chatgpt-mirai-qq-bot-web-search 0.1.14__py3-none-any.whl → 0.2.1__py3-none-any.whl

{chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info → chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: chatgpt-mirai-qq-bot-web-search
-Version: 0.1.14
+Version: 0.2.1
 Summary: WebSearch adapter for lss233/chatgpt-mirai-qq-bot
 Home-page: https://github.com/chuanSir123/web_search
 Author: chuanSir

chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/RECORD

@@ -0,0 +1,11 @@
+web_search/__init__.py,sha256=zVZLb5A-im5XETwohgxyE-UCxjSvYl6I2OC3LnEQhdQ,4360
+web_search/blocks.py,sha256=S3RsV9CCTKAsKUNhewg__ejEpJRDz7DTawtH05WRgE8,6732
+web_search/config.py,sha256=DhLiERBJR2V5Boglf7Aq9Rbc4vsvLIh67CrLDIPeqA0,398
+web_search/web_searcher.py,sha256=0zLgMsWCK71gStyWpFjup5WfxHx3tBTf3rGwM7Ae7Zs,13332
+web_search/example/roleplayWithWebSearch.yaml,sha256=C-dGy3z8gcRcmxzurssP-kPRLqMf1TYR-nnNUaJjISE,7468
+chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/LICENSE,sha256=ILBn-G3jdarm2w8oOrLmXeJNU3czuJvVhDLBASWdhM8,34522
+chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/METADATA,sha256=Gt59c1F8TCJFClQ0qqdMvrCQ2gpeHHcIVH9cbVli-zw,1738
+chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/WHEEL,sha256=tZoeGjtWxWRfdplE7E3d45VPlLNQnvbKiYnx7gwAy8A,92
+chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/entry_points.txt,sha256=o3kRDSdSmSdnCKlK6qS57aN0WpI4ab-Nxub2NwUrjf0,64
+chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/top_level.txt,sha256=PoNm8MJYw_y8RTMaNlY0ePLoNHxVUAE2IHDuL5fFubI,11
+chatgpt_mirai_qq_bot_web_search-0.2.1.dist-info/RECORD,,

web_search/blocks.py

@@ -1,16 +1,17 @@
-from typing import Any, Dict, List, Optional
+from typing import Any, Dict, List, Optional,Annotated
 import asyncio
-from kirara_ai.workflow.core.block import Block
-from kirara_ai.workflow.core.block.input_output import Input, Output
+from kirara_ai.workflow.core.block import Block, Input, Output, ParamMeta
 from .web_searcher import WebSearcher
 from .config import WebSearchConfig
 from kirara_ai.llm.format.message import LLMChatMessage
 from kirara_ai.llm.format.response import LLMChatResponse
+from kirara_ai.ioc.container import DependencyContainer
 
+def get_options_provider(container: DependencyContainer, block: Block) -> List[str]:
+    return ["bing", "google", "baidu"]
 class WebSearchBlock(Block):
     """Web搜索Block"""
     name = "web_search"
-
     inputs = {
         "llm_resp": Input(name="llm_resp",label="LLM 响应", data_type=LLMChatResponse, description="搜索关键词")
     }
@@ -19,13 +20,16 @@ class WebSearchBlock(Block):
         "results": Output(name="results",label="搜索结果",data_type= str, description="搜索结果")
     }
 
-    def __init__(self, name: str = None, max_results: Optional[int] = None, timeout: Optional[int] = None, fetch_content: Optional[bool] = None):
+    def __init__(self, name: str = None, max_results: Optional[int] = 3, timeout: Optional[int] = 10, fetch_content: Optional[bool] = True
+    ,engine: Annotated[Optional[str],ParamMeta(label="搜索引擎", description="要使用的搜索引擎", options_provider=get_options_provider),] = "bing", proxy: str = None,):
         super().__init__(name)
         self.searcher = None
         self.config = WebSearchConfig()
         self.max_results = max_results
         self.timeout = timeout
         self.fetch_content = fetch_content
+        self.engine=engine
+        self.proxy = proxy
 
