优化发帖

spider/task.py

@@ -432,73 +432,173 @@ def retry_get_new_video(page, cookies, post_count):
 
     Args:
         page: Playwright页面对象
-        cookies: Cookies
+        cookies: Cookies (此处未使用，保留接口)
         post_count: 初始帖子数量
 
     Returns:
-        函数执行结果或抛出Timeout（如果超时）
+        新视频的页面URL
+
+    Raises:
+        TimeoutError: 超时未获取到新视频
+        OperationFailed: 点击视频时出错
     """
-    max_duration = 10 * 60  # 5分钟（秒）
-    retry_interval = 30  # 30秒重试一次
+    max_duration = 10 * 60   # 5分钟（300秒）
+    retry_interval = 30      # 30秒重试一次
     start_time = time.time()
     attempt = 1
 
     while time.time() - start_time < max_duration:
-        # 获取当前帖子数量
+        # 1. 滚动到页面底部，触发懒加载新内容
+        print("滚动到页面底部，加载更多内容...")
+        page.evaluate("window.scrollTo(0, document.body.scrollHeight)")
+        # 等待新内容加载（可调整时间或使用网络空闲等待）
+        page.wait_for_timeout(3000)  # 等待3秒让新帖子渲染
+        # 可选：等待网络空闲确保动态内容加载完成
+        # page.wait_for_load_state("networkidle")
+
+        # 2. 获取当前帖子数量
         new_post_count = get_post_count(page, cookies)
-        # 如果新帖子数量大于初始帖子数量，则表示有新帖子上传
+        print(f"第{attempt}次检查: 初始帖子数={post_count}, 当前帖子数={new_post_count}")
+
+        # 3. 如果有新帖子（视频）
         if new_post_count > post_count:
+            print("检测到新视频，准备点击...")
             try:
-                # 尝试点击视频
+                # 获取所有带 aria-posinset 的 div（代表每个视频帖子）
                 comment_buttons = page.query_selector_all('//div[@aria-posinset]')
                 if comment_buttons:
-                    # 使用js去点击第一个评论按钮
-                    element = page.query_selector_all('//div[@aria-posinset]')[0]
-                    # 获取元素位置
+                    # 使用第一个视频元素（最新发布的通常在第一个）
+                    element = comment_buttons[0]
                     bounding_box = element.bounding_box()
 
                     if bounding_box:
-                        # 滚动到特定位置
+                        # 滚动到元素可视区域中央
                         page.evaluate(
-                            f''' () => {{ window.scrollTo({{ top: {bounding_box['y']} - window.innerHeight / 2, left: {bounding_box['x']} - window.innerWidth / 2, behavior: 'smooth' }}); }} ''')
+                            f''' () => {{ 
+                                window.scrollTo({{ 
+                                    top: {bounding_box['y']} - window.innerHeight / 2, 
+                                    left: {bounding_box['x']} - window.innerWidth / 2, 
+                                    behavior: 'smooth' 
+                                }}); 
+                            }} '''
+                        )
+                        page.wait_for_timeout(1000)  # 等待滚动完成
 
-                        # 等待滚动完成
-                        page.wait_for_timeout(1000)
-
-                        # 点击
+                        # 点击视频
                         element.click()
+                        # 等待视频页面加载
                         time.sleep(random.randint(3, 5))
+                        # 刷新页面确保视频播放器完全加载（可选）
                         page.reload(timeout=180000)
                         time.sleep(random.randint(3, 5))
                         return page.url
                     else:
-                        raise OperationFailed(f"未找到视频")
+                        raise OperationFailed("未找到视频元素的有效位置信息")
+                else:
+                    raise OperationFailed("未找到任何视频元素（div[@aria-posinset]）")
+
             except Exception as e:
                 raise OperationFailed(f"点击视频时出错: {e}")
 
-        # 计算下一次重试时间
+        # 4. 没有新视频，等待下一次重试
         elapsed = time.time() - start_time
         remaining_time = max_duration - elapsed
 
         if remaining_time > 0:
-            # 等待30秒或剩余时间（取较小值）
             sleep_time = min(retry_interval, remaining_time)
-            print(f"第{attempt}次尝试，等待 {sleep_time:.1f} 秒后重试... (剩余时间: {remaining_time:.1f}秒)")
+            print(f"第{attempt}次尝试未发现新视频，等待 {sleep_time:.1f} 秒后重试... (剩余时间: {remaining_time:.1f}秒)")
             time.sleep(sleep_time)
 
         attempt += 1
 
+    # 超时退出
     print("5分钟超时，退出重试")
     raise TimeoutError("未获取到新视频(可能视频上传失败)，已超时")
 
