benchmark-collector 1.5.2 → 1.6.0

package/README.md

@@ -11,6 +11,7 @@
 - **Cookie 自动注入** — 从本地 Chrome 提取并解密 Cookie，保留登录态（支持 macOS / Windows / Linux）
 - **对话采集** — 支持 DOM 检测和 API 拦截两种模式采集客服对话消息，内置平台预设选择器自动匹配，消息自动去重
 - **平台预设** — 内置常见客服平台（抖店、拼多多等）的对话容器选择器，无需手动指定
+- **历史会话批量采集** — 内置抖音飞鸽 / 拼多多客服后台的历史会话自动采集脚本，免人工操作即可批量拉取后台数据
 - **长时间录制优化** — 写入节流、原子保存、CDP 自动清理，支持数小时连续采集
 - **增量保存** — 每步操作实时保存，进程异常退出也不丢数据
 - **跨平台** — 支持 macOS / Windows / Linux
@@ -147,6 +148,49 @@ output/处理退款_2026-03-22T10-00-00/
 | `*_dom.html` | 页面 DOM 快照（含 iframe 内容） | DOM-based agent 训练 |
 | `*_conversation.json` | 对话消息（角色 / 内容 / 时间戳 / 附件） | 客服对话训练数据（DOM 模式） |
 
+## 历史会话批量采集
+
+针对抖音飞鸽（抖店）和拼多多商家后台的历史会话页面，提供两个独立的批量采集脚本，自动操作页面、拦截接口、翻页累积数据。
+
+### 抖音飞鸽
+
+页面：`https://jsls.jinritemai.com/im/shop/main/data/historyConversation`，接口：`fuzzySearchConversation`。
+
+```bash
+# 默认运行（自动注入本地 Chrome Cookie，有头模式）
+npm run feige-history
+
+# 自定义参数
+npx ts-node scripts/feige-history-collect.ts \
+  --max-pages 100 \   # 最大账号列表翻页数（默认 100）
+  --delay 1000 \      # 每次翻页间隔，避免限流（默认 1000ms）
+  --output ./data \   # 输出目录（默认 output）
+  --headless          # 无头模式
+```
+
+流程：打开页面 → 选"近30天" → 点击查询 → 拦截 `fuzzySearchConversation` → 自动翻页直到采集全部 → 保存到 `output/feige-history_<timestamp>.json`。
+
+### 拼多多商家后台
+
+页面：`https://mms.pinduoduo.com/mms-chat/search?msfrom=mms_sidenav`，接口：`getMessages`。
+
+```bash
+# 默认运行
+npm run pdd-history
+
+# 自定义参数
+npx ts-node scripts/pdd-history-collect.ts \
+  --max-accounts 200 \    # 最多采集消费者账号数（默认 500）
+  --max-msg-pages 20 \    # 每个账号最多翻多少页消息（默认 20）
+  --delay 600 \           # 操作间隔（默认 600ms）
+  --output ./data \
+  --headless
+```
+
+流程：打开页面 → 选"近30天" → 点击查询 → 遍历左侧 `.user-item` 消费者账号列表 → 每个账号点击后遍历右侧消息分页 → 拦截 `getMessages` 累积全部消息 → 保存到 `output/pdd-history_<timestamp>.json`。
+
+> **前置条件：** 需要先在本地 Chrome 中登录对应的商家后台，脚本会自动从 Chrome 解密提取 Cookie 完成认证。
+
 ## 回放与调试
 
 ```bash

package/dist/scripts/feige-history-collect.js

@@ -0,0 +1,375 @@
+#!/usr/bin/env node
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+/**
+ * 抖音飞鸽历史会话数据采集脚本
+ *
+ * 功能：
+ *   1. 打开飞鸽数据页面 (historyConversation)
+ *   2. 注入 Cookie 实现免登录
+ *   3. 点击"查询"触发数据加载
+ *   4. 拦截 fuzzySearchConversation 接口响应
+ *   5. 自动翻页直到获取所有数据
+ *   6. 保存完整数据到本地 JSON 文件
+ *
+ * 用法：
+ *   npx ts-node scripts/feige-history-collect.ts [--output <dir>] [--max-pages <n>] [--delay <ms>]
+ */
+const playwright_core_1 = require("playwright-core");
+const fs = __importStar(require("fs"));
+const path = __importStar(require("path"));
+const cookie_extract_1 = require("../src/cookie-extract");
+const PAGE_URL = 'https://jsls.jinritemai.com/im/shop/main/data/historyConversation';
+const API_PATTERN = 'fuzzySearchConversation';
+// ========== CLI 参数解析 ==========
+const args = process.argv.slice(2);
+function getArg(name, defaultValue) {
+    const idx = args.indexOf(`--${name}`);
+    return idx >= 0 && args[idx + 1] ? args[idx + 1] : defaultValue;
+}
+function hasFlag(name) {
+    return args.includes(`--${name}`);
+}
+const outputDir = getArg('output', 'output');
+const maxPages = parseInt(getArg('max-pages', '100'), 10);
+const delayMs = parseInt(getArg('delay', '1000'), 10);
+const headless = hasFlag('headless');
+// ========== 主逻辑 ==========
+class FeigeHistoryCollector {
+    constructor() {
+        this.allConversations = [];
+        this.currentPageData = [];
+        this.responseReceived = false;
+        this.totalFromApi = 0;
+    }
+    async run() {
+        console.log('飞鸽历史会话数据采集');
+        console.log(`  页面: ${PAGE_URL}`);
+        console.log(`  最大页数: ${maxPages}`);
+        console.log(`  请求间隔: ${delayMs}ms`);
+        console.log(`  输出目录: ${outputDir}`);
+        console.log('');
+        const launchOpts = {
+            headless,
+            handleSIGINT: false,
+            handleSIGTERM: false,
+            handleSIGHUP: false,
+        };
+        let browser;
+        try {
+            browser = await playwright_core_1.chromium.launch({ ...launchOpts, channel: 'chrome' });
+        }
+        catch {
+            browser = await playwright_core_1.chromium.launch(launchOpts);
+        }
