feat(extension): add rendered page collection fallback

README.md

@@ -3,6 +3,7 @@
 AiToEarn 自维护的 Chrome/Edge 扩展，也是 Gitea Release 唯一允许打包发布的扩展源码，用于：
 
 - 从当前网页采集标题、正文和图片。
+- 在普通 HTTP 采集不足时复用当前 Chrome 标签页，等待动态页面渲染后回传正文、HTML、图片和链接。
 - 在 AiToEarn 页面提供 `window.AIToEarnDistribution` 分发桥接。
 - 复用浏览器现有登录状态检测内容平台账号。
 - 将同一份 Markdown、HTML 或纯文本内容填入平台编辑器。
@@ -12,6 +13,8 @@ AiToEarn 自维护的 Chrome/Edge 扩展，也是 Gitea Release 唯一允许打
 
 扩展弹窗和后台通知中的“成功”只表示平台编辑器已打开并完成内容填充，不表示内容已经公开发布。右键菜单“提取此页面到 AiToEarn”会把最近一次提取结果暂存到浏览器本地，随后打开扩展即可继续选择平台。
 
+AiToEarn 后台触发浏览器采集时会优先复用同地址标签页；没有现成标签页时才在当前 Chrome 内创建后台标签页。普通采集完成后临时标签页会关闭，遇到登录、验证码或地区访问限制时则保留并激活页面，等待用户处理。
+
 ## 上游来源
 
 多平台检测与编辑器填充逻辑基于 `doocs/cose`，固定版本和许可证见：

background.js

@@ -1,6 +1,112 @@
 import './distribution/cose/background.js';
 
 const PREPARE_BATCH_MESSAGE = 'PREPARE_PLATFORM_BATCH';
+const CAPTURE_URL_MESSAGE = 'CAPTURE_URL';
+const CAPTURE_ATTEMPTS = 8;
+
+function sleep(milliseconds) {
+  return new Promise(resolve => setTimeout(resolve, milliseconds));
+}
+
+function normalizeCaptureUrl(value) {
+  const url = new URL(value);
+  if (!['http:', 'https:'].includes(url.protocol)) throw new Error('只支持采集 HTTP 或 HTTPS 页面');
+  if (url.username || url.password) throw new Error('采集地址不能包含账号或密码');
+  url.hash = '';
+  return url;
+}
+
+function comparableUrl(value) {
+  try {
+    return normalizeCaptureUrl(value).href.replace(/\/$/, '');
+  } catch {
+    return '';
+  }
+}
+
+async function waitForTabReady(tabId) {
+  const deadline = Date.now() + 30000;
+  while (Date.now() < deadline) {
+    const tab = await chrome.tabs.get(tabId);
+    if (tab.status === 'complete') return;
+    await sleep(300);
+  }
+  throw new Error('页面加载超时');
+}
+
+async function requestPageExtraction(tabId) {
+  try {
+    return await chrome.tabs.sendMessage(tabId, { action: 'extractContent' });
+  } catch {
+    await chrome.scripting.executeScript({ target: { tabId }, files: ['content.js'] });
+    return chrome.tabs.sendMessage(tabId, { action: 'extractContent' });
+  }
+}
+
+async function captureUrl(rawUrl) {
+  const requestedUrl = normalizeCaptureUrl(rawUrl);
+  const requestedComparableUrl = comparableUrl(requestedUrl.href);
+  const tabs = await chrome.tabs.query({});
+  let tab = tabs.find(candidate => candidate.id && comparableUrl(candidate.url) === requestedComparableUrl);
+  const createdTab = !tab;
+  let keepTabOpen = false;
+
+  if (!tab) {
+    tab = await chrome.tabs.create({ url: requestedUrl.href, active: false });
+  }
+  if (!tab.id) throw new Error('无法创建或复用采集标签页');
+
+  try {
+    await waitForTabReady(tab.id);
+    let lastBodyLength = 0;
+    let lastTitle = '';
+
+    for (let attempt = 1; attempt <= CAPTURE_ATTEMPTS; attempt += 1) {
+      const response = await requestPageExtraction(tab.id);
+      if (!response?.success) throw new Error(response?.error || '扩展无法读取页面内容');
+
+      const data = response.data || {};
+      const accessState = data.accessState || { status: 'ok', message: '', requiresUserAction: false };
+      if (accessState.requiresUserAction) {
+        keepTabOpen = true;
+        await chrome.tabs.update(tab.id, { active: true });
+        return {
+          success: false,
+          error: accessState.message || '页面需要人工操作后才能继续采集',
+          requiresUserAction: true,
+          accessState,
+          tabId: tab.id,
+          reusedTab: !createdTab,
+        };
+      }
+
+      const bodyLength = typeof data.body === 'string' ? data.body.trim().length : 0;
+      const minimumLength = data.pageType === 'article' ? 120 : 60;
+      lastBodyLength = bodyLength;
+      lastTitle = data.title || '';
+      if (bodyLength >= minimumLength) {
+        return {
+          success: true,
+          data: {
+            ...data,
+            url: requestedUrl.href,
+            finalUrl: data.url || requestedUrl.href,
+          },
+          tabId: tab.id,
+          reusedTab: !createdTab,
+        };
+      }
+
+      await sleep(Math.min(2500, 600 + attempt * 250));
+    }
+
+    throw new Error(`页面已打开，但动态正文仍未渲染完成（标题：${lastTitle || '未知'}，正文 ${lastBodyLength} 字）`);
+  } finally {
+    if (createdTab && !keepTabOpen) {
+      await chrome.tabs.remove(tab.id).catch(() => undefined);
+    }
+  }
+}
 