     def _ensure_searcher(self):
         """同步方式初始化searcher"""
@@ -62,11 +66,81 @@ class WebSearchBlock(Block):
                     query=query,
                     max_results=max_results,
                     timeout=timeout,
-                    fetch_content=fetch_content
+                    fetch_content=fetch_content,
+                    engine=self.engine,
+                    proxy = self.proxy,
+                )
+            )
+            return {"results": "\n以下是联网搜索的结果:\n-- 搜索结果开始 --"+results+"\n-- 搜索结果结束 --"}
+        except Exception as e:
+            print(e)
+            return {"results": f"搜索失败: {str(e)}"}
+class WebSearchByKeywordBlock(Block):
+    """Web搜索Block"""
+    name = "web_search_by_keyword"
+    description = "网络搜索，通过关键词进行网络搜索"
+
+    inputs = {
+        "keyword": Input(name="keyword",label="搜索关键字", data_type=str, description="搜索关键词")
+    }
+
+    outputs = {
+        "results": Output(name="results",label="搜索结果",data_type= str, description="搜索结果")
+    }
+
+    def __init__(self, name: str = None, max_results: Optional[int] = 3, timeout: Optional[int] = 10, fetch_content: Optional[bool] = True
+    ,engine: Annotated[Optional[str],ParamMeta(label="搜索引擎", description="要使用的搜索引擎", options_provider=get_options_provider),] = "bing", proxy: str = None,):
+        super().__init__(name)
+        self.searcher = None
+        self.config = WebSearchConfig()
+        self.max_results = max_results
+        self.timeout = timeout
+        self.fetch_content = fetch_content
+        self.engine=engine
+        self.proxy = proxy
+
+    def _ensure_searcher(self):
+        """同步方式初始化searcher"""
+        if not self.searcher:
+            try:
+                loop = asyncio.get_event_loop()
+            except RuntimeError:
+                # 如果在新线程中没有事件循环，则创建一个新的
+                loop = asyncio.new_event_loop()
+                asyncio.set_event_loop(loop)
+            self.searcher = loop.run_until_complete(WebSearcher.create())
+
+    def execute(self, **kwargs) -> Dict[str, Any]:
+        query = kwargs["keyword"]
+
+        if query == "" or query.startswith("无"):
+            return {"results": ""}
+        max_results = self.max_results
+        timeout = self.timeout
+        fetch_content = self.fetch_content
+        self._ensure_searcher()
+
+        try:
+            # 在新线程中创建事件循环
+            try:
+                loop = asyncio.get_event_loop()
+            except RuntimeError:
+                loop = asyncio.new_event_loop()
+                asyncio.set_event_loop(loop)
+
+            results = loop.run_until_complete(
+                self.searcher.search(
+                    query=query,
+                    max_results=max_results,
+                    timeout=timeout,
+                    fetch_content=fetch_content,
+                    engine=self.engine,
+                    proxy = self.proxy,
                 )
             )
             return {"results": "\n以下是联网搜索的结果:\n-- 搜索结果开始 --"+results+"\n-- 搜索结果结束 --"}
         except Exception as e:
+            print(e)
             return {"results": f"搜索失败: {str(e)}"}
 
 class AppendSystemPromptBlock(Block):