+        const context = await browser.newContext({ viewport: { width: 1440, height: 900 } });
+        // 注入 Cookie
+        console.log('🍪 从本地 Chrome 提取 Cookie...');
+        const cookies = (0, cookie_extract_1.extractCookiesForUrl)(PAGE_URL);
+        if (cookies.length > 0) {
+            let injected = 0;
+            for (const cookie of cookies) {
+                try {
+                    await context.addCookies([cookie]);
+                    injected++;
+                }
+                catch { }
+            }
+            console.log(`  ✅ 已注入 ${injected}/${cookies.length} 个 Cookie`);
+        }
+        else {
+            console.log('  ⚠️ 未找到 Cookie，可能需要先在 Chrome 中登录飞鸽');
+        }
+        this.page = await context.newPage();
+        await this.setupApiIntercept();
+        // 导航到页面
+        console.log('\n🌐 正在打开飞鸽数据页面...');
+        try {
+            await this.page.goto(PAGE_URL, { waitUntil: 'networkidle', timeout: 30000 });
+        }
+        catch (e) {
+            console.log(`  ⚠️ 页面加载超时，继续尝试: ${e.message?.split('\n')[0]}`);
+        }
+        await this.page.waitForTimeout(2000);
+        // 点击"近30天"快捷按钮，设置时间范围
+        console.log('\n📅 点击"近30天"按钮...');
+        const rangeSelected = await this.clickLast30DaysButton();
+        if (!rangeSelected) {
+            console.log('  ⚠️ 未找到"近30天"按钮，使用默认时间范围');
+        }
+        else {
+            console.log('  ✅ 已选择"近30天"');
+            await this.page.waitForTimeout(500);
+        }
+        // 点击"查询"按钮
+        console.log('\n🔍 点击查询按钮...');
+        this.responseReceived = false;
+        const searched = await this.clickSearchButton();
+        if (!searched) {
+            console.log('  ⚠️ 未找到查询按钮，尝试等待数据自动加载...');
+        }
+        // 等待第一次 API 响应
+        await this.waitForApiResponse(10000);
+        if (!this.responseReceived) {
+            console.log('  ⚠️ 未收到 API 响应，请确认页面已正确登录');
+            await browser.close();
+            process.exit(1);
+        }
+        // 自动翻页
+        let pageNum = 1;
+        console.log(`\n📄 开始翻页采集 (已获取第 1 页)...`);
+        while (pageNum < maxPages) {
+            if (this.totalFromApi > 0 && this.allConversations.length >= this.totalFromApi) {
+                console.log(`  ✅ 已获取全部 ${this.totalFromApi} 条数据`);
+                break;
+            }
+            if (this.currentPageData.length === 0 && pageNum > 1) {
+                console.log(`  ✅ 当前页无数据，采集结束`);
+                break;
+            }
+            this.responseReceived = false;
+            this.currentPageData = [];
+            const hasNext = await this.clickNextPage();
+            if (!hasNext) {
+                console.log(`  ✅ 没有下一页了，采集结束`);
+                break;
+            }
+            pageNum++;
+            await this.waitForApiResponse(10000);
+            if (!this.responseReceived) {
+                console.log(`  ⚠️ 第 ${pageNum} 页未收到响应，停止翻页`);
+                break;
+            }
+            await this.page.waitForTimeout(delayMs);
+        }
+        if (pageNum >= maxPages) {
+            console.log(`  ⚠️ 达到最大页数限制 (${maxPages})`);
+        }
+        // 保存数据
+        const result = {
+            conversations: this.allConversations,
+            totalPages: pageNum,
+            totalConversations: this.allConversations.length,
+            collectTime: new Date().toISOString(),
+        };
+        fs.mkdirSync(outputDir, { recursive: true });
+        const timestamp = new Date().toISOString().replace(/[:.]/g, '-').slice(0, 19);
+        const outputFile = path.join(outputDir, `feige-history_${timestamp}.json`);
+        fs.writeFileSync(outputFile, JSON.stringify(result, null, 2));
+        console.log(`\n✅ 采集完成`);
+        console.log(`   共 ${this.allConversations.length} 条会话, ${pageNum} 页`);
+        console.log(`   文件: ${outputFile}`);
+        await browser.close();
+    }
+    /** 设置 API 拦截 */
+    async setupApiIntercept() {
+        await this.page.route(`**/${API_PATTERN}*`, async (route) => {
+            const response = await route.fetch();
+            const body = await response.text();
+            try {