 async function preparePlatformBatch(platformIds, content) {
   await chrome.runtime.sendMessage({ type: 'START_SYNC_BATCH' });
@@ -86,6 +192,22 @@ chrome.contextMenus.onClicked.addListener((info, tab) => {
 });
 
 chrome.runtime.onMessage.addListener((request, sender, sendResponse) => {
+  if (request.type === CAPTURE_URL_MESSAGE) {
+    if (typeof request.url !== 'string' || !request.url.trim()) {
+      sendResponse({ success: false, error: '缺少要采集的页面地址' });
+      return false;
+    }
+    captureUrl(request.url)
+      .then(sendResponse)
+      .catch((error) => {
+        sendResponse({
+          success: false,
+          error: error instanceof Error ? error.message : String(error),
+        });
+      });
+    return true;
+  }
+
   if (request.type !== PREPARE_BATCH_MESSAGE) return false;
 
   const platformIds = Array.isArray(request.platformIds)

content.js

@@ -3,17 +3,47 @@
 (() => {
   'use strict';
 
+  if (globalThis.__AI_TO_EARN_CONTENT_SCRIPT__) return;
+  globalThis.__AI_TO_EARN_CONTENT_SCRIPT__ = true;
+
+  const contentSelectors = [
+    'article',
+    '[role="main"]',
+    'main',
+    '.amos-land-page',
+    '.post-content',
+    '.article-content',
+    '.entry-content',
+    '.content-body',
+    '.markdown-body',
+    '.rich_media_content',
+    '#article-content',
+    '#content',
+    '.content',
+    '.post',
+    '.article',
+    '.story-body',
+    '#cnblogs_post_body',
+    '.htmledit_views',
+    '#article_content',
+    '.blog-content-box'
+  ];
+
   // ===== Content Extraction =====
   function extractPageContent() {
     const data = {
       title: '',
       description: '',
       body: '',
+      html: '',
       images: [],
+      links: [],
       url: window.location.href,
       siteName: '',
       author: '',
-      publishDate: ''
+      publishDate: '',
+      pageType: 'article',
+      accessState: { status: 'ok', message: '', requiresUserAction: false }
     };
 
     // --- Title ---
@@ -43,10 +73,15 @@
       || '';
 
     // --- Body Content ---
-    data.body = extractMainContent();
+    const mainContent = extractMainContent();
+    data.body = mainContent.text.slice(0, 100000);
+    data.html = mainContent.html.slice(0, 900000);
 
     // --- Images ---
-    data.images = extractImages();
+    data.images = extractImages(mainContent.element);
+    data.links = extractLinks(mainContent.element);
+    data.pageType = detectPageType();
+    data.accessState = detectAccessState(data.body);
 
