sku_size: { type: DataTypes.STRING(128), allowNull: true },
+      brand_line: { type: DataTypes.TEXT, allowNull: true },
+      brand_store_url: { type: DataTypes.TEXT, allowNull: true },
+      ac_badge: { type: DataTypes.STRING(128), allowNull: true },
+      bestseller_hint: { type: DataTypes.TEXT, allowNull: true },
+      delivery_hint: { type: DataTypes.TEXT, allowNull: true },
+      social_proof: { type: DataTypes.TEXT, allowNull: true },
+      sustainability_hint: { type: DataTypes.TEXT, allowNull: true },
+      rating_stars: { type: DataTypes.STRING(64), allowNull: true },
+      review_count_text: { type: DataTypes.STRING(64), allowNull: true },
+      main_image: { type: DataTypes.TEXT, allowNull: true },
+      images_json: { type: DataTypes.TEXT('long'), allowNull: true },
+      bullets_json: { type: DataTypes.TEXT('long'), allowNull: true },
+      product_info_json: { type: DataTypes.TEXT('long'), allowNull: true },
+      detail_extra_lines_json: { type: DataTypes.TEXT('long'), allowNull: true }
+    },
+    {
+      tableName: 'amazon_product',
+      indexes: [
+        { unique: true, fields: ['asin'] },
+        { fields: ['created_at'] },
+        { fields: ['updated_at'] }
+      ]
+    }
+  );
+}

server/models/amazon_review.js

@@ -0,0 +1,31 @@
+import { DataTypes } from 'sequelize';
+
+export function define_amazon_review(sequelize) {
+  return sequelize.define(
+    'amazon_review',
+    {
+      id: { type: DataTypes.BIGINT.UNSIGNED, primaryKey: true, autoIncrement: true },
+      asin: { type: DataTypes.STRING(32), allowNull: true },
+      url: { type: DataTypes.TEXT, allowNull: false },
+      review_id: { type: DataTypes.STRING(64), allowNull: false },
+      author: { type: DataTypes.STRING(256), allowNull: true },
+      title: { type: DataTypes.TEXT, allowNull: true },
+      body: { type: DataTypes.TEXT('long'), allowNull: true },
+      rating_text: { type: DataTypes.STRING(64), allowNull: true },
+      review_date: { type: DataTypes.STRING(128), allowNull: true },
+      review_index: { type: DataTypes.INTEGER, allowNull: true },
+      batch_key: { type: DataTypes.STRING(64), allowNull: false },
+      batch_total: { type: DataTypes.INTEGER, allowNull: true },
+      batch_limit: { type: DataTypes.INTEGER, allowNull: true }
+    },
+    {
+      tableName: 'amazon_review',
+      indexes: [
+        { unique: true, fields: ['review_id'] },
+        { fields: ['asin'] },
+        { fields: ['batch_key'] },
+        { fields: ['created_at'] }
+      ]
+    }
+  );
+}

server/models/amazon_search_item.js

@@ -0,0 +1,30 @@
+import { DataTypes } from 'sequelize';
+
+export function define_amazon_search_item(sequelize) {
+  return sequelize.define(
+    'amazon_search_item',
+    {
+      id: { type: DataTypes.BIGINT.UNSIGNED, primaryKey: true, autoIncrement: true },
+      asin: { type: DataTypes.STRING(32), allowNull: false },
+      url: { type: DataTypes.TEXT, allowNull: false },
+      title: { type: DataTypes.TEXT, allowNull: true },
+      price: { type: DataTypes.STRING(64), allowNull: true },
+      rating: { type: DataTypes.FLOAT, allowNull: true },
+      rating_text: { type: DataTypes.STRING(64), allowNull: true },
+      review_count: { type: DataTypes.INTEGER, allowNull: true },
+      review_count_text: { type: DataTypes.STRING(64), allowNull: true },
+      rank_index: { type: DataTypes.INTEGER, allowNull: true, comment: '列表中的 index 字段' },
+      batch_key: { type: DataTypes.STRING(64), allowNull: false, comment: '一次列表抓取的批次 key' },
+      batch_total: { type: DataTypes.INTEGER, allowNull: true },
+      batch_limit: { type: DataTypes.INTEGER, allowNull: true }
+    },
+    {
+      tableName: 'amazon_search_item',
+      indexes: [
+        { fields: ['asin'] },
+        { fields: ['batch_key'] },
+        { fields: ['created_at'] }
+      ]
+    }
+  );
+}