+                const data = JSON.parse(body);
+                const conversations = this.extractConversationList(data);
+                this.totalFromApi = this.extractTotal(data);
+                if (conversations.length > 0) {
+                    this.currentPageData = conversations;
+                    this.allConversations.push(...conversations);
+                    console.log(`  📡 获取到 ${conversations.length} 条会话 (累计: ${this.allConversations.length}/${this.totalFromApi || '?'})`);
+                }
+                else {
+                    this.currentPageData = [];
+                    console.log(`  📡 该页无数据`);
+                }
+            }
+            catch (e) {
+                console.log(`  ⚠️ 解析响应失败: ${e.message}`);
+                this.currentPageData = [];
+            }
+            this.responseReceived = true;
+            await route.fulfill({ response });
+        });
+    }
+    /** 等待 API 响应（轮询 responseReceived 标志） */
+    async waitForApiResponse(timeout) {
+        const start = Date.now();
+        while (Date.now() - start < timeout && !this.responseReceived) {
+            await this.page.waitForTimeout(200);
+        }
+        await this.page.waitForTimeout(300);
+    }
+    /** 点击"近30天"快捷按钮（设置时间范围） */
+    async clickLast30DaysButton() {
+        const selectors = [
+            'button:has-text("近30天")',
+            'button:has-text("最近30天")',
+            'button:has-text("近三十天")',
+            'span:has-text("近30天")',
+            'a:has-text("近30天")',
+            'div[role="button"]:has-text("近30天")',
+            '[class*="tag"]:has-text("近30天")',
+            '[class*="quick"]:has-text("近30天")',
+            'label:has-text("近30天")',
+            // 兜底：包含"30天"文本的可点击元素
+            'button:has-text("30天")',
+            'span[class*="btn"]:has-text("30天")',
+        ];
+        for (const selector of selectors) {
+            try {
+                const btn = this.page.locator(selector).first();
+                if (await btn.isVisible({ timeout: 1000 })) {
+                    await btn.click();
+                    return true;
+                }
+            }
+            catch { }
+        }
+        return false;
+    }
+    /** 点击查询按钮 */
+    async clickSearchButton() {
+        const selectors = [
+            'button:has-text("查询")',
+            'button:has-text("搜索")',
+            'button:has-text("查找")',
+            '.ant-btn-primary:has-text("查询")',
+            '.ant-btn-primary:has-text("搜索")',
+            '[class*="search"] button',
+            '[class*="query"] button',
+            'button[type="submit"]',
+        ];
+        for (const selector of selectors) {
+            try {
+                const btn = this.page.locator(selector).first();
+                if (await btn.isVisible({ timeout: 1000 })) {
+                    await btn.click();
+                    return true;
+                }
+            }
+            catch { }
+        }
+        return false;
+    }
+    /** 点击下一页按钮 */
+    async clickNextPage() {
+        const selectors = [
+            '.ant-pagination-next:not(.ant-pagination-disabled)',
+            'li.ant-pagination-next:not(.ant-pagination-disabled) button',
+            'li.ant-pagination-next:not(.ant-pagination-disabled) a',
+            'button:has-text("下一页")',
+            'a:has-text("下一页")',
+            '.arco-pagination-item-next:not(.arco-pagination-item-disabled)',
+            '[class*="next"]:not([class*="disabled"]):not([disabled])',
+        ];
+        for (const selector of selectors) {
+            try {
+                const btn = this.page.locator(selector).first();
+                if (await btn.isVisible({ timeout: 1000 })) {
+                    const disabled = await btn.getAttribute('disabled');
+                    const ariaDisabled = await btn.getAttribute('aria-disabled');
+                    if (disabled !== null || ariaDisabled === 'true')
+                        continue;
+                    await btn.click();
+                    return true;
+                }
+            }
+            catch { }
+        }