     return data;
   }
@@ -58,34 +93,12 @@
 
   // ===== Main Content Extraction =====
   function extractMainContent() {
-    // Priority selectors for main content areas
-    const contentSelectors = [
-      'article',
-      '[role="main"]',
-      'main',
-      '.post-content',
-      '.article-content',
-      '.entry-content',
-      '.content-body',
-      '.markdown-body',
-      '#article-content',
-      '#content',
-      '.content',
-      '.post',
-      '.article',
-      '.story-body',
-      '#cnblogs_post_body',
-      '.htmledit_views',
-      '#article_content',
-      '.blog-content-box'
-    ];
-
     for (const selector of contentSelectors) {
       const el = document.querySelector(selector);
       if (el) {
         const text = cleanText(el.innerText);
         if (text.length > 50) {
-          return text;
+          return { text, html: el.outerHTML, element: el };
         }
       }
     }
@@ -97,16 +110,66 @@
         .map(p => p.textContent.trim())
         .filter(t => t.length > 20);
       if (texts.length > 0) {
-        return cleanText(texts.join('\n\n'));
+        return {
+          text: cleanText(texts.join('\n\n')),
+          html: `<main>${Array.from(paragraphs).map(p => p.outerHTML).join('')}</main>`,
+          element: document.body
+        };
       }
     }
 
     // Final fallback
-    return cleanText(document.body.innerText);
+    return {
+      text: cleanText(document.body?.innerText || ''),
+      html: document.body?.innerHTML || '',
+      element: document.body
+    };
+  }
+
+  function detectPageType() {
+    const route = `${window.location.pathname}${window.location.search}`.toLowerCase();
+    if (/\/trending\/|\/topic(?:\/|$)|topic_innerflow/.test(route)
+      || document.querySelector('.amos-land-page, [data-page-type="topic"]')) {
+      return 'topic';
+    }
+    if (/\/(?:search|tag|category|hot|feed)(?:\/|$)/.test(window.location.pathname.toLowerCase())
+      || document.querySelector('[data-page-type="listing"]')) {
+      return 'listing';
+    }
+    return 'article';
+  }
+
+  function detectAccessState(extractedBody) {
+    const visible = element => Boolean(element && (element.offsetWidth || element.offsetHeight || element.getClientRects().length));
+    const pageText = cleanText(document.body?.innerText || '').slice(0, 12000);
+    const lowerText = pageText.toLowerCase();
+    const captchaElement = Array.from(document.querySelectorAll(
+      '[class*="captcha"], [id*="captcha"], [class*="verify"], [id*="verify"], iframe[src*="captcha"]'
+    )).find(visible);
+    if (captchaElement || /(验证码|安全验证|人机验证|滑块验证|verify you are human|complete the security check|captcha)/i.test(pageText)) {
+      return { status: 'captcha_required', message: '页面需要完成验证码或安全验证', requiresUserAction: true };
+    }
+
+    const passwordInput = Array.from(document.querySelectorAll('input[type="password"]')).find(visible);
+    const loginRoute = /\/(?:login|signin|passport)(?:\/|$)/i.test(window.location.pathname);
+    if (passwordInput || (loginRoute && /(登录|登陆|sign in|log in)/i.test(pageText))) {
+      return { status: 'login_required', message: '页面需要登录后才能采集', requiresUserAction: true };
+    }
+
+    if (extractedBody.length < 120 && (
+      lowerText.includes('access denied')
+      || lowerText.includes('forbidden')
+      || lowerText.includes('not available in your region')
+      || /(所在地区|当前地区|地区限制|需要.*vpn|网络环境受限)/i.test(pageText)
+    )) {
+      return { status: 'access_restricted', message: '页面存在地区或网络访问限制', requiresUserAction: true };
+    }
+
+    return { status: 'ok', message: '', requiresUserAction: false };
   }
 
   // ===== Image Extraction =====
-  function extractImages() {
+  function extractImages(mainElement) {
     const images = new Set();
 
     // 1. Open Graph image
@@ -121,18 +184,17 @@
     // Skip favicons, they're not useful as content images
 
     // 4. Content area images (from article/main/content areas)
-    const contentSelectors = ['article', 'main', '.content', '.post-content', '#content'];
-    for (const selector of contentSelectors) {
-      const container = document.querySelector(selector);
-      if (container) {
-        container.querySelectorAll('img').forEach(img => {
-          const src = img.src || img.dataset.src || img.dataset.original;
-          if (src && !isSmallImage(img) && !isIconOrAvatar(src)) {
-            images.add(resolveUrl(src));
-          }
-        });
-        if (images.size > 0) break;
-      }
+    if (mainElement) {
+      mainElement.querySelectorAll('img').forEach(img => {
+        const src = img.currentSrc
+          || img.src
+          || img.dataset.src
+          || img.dataset.original
+          || img.dataset.lazySrc;
+        if (src && !isSmallImage(img) && !isIconOrAvatar(src)) {
+          images.add(resolveUrl(src));
+        }
+      });
     }
 