web_search/web_searcher.py

@@ -1,237 +1,325 @@
-from playwright.async_api import async_playwright
-import trafilatura
-import random
-import time
-import urllib.parse
-import asyncio
-import subprocess
-import sys
-from kirara_ai.logger import get_logger
-
-logger = get_logger("WebSearchPlugin")
-
-class WebSearcher:
-    def __init__(self):
-        self.playwright = None
-        self.browser = None
-        self.context = None
-
-    @classmethod
-    async def create(cls):
-        """创建 WebSearcher 实例的工厂方法"""
-        self = cls()
-        return self
-
-    async def _ensure_initialized(self):
-        """确保浏览器已初始化"""
-        try:
-            self.playwright = await async_playwright().start()
-            try:
-                self.browser = await self.playwright.chromium.launch(
-                    headless=True,
-                    chromium_sandbox=False,
-                    args=['--no-sandbox', '--disable-setuid-sandbox', '--disable-dev-shm-usage', '--disable-gpu']
-                )
-            except Exception as e:
-                if "Executable doesn't exist" in str(e):
-                    logger.info("Installing playwright browsers...")
-                    # 使用 python -m playwright install 安装浏览器
-                    process = subprocess.Popen(
-                        [sys.executable, "-m", "playwright", "install", "chromium"],
-                        stdout=subprocess.PIPE,
-                        stderr=subprocess.PIPE
-                    )
-                    stdout, stderr = process.communicate()
-                    if process.returncode != 0:
-                        raise RuntimeError(f"Failed to install playwright browsers: {stderr.decode()}")
-
-                    # 重试启动浏览器
-                    self.browser = await self.playwright.chromium.launch(
-                        headless=False,
-                        chromium_sandbox=False,
-                        args=['--no-sandbox', '--disable-setuid-sandbox', '--disable-dev-shm-usage', '--disable-gpu']
-                    )
-                else:
-                    raise
-            return await self.browser.new_context(
-                viewport={'width': 1920, 'height': 1080},
-                user_agent='Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/121.0.0.0 Safari/537.36'
-            )
-        except Exception as e:
-            logger.error(f"Failed to initialize WebSearcher: {e}")
-            await self.close()
-            raise
-
-    async def random_sleep(self, min_time=1, max_time=3):
-        """随机等待"""
-        await asyncio.sleep(random.uniform(min_time, max_time))
-
-    async def simulate_human_scroll(self, page):
-        """模拟人类滚动"""
-        for _ in range(3):
-            await page.mouse.wheel(0, random.randint(300, 700))
-            await self.random_sleep(0.3, 0.7)
-
-    async def get_webpage_content(self, url: str, timeout: int,context) -> str:
-        """获取网页内容"""
-        start_time = time.time()
-        try:
-            # 创建新标签页获取内容
-            page = await context.new_page()
-            try:
-                # 设置更严格的资源加载策略
-                await page.route("**/*", lambda route: route.abort()
-                    if route.request.resource_type in ['image', 'stylesheet', 'font', 'media']
-                    else route.continue_())
-
-                # 使用 domcontentloaded 而不是 networkidle
-                await page.goto(url, wait_until='domcontentloaded', timeout=timeout * 1000)
-
-                # 等待页面主要内容加载，但设置较短的超时时间
-                try:
-                    await page.wait_for_load_state('domcontentloaded', timeout=5000)
-                except Exception as e:
-                    logger.warning(f"Load state timeout for {url}, continuing anyway: {e}")
-
-                await self.random_sleep(1, 2)
-                await self.simulate_human_scroll(page)
-
-                content = await page.content()
-                text = trafilatura.extract(content)
-
-                await page.close()
-                logger.info(f"Content fetched - URL: {url} - Time: {time.time() - start_time:.2f}s")
-                return text or ""
-            except Exception as e:
-                await page.close()
-                logger.error(f"Failed to fetch content - URL: {url} - Error: {e}")
-                return ""
-        except Exception as e:
-            logger.error(f"Failed to create page - URL: {url} - Error: {e}")
-            return ""
-
-    async def process_search_result(self, result, idx: int, timeout: int, fetch_content: bool,context):
-        """处理单个搜索结果"""
-        try:
-            title_element = await result.query_selector('h2')
-            link_element = await result.query_selector('h2 a')
-            snippet_element = await result.query_selector('.b_caption p')
-
-            if not title_element or not link_element:
-                return None
-
-            title = await title_element.inner_text()
-            link = await link_element.get_attribute('href')
-            snippet = await snippet_element.inner_text() if snippet_element else "无简介"
-
-            if not link:
-                return None
-
-            result_text = f"[{idx+1}] {title}\nURL: {link}\n搜索简介: {snippet}"
-
-            if fetch_content:
-
-                content = await self.get_webpage_content(link, timeout,context)
-                if content:
-                    result_text += f"\n内容详情:\n{content}"
-
-            return result_text
-
-        except Exception as e:
-            logger.error(f"Failed to process result {idx}: {e}")