+        return false;
+    }
+    /** 从 API 响应中提取会话列表 */
+    extractConversationList(data) {
+        const paths = [
+            data?.data?.conversations,
+            data?.data?.list,
+            data?.data?.data,
+            data?.data?.records,
+            data?.data?.items,
+            data?.result?.conversations,
+            data?.result?.list,
+            data?.result?.data,
+            data?.conversations,
+            data?.list,
+        ];
+        for (const p of paths) {
+            if (Array.isArray(p) && p.length > 0)
+                return p;
+        }
+        return this.findFirstArray(data, 0);
+    }
+    /** 递归查找第一个看起来像会话列表的数组 */
+    findFirstArray(obj, depth) {
+        if (depth > 5 || !obj || typeof obj !== 'object')
+            return [];
+        if (Array.isArray(obj) && obj.length > 0 && typeof obj[0] === 'object')
+            return obj;
+        for (const val of Object.values(obj)) {
+            if (Array.isArray(val) && val.length > 0 && typeof val[0] === 'object') {
+                return val;
+            }
+        }
+        for (const val of Object.values(obj)) {
+            if (val && typeof val === 'object' && !Array.isArray(val)) {
+                const found = this.findFirstArray(val, depth + 1);
+                if (found.length > 0)
+                    return found;
+            }
+        }
+        return [];
+    }
+    /** 从响应中提取总数 */
+    extractTotal(data) {
+        const paths = [
+            data?.data?.total,
+            data?.data?.totalCount,
+            data?.data?.total_count,
+            data?.data?.pagination?.total,
+            data?.data?.page_info?.total,
+            data?.result?.total,
+            data?.total,
+        ];
+        for (const p of paths) {
+            if (typeof p === 'number' && p > 0)
+                return p;
+        }
+        return 0;
+    }
+}
+// ========== 启动 ==========
+const collector = new FeigeHistoryCollector();
+collector.run().catch((err) => {
+    console.error('采集出错:', err);
+    process.exit(1);
+});
+//# sourceMappingURL=feige-history-collect.js.map

package/dist/scripts/pdd-history-collect.js

@@ -0,0 +1,499 @@
+#!/usr/bin/env node
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+Object.defineProperty(exports, "__esModule", { value: true });
+/**
+ * 拼多多客服后台历史会话数据采集脚本
+ *
+ * 页面：https://mms.pinduoduo.com/mms-chat/search?msfrom=mms_sidenav
+ * 接口：POST https://mms.pinduoduo.com/latitude/search/message/getMessages
+ *
+ * 实际页面结构（已通过浏览器验证）：
+ *   - 顶部表单：查询方式 / 客服账号 / 时间 / 查询按钮
+ *   - 时间输入框 placeholder="开始时间  ~  结束时间", readonly
+ *   - 时间浮层快捷按钮：今天 / 近7天 / 近30天 (点击后自动应用并关闭浮层，无需确认)
+ *   - 查询按钮：button.BTN_primary_xxx 文字"查询"
+ *   - 查询结果：左侧账号列表 .cs-list > .user-item (每页 10 条) + 右侧 .message-list
+ *   - 分页器：.PGT_outerWrapper_xxx，下一页 .PGT_next_xxx
+ *   - 接口返回的消息一次性全量渲染到 .message-list，无懒加载
+ *
+ * 流程：
+ *   1. 注入 Cookie 实现免登录
+ *   2. 打开页面
+ *   3. 点击时间输入框 → 点击"近30天" (自动应用 + 关闭浮层)
+ *   4. 点击"查询" → 左侧出现 .user-item 列表
+ *   5. 遍历当前页所有 .user-item，逐个点击 → 拦截 getMessages 响应
+ *   6. 点击分页器"下一页"，重复步骤 5，直到没有下一页
+ *   7. 增量保存到 output/pdd-history_<timestamp>.json
+ *
+ * 用法：
+ *   npx ts-node scripts/pdd-history-collect.ts [--output <dir>] [--max-accounts <n>]
+ *     [--delay <ms>] [--headless]
+ */
+const playwright_core_1 = require("playwright-core");
+const fs = __importStar(require("fs"));
+const path = __importStar(require("path"));
+const cookie_extract_1 = require("../src/cookie-extract");
+const PAGE_URL = 'https://mms.pinduoduo.com/mms-chat/search?msfrom=mms_sidenav';
+const API_PATTERN = '/latitude/search/message/getMessages';
+// ========== CLI 参数 ==========
+const args = process.argv.slice(2);
+const getArg = (name, def) => {
+    const idx = args.indexOf(`--${name}`);
+    return idx >= 0 && args[idx + 1] ? args[idx + 1] : def;
+};
+const hasFlag = (name) => args.includes(`--${name}`);
+const outputDir = getArg('output', 'output');
+const maxAccounts = parseInt(getArg('max-accounts', '500'), 10);
+const maxMsgPages = parseInt(getArg('max-msg-pages', '20'), 10);
+const delayMs = parseInt(getArg('delay', '600'), 10);