     // 5. If no content images found, scan all images
@@ -148,6 +210,15 @@
     return Array.from(images).slice(0, 10);
   }
 
+  function extractLinks(mainElement) {
+    const links = new Set();
+    (mainElement || document.body)?.querySelectorAll('a[href]').forEach(anchor => {
+      const href = resolveUrl(anchor.getAttribute('href'));
+      if (/^https?:\/\//i.test(href)) links.add(href);
+    });
+    return Array.from(links).slice(0, 100);
+  }
+
   function isSmallImage(img) {
     return (img.naturalWidth > 0 && img.naturalWidth < 100) ||
            (img.naturalHeight > 0 && img.naturalHeight < 100);

distribution-bridge.js

@@ -5,6 +5,7 @@
     'getVersion',
     'listPlatforms',
     'checkPlatforms',
+    'captureUrl',
     'startBatch',
     'syncToPlatform',
   ]);
@@ -33,6 +34,11 @@
         const response = await sendRuntimeMessage({ type: 'CHECK_PLATFORM_STATUS', platforms: selected });
         return response?.status || {};
       }
+      case 'captureUrl': {
+        if (typeof payload?.url !== 'string' || !payload.url.trim())
+          throw new Error('缺少要采集的页面地址');
+        return chrome.runtime.sendMessage({ type: 'CAPTURE_URL', url: payload.url });
+      }
       case 'startBatch':
         return sendRuntimeMessage({ type: 'START_SYNC_BATCH' });
       case 'syncToPlatform': {

distribution-page-bridge.js

@@ -44,6 +44,7 @@
       getVersion: () => request('getVersion'),
       listPlatforms: () => request('listPlatforms'),
       checkPlatforms: platformIds => request('checkPlatforms', { platformIds }),
+      captureUrl: url => request('captureUrl', { url }),
       startBatch: () => request('startBatch'),
       syncToPlatform: (platformId, content) => request('syncToPlatform', { platformId, content }),
     },

distribution/cose/background.js

@@ -868,7 +868,7 @@ async function syncToPlatform(platformId, content) {
       await new Promise(resolve => setTimeout(resolve, 1000))
 
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.plainText || content.markdown || content.body || ''
       console.log('[COSE] 小红书 HTML 内容长度:', htmlContent?.length || 0)
 
       // 填充标题和内容
@@ -1583,7 +1583,7 @@ async function syncToPlatform(platformId, content) {
     // 微信公众号：直接注入 HTML 到编辑器
     if (platformId === 'wechat') {
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.wechatHtml || content.html || content.body || ''
       console.log('[COSE] 微信 HTML 内容长度:', htmlContent?.length || 0)
 
       // 等待额外时间确保编辑器完全加载
@@ -1753,7 +1753,7 @@ async function syncToPlatform(platformId, content) {
     // 抖音：使用剪贴板 HTML 粘贴到编辑器（类似微信公众号）
     if (platformId === 'douyin') {
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.plainText || content.markdown || content.body || ''
       console.log('[COSE] 抖音 HTML 内容长度:', htmlContent?.length || 0)
       console.log('[COSE] 开始注入抖音内容...')
 
@@ -1883,7 +1883,7 @@ async function syncToPlatform(platformId, content) {
       await new Promise(resolve => setTimeout(resolve, 3000))
 
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.html || content.body || ''
       console.log('[COSE] 搜狐号 HTML 内容长度:', htmlContent?.length || 0)
 
       // 填充标题和内容
@@ -1942,7 +1942,7 @@ async function syncToPlatform(platformId, content) {
     // B站专栏：使用 UEditor execCommand 插入 HTML
     if (platformId === 'bilibili') {
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.html || content.body || ''
       console.log('[COSE] B站专栏 HTML 内容长度:', htmlContent?.length || 0)
 
       // 等待 UEditor 就绪
@@ -2052,7 +2052,7 @@ async function syncToPlatform(platformId, content) {
       await new Promise(resolve => setTimeout(resolve, 3000))
 
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.plainText || content.markdown || content.body || ''
       console.log('[COSE] 微博头条 HTML 内容长度:', htmlContent?.length || 0)
 
       // 填充标题和内容
@@ -2717,7 +2717,7 @@ async function syncToPlatform(platformId, content) {
       await new Promise(resolve => setTimeout(resolve, 3000))
 
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.html || content.body || ''
       console.log('[COSE] 百家号 HTML 内容长度:', htmlContent?.length || 0)
 