server/models/crawl_run_record.js

@@ -15,8 +15,7 @@ export function define_crawl_run_record(sequelize) {
         allowNull: false,
         defaultValue: 'api',
         comment: 'api | cron'
-      },
-      schedule_task_id: { type: DataTypes.BIGINT.UNSIGNED, allowNull: true }
+      }
     },
     {
       tableName: 'crawl_run_record',

server/models/index.js

@@ -1,7 +1,9 @@
 import { Sequelize } from 'sequelize';
 import { get_sequelize_options } from '../config/database.js';
 import { define_crawl_run_record } from './crawl_run_record.js';
-import { define_schedule_task } from './schedule_task.js';
+import { define_amazon_product } from './amazon_product.js';
+import { define_amazon_search_item } from './amazon_search_item.js';
+import { define_amazon_review } from './amazon_review.js';
 
 const sequelize_options = get_sequelize_options();
 const { database, username, password, ...rest } = sequelize_options;
@@ -9,7 +11,7 @@ const { database, username, password, ...rest } = sequelize_options;
 export const sequelize = new Sequelize(database, username, password, rest);
 
 export const crawl_run_record = define_crawl_run_record(sequelize);
-export const schedule_task = define_schedule_task(sequelize);
 
-schedule_task.hasMany(crawl_run_record, { foreignKey: 'schedule_task_id', as: 'records' });
-crawl_run_record.belongsTo(schedule_task, { foreignKey: 'schedule_task_id', as: 'schedule_task' });
+export const amazon_product = define_amazon_product(sequelize);
+export const amazon_search_item = define_amazon_search_item(sequelize);
+export const amazon_review = define_amazon_review(sequelize);

server/models/schedule_task.js

@@ -1,19 +1 @@
-import { DataTypes } from 'sequelize';
-
-export function define_schedule_task(sequelize) {
-  return sequelize.define(
-    'schedule_task',
-    {
-      id: { type: DataTypes.BIGINT.UNSIGNED, primaryKey: true, autoIncrement: true },
-      name: { type: DataTypes.STRING(128), allowNull: false },
-      cron_expression: { type: DataTypes.STRING(64), allowNull: false },
-      action_name: { type: DataTypes.STRING(128), allowNull: false },
-      payload_json: { type: DataTypes.TEXT, allowNull: true },
-      enabled: { type: DataTypes.BOOLEAN, allowNull: false, defaultValue: true },
-      last_run_at: { type: DataTypes.DATE, allowNull: true }
-    },
-    {
-      tableName: 'schedule_task'
-    }
-  );
-}
+// 已废弃：按需求改为写死定时任务（见 config/cron_tasks.js）

server/routes/schedule_task.js

@@ -1,73 +1 @@
-import Router from '@koa/router';
-import { schedule_task } from '../models/index.js';
-import { safe_json_parse, safe_json_stringify } from '../services/json_utils.js';
-
-export const schedule_task_router = new Router();
-
-schedule_task_router.post('/api/schedule_task/create', async (ctx) => {
-  const { name, cron_expression, action_name, payload } = ctx.request.body || {};
-  if (!name || !cron_expression || !action_name) {
-    ctx.status = 400;
-    ctx.body = { ok: false, error: '缺少 name/cron_expression/action_name' };
-    return;
-  }
-
-  const payload_json = payload ? safe_json_stringify(payload) : null;
-  const row = await schedule_task.create({
-    name,
-    cron_expression,
-    action_name,
-    payload_json,
-    enabled: true
-  });
-
-  ctx.body = { ok: true, data: { id: row.id } };
-});
-
-schedule_task_router.post('/api/schedule_task/list', async (ctx) => {
-  const rows = await schedule_task.findAll({ order: [['id', 'desc']] });
-  ctx.body = {
-    ok: true,
-    data: rows.map((r) => ({
-      id: r.id,
-      name: r.name,
-      cron_expression: r.cron_expression,
-      action_name: r.action_name,
-      payload: safe_json_parse(r.payload_json),
-      enabled: r.enabled,
-      last_run_at: r.last_run_at
-    }))
-  };
-});
-
-schedule_task_router.post('/api/schedule_task/set_enabled', async (ctx) => {
-  const { id, enabled } = ctx.request.body || {};
-  if (!id || typeof enabled !== 'boolean') {
-    ctx.status = 400;
-    ctx.body = { ok: false, error: '缺少 id/enabled(boolean)' };
-    return;
-  }
-
-  const row = await schedule_task.findByPk(id);
-  if (!row) {
-    ctx.status = 404;
-    ctx.body = { ok: false, error: '任务不存在' };
-    return;
-  }
-
-  row.enabled = enabled;
-  await row.save();
-  ctx.body = { ok: true };
-});
-
-schedule_task_router.post('/api/schedule_task/delete', async (ctx) => {
-  const { id } = ctx.request.body || {};
-  if (!id) {
-    ctx.status = 400;
-    ctx.body = { ok: false, error: '缺少 id' };
-    return;
-  }
-
-  await schedule_task.destroy({ where: { id } });
-  ctx.body = { ok: true };
-});
+// 已废弃：按需求改为写死定时任务（见 config/cron_tasks.js）