+const headless = hasFlag('headless');
+class PddHistoryCollector {
+    constructor() {
+        this.currentMessages = new Map(); // 当前账号的消息（按 msg_id 去重）
+        this.currentResponseCount = 0;
+        this.results = [];
+        this.outputFile = '';
+    }
+    async run() {
+        console.log('拼多多客服历史会话采集');
+        console.log(`  页面: ${PAGE_URL}`);
+        console.log(`  最大账号数: ${maxAccounts}`);
+        console.log(`  每账号最大消息分页: ${maxMsgPages}`);
+        console.log(`  请求间隔: ${delayMs}ms`);
+        console.log(`  输出目录: ${outputDir}`);
+        console.log('');
+        fs.mkdirSync(outputDir, { recursive: true });
+        const ts = new Date().toISOString().replace(/[:.]/g, '-').slice(0, 19);
+        this.outputFile = path.join(outputDir, `pdd-history_${ts}.json`);
+        const launchOpts = {
+            headless,
+            handleSIGINT: false,
+            handleSIGTERM: false,
+            handleSIGHUP: false,
+        };
+        let browser;
+        try {
+            browser = await playwright_core_1.chromium.launch({ ...launchOpts, channel: 'chrome' });
+        }
+        catch {
+            browser = await playwright_core_1.chromium.launch(launchOpts);
+        }
+        const context = await browser.newContext({ viewport: { width: 1440, height: 900 } });
+        // Cookie 注入
+        console.log('🍪 从本地 Chrome 提取 Cookie...');
+        const cookies = (0, cookie_extract_1.extractCookiesForUrl)(PAGE_URL);
+        if (cookies.length > 0) {
+            let injected = 0;
+            for (const c of cookies) {
+                try {
+                    await context.addCookies([c]);
+                    injected++;
+                }
+                catch { }
+            }
+            console.log(`  ✅ 已注入 ${injected}/${cookies.length} 个 Cookie`);
+        }
+        else {
+            console.log('  ⚠️ 未找到 Cookie，请先在 Chrome 中登录拼多多商家后台');
+        }
+        this.page = await context.newPage();
+        await this.setupApiIntercept();
+        // 1. 导航
+        console.log('\n🌐 打开页面...');
+        try {
+            await this.page.goto(PAGE_URL, { waitUntil: 'networkidle', timeout: 30000 });
+        }
+        catch (e) {
+            console.log(`  ⚠️ 加载超时，继续: ${e.message?.split('\n')[0]}`);
+        }
+        await this.page.waitForTimeout(2000);
+        // 2. 选择"近30天"
+        console.log('\n📅 选择时间范围"近30天"...');
+        const ok30 = await this.selectLast30Days();
+        if (!ok30) {
+            console.log('  ⚠️ 时间选择失败，使用页面默认');
+        }
+        else {
+            console.log('  ✅ 已选择近30天');
+        }
+        // 3. 点击查询
+        console.log('\n🔍 点击查询按钮...');
+        const searched = await this.clickSearchButton();
+        if (!searched) {
+            console.log('  ❌ 未找到查询按钮');
+            await browser.close();
+            process.exit(1);
+        }
+        // 等待左侧 user-item 列表出现
+        try {
+            await this.page.waitForSelector('.user-item', { timeout: 15000 });
+        }
+        catch {
+            console.log('  ⚠️ 等待 .user-item 超时，可能查询无结果');
+        }
+        await this.page.waitForTimeout(1500);
+        // 4. 遍历左侧账号列表（左侧没有分页器，可能通过滚动加载；这里先取全部已渲染项，
+        //    若数量增长则继续滚动加载更多）
+        console.log('\n👥 开始遍历消费者账号...');
+        let processed = 0;
+        let lastSeenCount = 0;
+        let stableRounds = 0;
+        while (processed < maxAccounts) {
+            const itemCount = await this.page.locator('.user-item').count();
+            if (itemCount === 0) {
+                console.log(`  ⚠️ 无账号，结束`);
+                break;
+            }
+            // 处理新出现的账号
+            for (let i = lastSeenCount; i < itemCount && processed < maxAccounts; i++) {
+                await this.processOneAccount(i, 1, processed + 1);
+                processed++;
+                this.saveResults();
+                await this.page.waitForTimeout(delayMs);
+            }
+            lastSeenCount = itemCount;
+            if (processed >= maxAccounts) {
+                console.log(`  ⚠️ 达到最大账号数限制 (${maxAccounts})`);
+                break;
+            }