-            return None
-
-    async def search(self, query: str, max_results: int = 3, timeout: int = 10, fetch_content: bool = True) -> str:
-        """执行搜索"""
-        context = await self._ensure_initialized()
-
-        search_start_time = time.time()
-        page = None
-        try:
-            encoded_query = urllib.parse.quote(query)
-            page = await context.new_page()
-
-            # 添加重试逻辑
-            max_retries = 3
-            for attempt in range(max_retries):
-                try:
-                    logger.info(f"Attempting to load search page (attempt {attempt + 1}/{max_retries})")
-                    await page.goto(
-                        f"https://www.bing.com/search?q={encoded_query}",
-                        wait_until='domcontentloaded',
-                        timeout=timeout * 1000
-                    )
-
-                    # 检查页面是否为空
-                    content = await page.content()
-                    if 'b_algo' not in content:
-                        if attempt < max_retries - 1:
-                            await page.reload()
-                            await self.random_sleep(1, 2)
-                            continue
-                    else:
-                        break
-                except Exception as e:
-                    logger.warning(f"Page navigation failed on attempt {attempt + 1}: {e}")
-                    if attempt < max_retries - 1:
-                        await self.random_sleep(1, 2)
-                        continue
-                    else:
-                        raise
-
-            # 使用更可靠的选择器等待策略
-            try:
-                selectors = ['.b_algo', '#b_results .b_algo', 'main .b_algo']
-                results = None
-
-                for selector in selectors:
-                    try:
-                        await page.wait_for_selector(selector, timeout=5000)
-                        results = await page.query_selector_all(selector)
-                        if results and len(results) > 0:
-                            break
-                    except Exception:
-                        continue
-
-                if not results:
-                    logger.error("No search results found with any selector")
-                    return "搜索结果加载失败"
-
-            except Exception as e:
-                logger.error(f"Failed to find search results: {e}")
-                return "搜索结果加载失败"
-
-            logger.info(f"Found {len(results)} search results")
-
-            tasks = []
-            for idx, result in enumerate(results[:max_results]):
-                tasks.append(self.process_search_result(result, idx, timeout, fetch_content,context))
-
-            detailed_results = []
-            completed_results = await asyncio.gather(*tasks)
-
-            for result in completed_results:
-                if result:
-                    detailed_results.append(result)
-
-            total_time = time.time() - search_start_time
-            results = "\n---\n".join(detailed_results) if detailed_results else "未找到相关结果"
-            logger.info(f"Search completed - Query: {query} - Time: {total_time:.2f}s - Found {len(detailed_results)} valid results")
-            return results
-
-        except Exception as e:
-            logger.error(f"Search failed - Query: {query} - Error: {e}", exc_info=True)
-            return f"搜索失败: {str(e)}"
-        finally:
-            if page:
-                try:
-                    await page.close()
-                except Exception as e:
-                    logger.error(f"Error closing page: {e}")
-
-    async def close(self):
-        """关闭浏览器"""
-        if self.browser:
-            await self.browser.close()
-        if self.playwright:
-            await self.playwright.stop()
+from playwright.async_api import async_playwright
+import trafilatura
+import random
+import time
+import urllib.parse
+import asyncio
+import subprocess
+import sys
+from kirara_ai.logger import get_logger
+import os
+
+logger = get_logger("WebSearchPlugin")
+
+class WebSearcher:
+    def __init__(self):
+        self.playwright = None
+        self.browser = None
+        self.context = None
+        self.search_engines = {
+            'bing': {
+                'url': 'https://www.bing.com/search?q={}',
+                'selectors': ['.b_algo', '#b_results .b_algo', 'main .b_algo'],
+                'title_selector': 'h2',
+                'link_selector': 'h2 a',
+                'snippet_selector': '.b_caption p'
+            },
+            'google': {
+                'url': 'https://www.google.com/search?q={}',
+                'selectors': ['.MjjYud', 'div.g', 'div[data-hveid]'],
+                'title_selector': 'h3.LC20lb',
+                'link_selector': 'a[jsname="UWckNb"], div.yuRUbf a',
+                'snippet_selector': 'div.VwiC3b'
+            },
+            'baidu': {
+                'url': 'https://www.baidu.com/s?wd={}',
+                'selectors': ['.result', '.result-op'],
+                'title_selector': 'h3',
+                'link_selector': 'h3 a',
+                'snippet_selector': '.content-right_8Zs40'
+            }
+        }
+
+    @classmethod
+    async def create(cls):
+        """创建 WebSearcher 实例的工厂方法"""
+        self = cls()
+        return self
+
+    async def _ensure_initialized(self,proxy):
+        """确保浏览器已初始化"""
+        try:
+            self.playwright = await async_playwright().start()
+
+            # 创建用户数据目录路径
+            user_data_dir = os.path.join(os.path.expanduser("~"), ".playwright_user_data")