server/services/amazon_persist.js

@@ -0,0 +1,110 @@
+import {
+  amazon_product,
+  amazon_search_item,
+  amazon_review
+} from '../models/index.js';
+import { safe_json_stringify } from './json_utils.js';
+
+function build_batch_key(prefix) {
+  return `${prefix}_${Date.now()}_${Math.random().toString().slice(2, 8)}`;
+}
+
+function pick_asin_from_url(url) {
+  if (!url) return null;
+  const m = String(url).match(/\/dp\/([A-Z0-9]{8,16})/i);
+  return m && m[1] ? m[1].toUpperCase() : null;
+}
+
+export async function persist_amazon_result(result) {
+  if (!result || !result.stage) {
+    return;
+  }
+
+  if (result.stage === 'detail') {
+    const asin = result.asin || pick_asin_from_url(result.url);
+    if (!asin) {
+      return;
+    }
+
+    await amazon_product.upsert({
+      asin,
+      url: result.url || '',
+      title: result.title || null,
+      price: result.price || null,
+      sku: result.sku || null,
+      sku_color: result.sku_color || null,
+      sku_size: result.sku_size || null,
+      brand_line: result.brand_line || null,
+      brand_store_url: result.brand_store_url || null,
+      ac_badge: result.ac_badge || null,
+      bestseller_hint: result.bestseller_hint || null,
+      delivery_hint: result.delivery_hint || null,
+      social_proof: result.social_proof || null,
+      sustainability_hint: result.sustainability_hint || null,
+      rating_stars: result.rating_stars || null,
+      review_count_text: result.review_count_text || null,
+      main_image: result.main_image || null,
+      images_json: safe_json_stringify(result.images || []),
+      bullets_json: safe_json_stringify(result.bullets || []),
+      product_info_json: safe_json_stringify(result.product_info || {}),
+      detail_extra_lines_json: safe_json_stringify(result.detail_extra_lines || [])
+    });
+
+    return;
+  }
+
+  if (result.stage === 'list') {
+    const batch_key = build_batch_key('list');
+    const items = Array.isArray(result.items) ? result.items : [];
+
+    for (const it of items) {
+      const asin = it.asin || pick_asin_from_url(it.url);
+      if (!asin || !it.url) continue;
+
+      await amazon_search_item.create({
+        asin,
+        url: it.url,
+        title: it.title || null,
+        price: it.price || null,
+        rating: typeof it.rating === 'number' ? it.rating : null,
+        rating_text: it.rating_text || null,
+        review_count: typeof it.review_count === 'number' ? it.review_count : null,
+        review_count_text: it.review_count_text || null,
+        rank_index: typeof it.index === 'number' ? it.index : null,
+        batch_key,
+        batch_total: typeof result.total === 'number' ? result.total : null,
+        batch_limit: typeof result.limit === 'number' ? result.limit : null
+      });
+    }
+
+    return;
+  }
+
+  if (result.stage === 'reviews') {
+    const batch_key = build_batch_key('reviews');
+    const asin = pick_asin_from_url(result.url);
+    const items = Array.isArray(result.items) ? result.items : [];
+
+    for (const it of items) {
+      const review_id = it.review_id;
+      if (!review_id) continue;
+
+      const asin_value = asin || pick_asin_from_url(it.url) || pick_asin_from_url(result.url);
+
+      await amazon_review.upsert({
+        asin: asin_value || null,
+        url: result.url || '',
+        review_id,
+        author: it.author || null,
+        title: it.title || null,
+        body: it.body || null,
+        rating_text: it.rating_text || null,
+        review_date: it.date || null,
+        review_index: typeof it.index === 'number' ? it.index : null,
+        batch_key,
+        batch_total: typeof result.total === 'number' ? result.total : null,
+        batch_limit: typeof result.limit === 'number' ? result.limit : null
+      });
+    }
+  }
+}