+            // 滚动 .cs-list 到底部尝试加载更多
+            const grew = await this.scrollAccountListAndCheck();
+            if (!grew) {
+                stableRounds++;
+                if (stableRounds >= 2) {
+                    console.log('  ✅ 账号列表无更多项，采集结束');
+                    break;
+                }
+            }
+            else {
+                stableRounds = 0;
+            }
+            await this.page.waitForTimeout(800);
+        }
+        this.saveResults();
+        const total = this.results.reduce((s, r) => s + r.messages.length, 0);
+        console.log(`\n✅ 采集完成`);
+        console.log(`   共 ${this.results.length} 个账号, ${total} 条消息`);
+        console.log(`   文件: ${this.outputFile}`);
+        await browser.close();
+    }
+    // ========== 单个账号 ==========
+    async processOneAccount(itemIdx, pageIdx, globalIdx) {
+        this.currentMessages.clear();
+        this.currentResponseCount = 0;
+        const items = this.page.locator('.user-item');
+        if (itemIdx >= (await items.count()))
+            return;
+        const item = items.nth(itemIdx);
+        const label = (await item.innerText().catch(() => '')).trim().replace(/\s+/g, ' ').slice(0, 80);
+        console.log(`\n[#${globalIdx} 页${pageIdx}/项${itemIdx + 1}] ${label || '(无文本)'}`);
+        try {
+            await item.scrollIntoViewIfNeeded({ timeout: 3000 });
+        }
+        catch { }
+        // 第一页第一个账号默认已被选中并已加载消息（等于无需点击就有响应）。
+        // 我们仍然点击一次，强制重新触发请求，以保证拦截能稳定捕获。
+        try {
+            await item.click({ timeout: 5000 });
+        }
+        catch (e) {
+            console.log(`  ⚠️ 点击失败: ${e.message?.split('\n')[0]}`);
+            return;
+        }
+        // 等 getMessages 响应（首页第一项可能没有新请求 — 因为页面初始化时已请求过；做兜底）
+        const got = await this.waitForResponse(8000);
+        if (!got && pageIdx === 1 && itemIdx === 0) {
+            console.log(`  ℹ️ 首项无新请求，跳过`);
+        }
+        // 遍历该账号的消息分页（拼多多右侧消息分页，每个账号页数不同）
+        const msgPagesVisited = await this.iterateMessagePages();
+        const messages = Array.from(this.currentMessages.values());
+        this.results.push({
+            accountKey: `p${pageIdx}_i${itemIdx + 1}`,
+            accountLabel: label,
+            pageIndex: pageIdx,
+            itemIndex: itemIdx + 1,
+            messagePagesVisited: msgPagesVisited,
+            messages,
+            responseCount: this.currentResponseCount,
+        });
+        console.log(`  ✅ ${messages.length} 条消息 (${this.currentResponseCount} 次响应)`);
+    }
+    // ========== 时间选择 ==========
+    /** 选择"近30天"（点击输入框 → 点近30天，浮层会自动关闭） */
+    async selectLast30Days() {
+        const input = this.page.locator('input[placeholder*="开始时间"][readonly]').first();
+        try {
+            if (!(await input.isVisible({ timeout: 3000 })))
+                return false;
+            await input.click({ timeout: 2000 });
+        }
+        catch {
+            return false;
+        }
+        // 等浮层
+        try {
+            await this.page.waitForSelector('.RPR_quickPickerWrapper_1198e34, [class*="RPR_quickPickerWrapper"]', { timeout: 3000 });
+        }
+        catch {
+            // 浮层 class 可能变化，继续尝试找按钮
+        }
+        await this.page.waitForTimeout(300);
+        // 点击"近30天"
+        const last30 = this.page.locator('button', { hasText: /^近30天$/ }).first();
+        try {
+            if (await last30.isVisible({ timeout: 2000 })) {
+                await last30.click({ timeout: 2000 });
+                await this.page.waitForTimeout(400);
+                return true;
+            }
+        }
+        catch { }
+        return false;
+    }
+    // ========== 查询按钮 ==========
+    async clickSearchButton() {
+        // 优先 BTN_primary 中的"查询"
+        const candidates = [
+            this.page.locator('button[class*="BTN_primary"]', { hasText: /^查询$/ }),
+            this.page.locator('button', { hasText: /^查询$/ }),
+        ];
+        for (const c of candidates) {
+            try {
+                const btn = c.first();
+                if (await btn.isVisible({ timeout: 1500 })) {
+                    await btn.click({ timeout: 2000 });