+            os.makedirs(user_data_dir, exist_ok=True)
+
+            # 合并所有选项到一个字典
+            context_options = {
+                'headless': True,
+                'chromium_sandbox': False,
+                'slow_mo': 50,  # 减慢操作速度，更像人类
+                'args': [
+                    '--no-sandbox',
+                    '--disable-setuid-sandbox',
+                    '--disable-dev-shm-usage',
+                    '--disable-blink-features=AutomationControlled',  # 隐藏自动化控制痕迹
+                    '--disable-features=IsolateOrigins,site-per-process',
+                ],
+                'ignore_default_args': ['--enable-automation'],  # 屏蔽自动化标志
+                'viewport': {'width': 1920, 'height': 1080},
+                'user_agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.0.0 Safari/537.36',
+                'locale': 'zh-CN',
+                'timezone_id': 'Asia/Shanghai',
+                'color_scheme': 'dark',  # 或 'light'，根据用户习惯
+                'device_scale_factor': 1.75,  # 高DPI设备
+                'has_touch': True,  # 支持触摸
+                'is_mobile': False,
+                'reduced_motion': 'no-preference'
+            }
+
+            # 如果是 Google 搜索，添加代理设置
+            if proxy:
+                context_options['proxy'] = {
+                    'server': proxy
+                }
+
+            try:
+                # 使用 launch_persistent_context 代替分开的 launch 和 new_context
+                self.context = await self.playwright.chromium.launch_persistent_context(
+                    user_data_dir=user_data_dir,
+                    **context_options
+                )
+
+                self.browser = None  # 不再需要单独的browser引用
+
+            except Exception as e:
+                if "Executable doesn't exist" in str(e):
+                    logger.info("Installing playwright browsers...")
+                    process = subprocess.Popen(
+                        [sys.executable, "-m", "playwright", "install", "chromium"],
+                        stdout=subprocess.PIPE,
+                        stderr=subprocess.PIPE
+                    )
+                    stdout, stderr = process.communicate()
+                    if process.returncode != 0:
+                        raise RuntimeError(f"Failed to install playwright browsers: {stderr.decode()}")
+
+                    # 重试使用 launch_persistent_context
+                    self.context = await self.playwright.chromium.launch_persistent_context(
+                        user_data_dir=user_data_dir,
+                        **context_options
+                    )
+                else:
+                    raise
+
+            # 注入脚本来伪装webdriver标记
+            await self.context.add_init_script("""
+                Object.defineProperty(navigator, 'webdriver', {
+                    get: () => false,
+                });
+
+                // 防止 iframe 检测
+                window.parent.document;
+
+                // 防止检测到 Chrome Devtools 协议
+                delete window.cdc_adoQpoasnfa76pfcZLmcfl_Array;
+                delete window.cdc_adoQpoasnfa76pfcZLmcfl_Promise;
+                delete window.cdc_adoQpoasnfa76pfcZLmcfl_Symbol;
+            """)
+
+            return self.context
+
+        except Exception as e:
+            logger.error(f"Failed to initialize WebSearcher: {e}")
+            await self.close()
+            raise
+
+    async def simulate_human_scroll(self, page):
+        """模拟人类滚动"""
+        for _ in range(3):
+            await page.mouse.wheel(0, random.randint(300, 700))
+
+    async def get_webpage_content(self, url: str, timeout: int,context) -> str:
+        """获取网页内容"""
+        start_time = time.time()
+        try:
+            # 创建新标签页获取内容
+            page = await context.new_page()
+            try:
+                # 设置更严格的资源加载策略
+                await page.route("**/*", lambda route: route.abort()
+                    if route.request.resource_type in ['image', 'stylesheet', 'font', 'media']
+                    else route.continue_())
+
+                # 使用 domcontentloaded 而不是 networkidle
+                await page.goto(url, wait_until='domcontentloaded', timeout=timeout * 1000)
+
+                # 等待页面主要内容加载，但设置较短的超时时间
+                try:
+                    await page.wait_for_load_state('domcontentloaded', timeout=5000)
+                except Exception as e:
+                    logger.warning(f"Load state timeout for {url}, continuing anyway: {e}")
+
+                await self.simulate_human_scroll(page)
+
+                content = await page.content()
+                text = trafilatura.extract(content)
+
+                await page.close()
+                logger.info(f"Content fetched - URL: {url} - Time: {time.time() - start_time:.2f}s")
+                return text or ""
+            except Exception as e:
+                await page.close()
+                logger.error(f"Failed to fetch content - URL: {url} - Error: {e}")
+                return ""
+        except Exception as e:
+            logger.error(f"Failed to create page - URL: {url} - Error: {e}")
+            return ""
+
+    async def process_search_result(self, result, idx: int, timeout: int, fetch_content: bool, context, engine='bing'):
+        """处理单个搜索结果"""
+        try:
+            engine_config = self.search_engines[engine]
+            title_element = await result.query_selector(engine_config['title_selector'])
+            link_element = await result.query_selector(engine_config['link_selector'])