server/services/cron_manager.js

@@ -1,30 +1 @@
-import cron from 'node-cron';
-
-const task_id_to_cron_job = new Map();
-
-export function stop_all_cron_jobs() {
-  for (const job of task_id_to_cron_job.values()) {
-    job.stop();
-  }
-  task_id_to_cron_job.clear();
-}
-
-export function upsert_cron_job(schedule_task_id, cron_expression, on_tick) {
-  const existing = task_id_to_cron_job.get(schedule_task_id);
-  if (existing) {
-    existing.stop();
-    task_id_to_cron_job.delete(schedule_task_id);
-  }
-
-  const job = cron.schedule(cron_expression, on_tick, { scheduled: true });
-  task_id_to_cron_job.set(schedule_task_id, job);
-}
-
-export function remove_cron_job(schedule_task_id) {
-  const job = task_id_to_cron_job.get(schedule_task_id);
-  if (!job) {
-    return;
-  }
-  job.stop();
-  task_id_to_cron_job.delete(schedule_task_id);
-}
+// 已废弃：按需求改为写死定时任务（见 config/cron_tasks.js）

server/services/puppeteer_runner.js

@@ -1,4 +1,5 @@
 import dotenv from 'dotenv';
+import fs from 'node:fs';
 import path from 'node:path';
 import puppeteer from 'puppeteer';
 
@@ -30,17 +31,64 @@ function get_extension_id_from_targets(targets) {
   return null;
 }
 
+async function wait_for_extension_id(browser, timeout_ms) {
+  const existing = get_extension_id_from_targets(browser.targets());
+  if (existing) {
+    return existing;
+  }
+
+  const target = await browser
+    .waitForTarget((t) => {
+      const url = t.url();
+      return typeof url === 'string' && url.startsWith('chrome-extension://');
+    }, { timeout: timeout_ms })
+    .catch(() => null);
+
+  if (!target) {
+    return null;
+  }
+
+  return get_extension_id_from_targets([target]);
+}
+
+function get_chrome_executable_path() {
+  // 优先环境变量，方便你后续切换版本
+  const from_env = process.env.CHROME_EXECUTABLE_PATH;
+  if (from_env) {
+    return path.resolve(from_env);
+  }
+
+  // 默认使用项目根目录的 chrome-win/chrome.exe
+  // 当前进程 cwd 通常是 server/，所以回到上一级
+  return path.resolve(process.cwd(), '../chrome-win/chrome.exe');
+}
+
 export async function get_or_create_browser() {
   if (browser_singleton) {
     return browser_singleton;
   }
 
-  const extension_path = path.resolve(get_crx_src_path());
+  const chrome_executable_path = get_chrome_executable_path();
+  if (!fs.existsSync(chrome_executable_path)) {
+    throw new Error(`Chrome 不存在: ${chrome_executable_path}`);
+  }
+
+  const raw_extension_path = path.resolve(get_crx_src_path());
+  const manifest_path = path.resolve(raw_extension_path, 'manifest.json');
+  if (!fs.existsSync(manifest_path)) {
+    throw new Error(`扩展 manifest.json 不存在: ${manifest_path}`);
+  }
+
+  const extension_path = raw_extension_path.replace(/\\/g, '/');
   const headless = String(process.env.PUPPETEER_HEADLESS || 'false') === 'true';
+  const user_data_dir = path.resolve(process.cwd(), 'puppeteer_profile');
 
   browser_singleton = await puppeteer.launch({
+    executablePath: chrome_executable_path,
     headless,
     args: [
+      `--user-data-dir=${user_data_dir}`,
+      '--enable-extensions',
       `--disable-extensions-except=${extension_path}`,
       `--load-extension=${extension_path}`,
       '--no-default-browser-check',
@@ -58,11 +106,19 @@ export async function invoke_extension_action(action_name, action_payload) {
   const page = await browser.newPage();
   await page.goto('about:blank');
 