 
-def playwright_post(cookies, content, image_key=None):
+def _is_video_media(image_key):
+    return bool(image_key and image_key.lower().endswith('.mp4'))
+
+
+def _download_post_media(image_key):
+    if not image_key:
+        return None
+
+    filename = image_key.split('/')[-1]
+    unique_filename = f"{uuid.uuid4()}_{filename}"
+    file_path = os.path.join(BASE_PATH, 'files', unique_filename)
+    client.fget_object(BUCKET, image_key, file_path)
+    return file_path
+
+
+def _open_post_composer(page, has_media):
+    if has_media:
+        composer = page.locator('//div[contains(@aria-placeholder, "What\'s on your mind")]').first
+        composer.wait_for(state='visible', timeout=300000)
+        return composer
+
+    triggers = [
+        '//span[contains(text(), "What\'s on your mind")]',
+        '//div[@role="button"]//span[contains(text(), "What\'s on your mind")]',
+    ]
+    last_error = None
+    for selector in triggers:
+        try:
+            page.locator(selector).first.click(timeout=30000)
+            composer = page.locator('//div[contains(@aria-placeholder, "What\'s on your mind")]').first
+            composer.wait_for(state='visible', timeout=300000)
+            return composer
+        except Error as e:
+            last_error = e
+
+    raise OperationFailed(f'未能打开发布输入框: {last_error}')
+
+
+def _fill_post_content(page, content, has_media):
+    composer = _open_post_composer(page, has_media=has_media)
+    composer.fill(content, timeout=300000)
+
+
+def _wait_post_submit_result(page):
+    page.wait_for_timeout(15000)
+
+
+def _get_latest_post_url(page):
+    candidates = [
+        '//div[@aria-posinset="1"]//a[@role="link"]',
+        '//a[contains(@href, "/posts/")]',
+        '//a[contains(@href, "permalink")]',
+    ]
+    for selector in candidates:
+        locator = page.locator(selector)
+        count = locator.count()
+        if count == 0:
+            continue
+        for index in range(count):
+            href = locator.nth(index).get_attribute('href')
+            if href and ('/posts/' in href or 'permalink' in href):
+                if href.startswith('/'):
+                    return f'https://www.facebook.com{href}'
+                return href
+
+    page.reload(timeout=180000)
+    return page.url
+
+
+def _is_facebook_home(page):
+    current_url = (page.url or '').rstrip('/')
+    return current_url in {'https://www.facebook.com', 'https://facebook.com'}
+
+
+def playwright_post(cookies, content, image_key=None, dry_run=False):
     path = os.path.join(BASE_PATH, 'chrome', '130-0008', 'chrome.exe')
     with lock:
         with sync_playwright() as playwright:
             update_windows_distinguish()
             max_browser_retries = 3
             last_error = None
+            parsed_cookies = parse_cookies(cookies)
+            is_video = _is_video_media(image_key)
 
             for browser_attempt in range(max_browser_retries):
                 browser = None
@@ -509,51 +609,52 @@ def playwright_post(cookies, content, image_key=None):
                         headless=False, args=['--start-maximized'], executable_path=path
                     )
                     context = browser.new_context(no_viewport=True)
-                    context.add_cookies(parse_cookies(cookies))
+                    context.add_cookies(parsed_cookies)
                     page = context.new_page()
+                    page.set_default_timeout(30000)
+                    page.set_default_navigation_timeout(180000)
 
-                    check_account_status(page, parse_cookies(cookies))
-                    # 声明默认发布视频数量
+                    check_account_status(page, parsed_cookies)
                     video_count = 0
 
                     url = 'https://www.facebook.com'
-                    # 先获取视频数量
-                    if image_key is not None and ".mp4" in image_key:
+                    if is_video:
                         video_count = get_post_count(page, cookies)
 
-                    # check_account_status 已经把页面带到 Facebook 首页，避免重复二次跳转导致页面进程崩溃
-                    if "facebook.com" not in page.url:
+                    if not _is_facebook_home(page):
                         retry_goto(page, url)
-                    time.sleep(random.randint(3, 10))
-                    time.sleep(5)
+                    sleep(3, 5)
 
                     if image_key:
-                        filename = image_key.split('/')[-1]
-                        file_path = os.path.join(BASE_PATH, 'files', filename)
-                        client.fget_object(BUCKET, image_key, file_path)
-
+                        file_path = _download_post_media(image_key)
                         sleep(1, 2)
                         page.locator('input[accept="image/*,image/heif,image/heic,video/*,video/mp4,video/x-m4v,video/x-matroska,.mkv"]').set_input_files(file_path)
-                        time.sleep(5)
+                        page.locator('//div[contains(@aria-placeholder, "What\'s on your mind")]').first.wait_for(
+                            state='visible', timeout=300000
+                        )
 
-                    if not image_key:
-                        page.click('''//span[contains(text(), "What's on your mind")]''')
+                    _fill_post_content(page, content, has_media=bool(image_key))
                     _edit_privacy(page)
-                    # 修改后 (使用 fill)
-                    page.fill('//div[contains(@aria-placeholder, "What\'s on your mind")]', content,
-                              timeout=300000)
-                    page.click('//div[@aria-label="Post"]', timeout=300000)
-                    time.sleep(15)
-                    post_index = page.locator('//div[@aria-posinset="1"]//a[@role="link"]').nth(2)
-                    post_index.click(timeout=600000)
-                    time.sleep(5)
-                    page.reload(timeout=180000)
-                    post_url = page.url
-                    # 视频格式要单独去获取链接
-                    if image_key is not None and ".mp4" in image_key:
-                        post_url = retry_get_new_video(page, cookies, video_count)
-                        time.sleep(random.randint(3, 10))
 
+                    post_button = page.locator('//div[@aria-label="Post"]').first
+                    post_button.wait_for(state='visible', timeout=300000)
+                    if dry_run:
+                        screenshot_content = _full_screenshot()
+                        key = f'screenshot/{uuid.uuid4()}.png'
+                        put_object(key, screenshot_content)
+                        return {
+                            'response_url': page.url,
+                            'screenshot_key': key,
+                            'dry_run': True,
+                            'message': '已完成到发布前校验，未实际点击 Post'
+                        }
+                    post_button.click(timeout=300000)
+                    _wait_post_submit_result(page)
+
+                    if is_video:
+                        post_url = retry_get_new_video(page, cookies, video_count)
+                    else:
+                        post_url = _get_latest_post_url(page)
                     screenshot_content = _full_screenshot()
                     key = f'screenshot/{uuid.uuid4()}.png'
                     put_object(key, screenshot_content)