+                    return true;
+                }
+            }
+            catch { }
+        }
+        return false;
+    }
+    // ========== 左侧账号列表滚动加载 ==========
+    /** 滚动 .cs-list 到底部，返回是否新增了项 */
+    async scrollAccountListAndCheck() {
+        return await this.page.evaluate(() => {
+            const list = document.querySelector('.cs-list');
+            if (!list)
+                return false;
+            const before = list.querySelectorAll('.user-item').length;
+            list.scrollTop = list.scrollHeight;
+            return new Promise(resolve => {
+                setTimeout(() => {
+                    const after = list.querySelectorAll('.user-item').length;
+                    resolve(after > before);
+                }, 1500);
+            });
+        });
+    }
+    // ========== 右侧消息分页（每个账号内部） ==========
+    /**
+     * 遍历当前选中账号的所有消息分页。
+     * 拼多多分页器结构：
+     *   - .table-section-right-footer 内的 [class*="PGT_outerWrapper"]
+     *   - 当前激活：[class*="PGT_pagerItemActive"]
+     *   - 下一页按钮：[class*="PGT_next"]，禁用时含 PGT_disabled
+     *
+     * 策略：循环点击下一页，直到 next 按钮 disabled 或没有新响应。
+     * 返回访问过的页数（含第 1 页）。
+     */
+    async iterateMessagePages() {
+        let visited = 1; // 当前已经在第 1 页
+        while (visited < maxMsgPages) {
+            // 检查右侧分页器是否存在 + 下一页是否可用
+            const status = await this.page.evaluate(() => {
+                const footer = document.querySelector('.table-section-right-footer');
+                if (!footer)
+                    return { exists: false, disabled: true };
+                const next = footer.querySelector('[class*="PGT_next"]');
+                if (!next)
+                    return { exists: false, disabled: true };
+                const cls = next.className || '';
+                const ariaDisabled = next.getAttribute('aria-disabled');
+                const r = next.getBoundingClientRect();
+                return {
+                    exists: r.width > 0 && r.height > 0,
+                    disabled: /PGT_disabled|disabled/.test(cls) || ariaDisabled === 'true',
+                };
+            });
+            if (!status.exists || status.disabled)
+                break;
+            // 点击下一页
+            const baseline = this.currentResponseCount;
+            const clicked = await this.page.evaluate(() => {
+                const footer = document.querySelector('.table-section-right-footer');
+                if (!footer)
+                    return false;
+                const next = footer.querySelector('[class*="PGT_next"]');
+                if (!next)
+                    return false;
+                next.click();
+                return true;
+            });
+            if (!clicked)
+                break;
+            // 等响应
+            const got = await this.waitForResponseSince(baseline, 8000);
+            if (!got) {
+                // 如果没新响应（可能数据已缓存或异常），仍然让 visited++ 但只重试 1 次
+                console.log(`  ⚠️ 第 ${visited + 1} 页消息无响应，停止该账号翻页`);
+                break;
+            }
+            visited++;
+            console.log(`    📄 已翻到消息第 ${visited} 页`);
+            await this.page.waitForTimeout(delayMs);
+        }
+        if (visited >= maxMsgPages) {
+            console.log(`  ⚠️ 该账号达到消息分页上限 ${maxMsgPages}，停止翻页`);
+        }
+        return visited;
+    }
+    // ========== API 拦截 ==========
+    async setupApiIntercept() {
+        await this.page.route(`**${API_PATTERN}*`, async (route) => {
+            const response = await route.fetch();
+            const body = await response.text();
+            try {
+                const data = JSON.parse(body);
+                const messages = this.extractMessages(data);
+                for (const m of messages) {
+                    const id = this.getMessageId(m);
+                    if (id && !this.currentMessages.has(id)) {
+                        this.currentMessages.set(id, m);
+                    }
+                }
+                this.currentResponseCount++;
+                if (messages.length > 0) {