-  const targets = await browser.targets();
-  const extension_id = get_extension_id_from_targets(targets);
+  // 尝试先打开 chrome://extensions 触发扩展初始化（某些环境下扩展 target 不会立刻出现）
+  try {
+    await page.goto('chrome://extensions/', { waitUntil: 'domcontentloaded' });
+  } catch (err) {
+    // ignore
+  }
+
+  const extension_id = await wait_for_extension_id(browser, 15000);
   if (!extension_id) {
     await page.close();
-    throw new Error('未找到扩展 extension_id（请确认 CRX_SRC_PATH 指向 src 且成功加载）');
+    throw new Error(
+      '未找到扩展 extension_id：Chrome 未加载扩展（常见原因：MV2 被禁用/企业策略未生效/CRX_SRC_PATH 不正确/使用了 headless）'
+    );
   }
 
   const bridge_url = `chrome-extension://${extension_id}/bridge/bridge.html`;

server/services/schedule_loader.js

@@ -1,33 +1,30 @@
-import { schedule_task } from '../models/index.js';
-import { safe_json_parse } from './json_utils.js';
+import cron from 'node-cron';
+import { cron_task_list } from '../config/cron_tasks.js';
 import { execute_action_and_record } from './task_executor.js';
-import { remove_cron_job, upsert_cron_job } from './cron_manager.js';
 
-export async function reload_all_schedules() {
-  const rows = await schedule_task.findAll();
+const cron_jobs = [];
 
-  for (const row of rows) {
-    if (!row.enabled) {
-      remove_cron_job(row.id);
-      continue;
-    }
-
-    upsert_cron_job(row.id, row.cron_expression, async () => {
+export function start_all_cron_tasks() {
+  for (const task of cron_task_list) {
+    const job = cron.schedule(task.cron_expression, async () => {
       try {
-        await schedule_task.update(
-          { last_run_at: new Date() },
-          { where: { id: row.id } }
-        );
-
         await execute_action_and_record({
-          action_name: row.action_name,
-          action_payload: safe_json_parse(row.payload_json) || {},
-          source: 'cron',
-          schedule_task_id: row.id
+          action_name: task.action_name,
+          action_payload: task.action_payload || {},
+          source: 'cron'
         });
       } catch (err) {
-        // cron 执行失败已在 crawl_run_record 落库，避免重复抛出影响其它任务
+        // 失败会在 crawl_run_record 落库
       }
     });
+
+    cron_jobs.push(job);
   }
 }
+
+export function stop_all_cron_tasks() {
+  for (const job of cron_jobs) {
+    job.stop();
+  }
+  cron_jobs.length = 0;
+}

server/services/task_executor.js

@@ -1,14 +1,10 @@
 import { crawl_run_record } from '../models/index.js';
 import { safe_json_stringify } from './json_utils.js';
 import { invoke_extension_action } from './puppeteer_runner.js';
+import { persist_amazon_result } from './amazon_persist.js';
 
 export async function execute_action_and_record(params) {
-  const {
-    action_name,
-    action_payload,
-    source,
-    schedule_task_id
-  } = params;
+  const { action_name, action_payload, source } = params;
 
   const request_payload = safe_json_stringify(action_payload || {});
 
@@ -18,8 +14,13 @@ export async function execute_action_and_record(params) {
 
   try {
     const result = await invoke_extension_action(action_name, action_payload || {});
+
     ok = true;
     result_payload = safe_json_stringify(result);
+
+    // 按 stage 自动入库（不影响原始 run_record 记录）
+    await persist_amazon_result(result);
+
     return result;
   } catch (err) {
     ok = false;
@@ -32,8 +33,7 @@ export async function execute_action_and_record(params) {
       ok,
       result_payload,
       error_message,
-      source,
-      schedule_task_id: schedule_task_id || null
+      source
     });
   }
 }