+            snippet_element = await result.query_selector(engine_config['snippet_selector'])
+
+            if not title_element or not link_element:
+                return None
+
+            title = await title_element.inner_text()
+            link = await link_element.get_attribute('href')
+
+            # 对于百度搜索需要特殊处理链接
+            if engine == 'baidu':
+                try:
+                    # 创建新页面来获取真实URL
+                    new_page = await context.new_page()
+                    await new_page.goto(link, wait_until='domcontentloaded', timeout=5000)
+                    real_url = new_page.url
+                    await new_page.close()
+                    link = real_url
+                except Exception as e:
+                    logger.warning(f"Failed to get real URL from Baidu: {e}")
+
+            snippet = await snippet_element.inner_text() if snippet_element else "无简介"
+
+            if not link:
+                return None
+
+            result_text = f"[{idx+1}] {title}\nURL: {link}\n搜索简介: {snippet}"
+
+            if fetch_content:
+
+                content = await self.get_webpage_content(link, timeout,context)
+                if content:
+                    result_text += f"\n内容详情:\n{content}"
+
+            return result_text
+
+        except Exception as e:
+            logger.error(f"Failed to process result {idx}: {e}")
+            return None
+
+    async def search(self, query: str, max_results: int = 3, timeout: int = 10, fetch_content: bool = True, engine: str = 'bing', proxy: str = None) -> str:
+        """执行搜索"""
+        if engine not in self.search_engines:
+            return f"不支持的搜索引擎: {engine}"
+
+        # 设置当前搜索引擎
+        self.current_engine = engine
+        context = await self._ensure_initialized(proxy)
+        engine_config = self.search_engines[engine]
+        search_start_time = time.time()
+        page = None
+
+        try:
+            encoded_query = urllib.parse.quote(query)
+            page = await context.new_page()
+
+            # Google搜索特定处理
+            await page.goto(
+                                engine_config['url'].format(encoded_query),
+                                wait_until='load',
+                                timeout=timeout * 1000
+                            )
+
+            # 使用搜索引擎特定的选择器
+            results = None
+
+            # 对于Google，让页面有更多时间加载
+            if engine == 'google':
+                await self.simulate_human_scroll(page)
+
+            for selector in engine_config['selectors']:
+                try:
+                    logger.info(f"Trying selector: {selector}")
+                    await page.wait_for_selector(selector, timeout=8000)  # 增加等待时间
+                    results = await page.query_selector_all(selector)
+                    if results and len(results) > 0:
+                        logger.info(f"Found {len(results)} results with selector {selector}")
+                        break
+                except Exception as e:
+                    logger.warning(f"Selector {selector} failed: {e}")
+                    continue
+
+            if not results:
+                # 尝试直接使用 JavaScript 获取元素
+                if engine == 'google':
+                    try:
+                        # 使用更通用的JavaScript选择器尝试获取结果
+                        results = await page.evaluate("""
+                            () => {
+                                const elements = document.querySelectorAll('div[data-sokoban-container], div.g, .MjjYud');
+                                return Array.from(elements).length;
+                            }
+                        """)
+                        logger.info(f"JavaScript found {results} elements")
+
+                        # 如果找到了元素，使用evaluate来处理它们
+                        if results > 0:
+                            # 自定义处理逻辑...
+                            pass
+                    except Exception as e:
+                        logger.error(f"JavaScript evaluation failed: {e}")
+
+                logger.error("No search results found with any selector")
+                await page.screenshot(path=f'search_failed_{engine}.png')
+                return "搜索结果加载失败"
+
+            logger.info(f"Found {len(results)} search results")
+
+            tasks = []
+            for idx, result in enumerate(results[:max_results]):
+                tasks.append(self.process_search_result(result, idx, timeout, fetch_content, context, engine))
+
+            detailed_results = []
+            completed_results = await asyncio.gather(*tasks)
+
+            for result in completed_results:
+                if result:
+                    detailed_results.append(result)
+
+            total_time = time.time() - search_start_time
+            results = "\n---\n".join(detailed_results) if detailed_results else "未找到相关结果"
+            logger.info(f"Search completed - Query: {query} - Time: {total_time:.2f}s - Found {len(detailed_results)} valid results")
+            return results
+
+        except Exception as e:
+            logger.error(f"Search failed - Query: {query} - Error: {e}", exc_info=True)
+            return f"搜索失败: {str(e)}"
+        finally:
+            if page:
+                try:
+                    await page.close()
+                except Exception as e:
+                    logger.error(f"Error closing page: {e}")
+
+    async def close(self):
+        """关闭浏览器"""
+        if self.context:
+            await self.context.close()
+        if self.playwright:
+            await self.playwright.stop()

chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/RECORD

@@ -1,11 +0,0 @@
-web_search/__init__.py,sha256=zVZLb5A-im5XETwohgxyE-UCxjSvYl6I2OC3LnEQhdQ,4360
-web_search/blocks.py,sha256=F1XJt7n1mHqMhkoU5Du1IB7NEVCx3Cl9vHkoY7plwls,3611
-web_search/config.py,sha256=DhLiERBJR2V5Boglf7Aq9Rbc4vsvLIh67CrLDIPeqA0,398
-web_search/web_searcher.py,sha256=HsXs5ctvBryupK4hX8qbJspx1-rWEPnMYfulEkC1WMw,9707
-web_search/example/roleplayWithWebSearch.yaml,sha256=C-dGy3z8gcRcmxzurssP-kPRLqMf1TYR-nnNUaJjISE,7468
-chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/LICENSE,sha256=ILBn-G3jdarm2w8oOrLmXeJNU3czuJvVhDLBASWdhM8,34522
-chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/METADATA,sha256=6UNFYoIQXPBsJ1efkuD7UtKr5R1BRJ_9XnGI-dqdRUE,1739
-chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/WHEEL,sha256=tZoeGjtWxWRfdplE7E3d45VPlLNQnvbKiYnx7gwAy8A,92
-chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/entry_points.txt,sha256=o3kRDSdSmSdnCKlK6qS57aN0WpI4ab-Nxub2NwUrjf0,64
-chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/top_level.txt,sha256=PoNm8MJYw_y8RTMaNlY0ePLoNHxVUAE2IHDuL5fFubI,11
-chatgpt_mirai_qq_bot_web_search-0.1.14.dist-info/RECORD,,