+                    console.log(`  📡 +${messages.length} 条消息 (累计 ${this.currentMessages.size})`);
+                }
+            }
+            catch (e) {
+                console.log(`  ⚠️ 解析响应失败: ${e.message?.split('\n')[0]}`);
+            }
+            await route.fulfill({ response });
+        });
+    }
+    /** 等待至少一次新的 getMessages 响应 */
+    async waitForResponse(timeout) {
+        return this.waitForResponseSince(this.currentResponseCount, timeout);
+    }
+    /** 从指定 baseline 等待新响应 */
+    async waitForResponseSince(baseline, timeout) {
+        const start = Date.now();
+        while (Date.now() - start < timeout) {
+            if (this.currentResponseCount > baseline) {
+                await this.page.waitForTimeout(300); // 等数据落地
+                return true;
+            }
+            await this.page.waitForTimeout(150);
+        }
+        return false;
+    }
+    // ========== 响应解析 ==========
+    extractMessages(data) {
+        const paths = [
+            data?.result?.messages,
+            data?.result?.list,
+            data?.result?.data,
+            data?.result?.message_list,
+            data?.result?.records,
+            data?.result?.msg_list,
+            data?.data?.messages,
+            data?.data?.list,
+            data?.data?.message_list,
+            data?.messages,
+            data?.list,
+        ];
+        for (const p of paths) {
+            if (Array.isArray(p) && p.length > 0)
+                return p;
+        }
+        return this.findMessageArray(data, 0);
+    }
+    findMessageArray(obj, depth) {
+        if (depth > 6 || !obj || typeof obj !== 'object')
+            return [];
+        if (Array.isArray(obj) && obj.length > 0 && this.looksLikeMessages(obj))
+            return obj;
+        for (const val of Object.values(obj)) {
+            if (Array.isArray(val) && val.length > 0 && this.looksLikeMessages(val)) {
+                return val;
+            }
+        }
+        for (const val of Object.values(obj)) {
+            if (val && typeof val === 'object' && !Array.isArray(val)) {
+                const found = this.findMessageArray(val, depth + 1);
+                if (found.length > 0)
+                    return found;
+            }
+        }
+        return [];
+    }
+    looksLikeMessages(arr) {
+        const sample = arr[0];
+        if (!sample || typeof sample !== 'object')
+            return false;
+        return (sample.msg_id !== undefined ||
+            sample.message_id !== undefined ||
+            sample.msgId !== undefined ||
+            sample.id !== undefined ||
+            sample.content !== undefined ||
+            sample.from !== undefined ||
+            sample.ts !== undefined);
+    }
+    getMessageId(m) {
+        return String(m?.msg_id ?? m?.message_id ?? m?.msgId ?? m?.id ?? m?.seq ?? m?.sequence ??
+            `auto_${JSON.stringify(m).slice(0, 200)}`);
+    }
+    // ========== 持久化 ==========
+    saveResults() {
+        const data = {
+            collectTime: new Date().toISOString(),
+            totalAccounts: this.results.length,
+            totalMessages: this.results.reduce((s, r) => s + r.messages.length, 0),
+            accounts: this.results,
+        };
+        fs.writeFileSync(this.outputFile, JSON.stringify(data, null, 2));
+    }
+}
+// ========== 启动 ==========
+const collector = new PddHistoryCollector();
+collector.run().catch((err) => {
+    console.error('采集出错:', err);
+    process.exit(1);
+});
+//# sourceMappingURL=pdd-history-collect.js.map

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "benchmark-collector",
-  "version": "1.5.2",
+  "version": "1.6.0",
   "description": "Playwright-based benchmark data collector for web agent training",
   "main": "dist/src/index.js",
   "types": "dist/src/index.d.ts",
@@ -18,6 +18,8 @@
   ],
   "scripts": {
     "collect": "ts-node scripts/collect.ts",
+    "feige-history": "ts-node scripts/feige-history-collect.ts",
+    "pdd-history": "ts-node scripts/pdd-history-collect.ts",
     "build": "tsc",
     "prepublishOnly": "npm run build",
     "postinstall": "node dist/scripts/postinstall.js",