       // 填充标题和内容
@@ -2826,7 +2826,7 @@ async function syncToPlatform(platformId, content) {
       await new Promise(resolve => setTimeout(resolve, 3000))
 
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.html || content.body || ''
       console.log('[COSE] 少数派 HTML 内容长度:', htmlContent?.length || 0)
 
       // 填充标题和内容
@@ -2896,7 +2896,7 @@ async function syncToPlatform(platformId, content) {
       await new Promise(resolve => setTimeout(resolve, 3000))
 
       // 使用剪贴板 HTML（带完整样式）或降级到 body
-      const htmlContent = content.wechatHtml || content.body
+      const htmlContent = content.html || content.body || ''
       console.log('[COSE] 支付宝开放平台 HTML 内容长度:', htmlContent?.length || 0)
 
       // 填充标题和内容
@@ -3196,7 +3196,7 @@ async function syncToPlatform(platformId, content) {
     // 豆瓣：向首页分享框注入内容
     if (platformId === 'douban') {
       // 使用纯文本内容（豆瓣分享框不支持富文本）
-      const textContent = content.markdown || content.body || ''
+      const textContent = content.plainText || content.markdown || content.body || ''
       console.log('[COSE] 豆瓣文本内容长度:', textContent?.length || 0)
 
       // 等待页面加载
@@ -3389,7 +3389,7 @@ async function syncToPlatform(platformId, content) {
 
 // 在目标页面执行的填充函数
 function fillContentOnPage(content, platformId) {
-  const { title, body, markdown, wechatHtml } = content
+  const { title, body, html, markdown, plainText } = content
 
   // 等待元素出现的工具函数
   function waitFor(selector, timeout = 10000) {
@@ -3422,7 +3422,7 @@ function fillContentOnPage(content, platformId) {
   // 根据平台填充内容
   async function fill() {
     const host = window.location.hostname
-    const contentToFill = markdown || body || ''
+    const contentToFill = markdown || plainText || body || ''
 
     // 知乎专栏 - 由 syncToPlatform 单独处理（使用导入文档功能）
     if (host.includes('zhihu.com')) {
@@ -3839,7 +3839,7 @@ function fillContentOnPage(content, platformId) {
        * Medium 使用 contenteditable 编辑器，通过 paste 事件注入 HTML 内容
        * 使用剪贴板 HTML（带完整样式）或降级到 body
        */
-      const htmlContent = wechatHtml || body || ''
+      const htmlContent = html || body || ''
       const contentEl = document.querySelector('p.graf--p')
 
       if (contentEl && htmlContent) {

distribution/cose/core/platforms/medium.js

@@ -19,8 +19,8 @@ const MediumPlatform = {
  * 3. 通过 paste 事件填充 HTML 内容到编辑器
  */
 async function fillMediumContent(content, waitFor, setInputValue) {
-  const { title, body, wechatHtml } = content
-  const htmlContent = wechatHtml || body || ''
+  const { title, body, html } = content
+  const htmlContent = html || body || ''
 
   console.log('[COSE] Medium 开始同步...')
 

distribution/cose/core/platforms/wangyihao.js

@@ -94,7 +94,7 @@ async function syncWangyihaoContent(tab, content, helpers) {
   await injectUtils(chrome, tab.id)
 
   // 使用剪贴板 HTML（带完整样式）或降级到 body
-  const htmlContent = content.wechatHtml || content.body
+  const htmlContent = content.html || content.body || ''
   console.log('[COSE] 网易号 HTML 内容长度:', htmlContent?.length || 0)
 
   // 在页面中执行：填充标题和粘贴 HTML 内容

distribution/cose/core/platforms/wechat.js

@@ -331,7 +331,7 @@ async function syncWechatContent(tab, content, helpers) {
   await waitForTab(tab.id)
 
   // 使用剪贴板 HTML（带完整样式）或降级到 body
-  const htmlContent = content.wechatHtml || content.body
+  const htmlContent = content.wechatHtml || content.html || content.body || ''
   console.log('[COSE] 微信 HTML 内容长度:', htmlContent?.length || 0)
 
   // 步骤4：使用 MutationObserver 监听编辑器出现

manifest.json

@@ -1,7 +1,7 @@
 {
   "manifest_version": 3,
   "name": "AiToEarn - 内容营销助手",
-  "version": "1.1.0",
+  "version": "1.2.0",
   "description": "AiToEarn 自维护内容扩展：网页采集、账号检测与多平台文章分发",
   "permissions": [
     "activeTab",