cnks 0.1.0__py3-none-any.whl → 0.1.1__py3-none-any.whl

cnks/server.py

@@ -7,8 +7,10 @@ import subprocess
 import sys
 import time
 import logging
+import webbrowser
 from pathlib import Path
 from urllib.parse import quote
+from typing import Dict, List, Any, Optional, Union
 
 from mcp.server.models import InitializationOptions
 import mcp.types as types
@@ -97,6 +99,8 @@ def open_chrome(url):
 
 async def search_with_playwright(keywords):
     """使用playwright在知网搜索关键词"""
+    global page_content
+    
     if not PLAYWRIGHT_AVAILABLE:
         return "需要安装playwright模块：uv add playwright"
     
@@ -216,7 +220,7 @@ async def search_with_playwright(keywords):
                                             # 查找所有包含"article/abstract?v="字样的链接
                                             links_count = await find_and_count_abstract_links(page)
                                             
-                                            return f"已完成全部操作：搜索关键词、设置每页显示50条、勾选CSSCI来源类别。找到{links_count}条包含article/abstract?v=的链接。浏览器将保持打开状态。"
+                                            return links_count
                                         else:
                                             logger.debug("[DEBUG] 在来源类别区域未找到CSSCI选项")
                                             
@@ -231,9 +235,11 @@ async def search_with_playwright(keywords):
                                                 # 查找所有包含"article/abstract?v="字样的链接
                                                 links_count = await find_and_count_abstract_links(page)
                                                 
-                                                return f"已完成全部操作：搜索关键词、设置每页显示50条、勾选CSSCI来源类别。找到{links_count}条包含article/abstract?v=的链接。浏览器将保持打开状态。"
+                                                return links_count
                                             else:
-                                                return "已完成搜索和设置每页显示50条，但未找到CSSCI选项。浏览器将保持打开状态。"
+                                                # 查找所有包含"article/abstract?v="字样的链接
+                                                links_count = await find_and_count_abstract_links(page)
+                                                return links_count
                                     else:
                                         logger.debug("[DEBUG] 未找到来源类别区域")
                                         
@@ -248,61 +254,127 @@ async def search_with_playwright(keywords):
                                             # 查找所有包含"article/abstract?v="字样的链接
                                             links_count = await find_and_count_abstract_links(page)
                                             
-                                            return f"已完成全部操作：搜索关键词、设置每页显示50条、勾选CSSCI来源类别。找到{links_count}条包含article/abstract?v=的链接。浏览器将保持打开状态。"
+                                            return links_count
                                         else:
-                                            return "已完成搜索和设置每页显示50条，但未找到来源类别区域或CSSCI选项。浏览器将保持打开状态。"
+                                            # 查找所有包含"article/abstract?v="字样的链接
+                                            links_count = await find_and_count_abstract_links(page)
+                                            return links_count
                                 except Exception as e:
                                     logger.debug(f"[DEBUG] 勾选CSSCI选项时出错: {str(e)}")
-                                    return f"已完成搜索和设置每页显示50条，但勾选CSSCI时出错: {str(e)}。浏览器将保持打开状态。"
+                                    # 查找所有包含"article/abstract?v="字样的链接
+                                    links_count = await find_and_count_abstract_links(page)
+                                    return links_count
                                 
-                                return "已完成全部操作：搜索关键词、点击排序下拉框、选择每页显示50条。浏览器将保持打开状态。"
+                                # 查找所有包含"article/abstract?v="字样的链接
+                                links_count = await find_and_count_abstract_links(page)
+                                return links_count
                             else:
                                 logger.debug("[DEBUG] 未找到'50'选项")
-                                return "已搜索并点击下拉框，但未找到'50'选项。浏览器将保持打开状态。"
+                                page_content = {
+                                    "count": 0,
+                                    "links": [],
+                                    "error": "已搜索并点击下拉框，但未找到'50'选项"
+                                }
+                                return 0
                         else:
                             logger.debug("[DEBUG] 未找到排序下拉框")
-                            return "已搜索，但未找到排序下拉框。浏览器将保持打开状态。"
+                            page_content = {
+                                "count": 0,
+                                "links": [],
+                                "error": "已搜索，但未找到排序下拉框"
+                            }
+                            return 0
                     except Exception as e:
                         logger.debug(f"[DEBUG] 点击下拉框或选项时出错: {str(e)}")
-                        return f"已搜索，但在点击下拉框或选项时出错: {str(e)}。浏览器将保持打开状态。"
-                    
-                    # 不关闭浏览器，让它保持打开状态
-                    # 注意：不调用 browser.close() 和 playwright.stop()
+                        page_content = {
+                            "count": 0,
+                            "links": [],
+                            "error": f"已搜索，但在点击下拉框或选项时出错: {str(e)}"
+                        }
+                        return 0
                 else:
                     # 不关闭浏览器
-                    return f"已填写搜索关键词: {keywords}，但未找到搜索按钮。请手动点击搜索。"
+                    page_content = {
+                        "count": 0,
+                        "links": [],
+                        "error": f"已填写搜索关键词: {keywords}，但未找到搜索按钮"
+                    }
+                    return 0
             else:
                 # 不关闭浏览器
-                return f"未找到搜索框。已打开知网页面，请手动搜索: {keywords}"
+                page_content = {
+                    "count": 0,
+                    "links": [],
+                    "error": f"未找到搜索框，无法搜索: {keywords}"
+                }
+                return 0
         except Exception as e:
             logger.debug(f"[DEBUG] 填写搜索框或点击搜索按钮时出错: {str(e)}")
             # 不关闭浏览器
-            return f"自动搜索过程中出错，请手动在页面中搜索: {keywords}"
+            page_content = {
+                "count": 0,
+                "links": [],
+                "error": f"自动搜索过程中出错: {str(e)}"
+            }
+            return 0
     except Exception as e:
         error_msg = str(e)
         logger.debug(f"[DEBUG] Playwright错误: {error_msg}")
         
         # 如果是找不到Chrome的错误，提供更明确的指导
         if "Executable doesn't exist" in error_msg and "ms-playwright" in error_msg:
-            return f"需要安装Playwright的浏览器: playwright install\n如果您想使用系统Chrome，请重新启动服务器。\n\n{error_msg}"
-        
-        # 如果Playwright启动失败，使用传统方式打开Chrome
-        return f"使用Playwright启动Chrome失败: {error_msg}。尝试使用传统方式打开浏览器。"
+            error_message = f"需要安装Playwright的浏览器: playwright install\n如果您想使用系统Chrome，请重新启动服务器。\n\n{error_msg}"
+        else:
+            error_message = f"使用Playwright启动Chrome失败: {error_msg}"
+            
+        page_content = {
+            "count": 0,
+            "links": [],
+            "error": error_message
+        }
+        return 0
 
 def search_with_direct_chrome(keywords):
     """直接使用Chrome搜索，不使用playwright"""
-    logger.debug("[DEBUG] 正在使用search_with_direct_chrome函数")
-    
-    # 构建知网搜索URL - 知网不支持URL参数搜索，所以只能打开页面
-    url = "https://kns.cnki.net/kns8s/search"
+    global page_content
     
-    # 打开Chrome
-    result = open_chrome(url)
+    logger.debug("[DEBUG] 正在使用search_with_direct_chrome函数")
     
-    if result is True:
-        return f"已打开知网页面。请在搜索框中输入并搜索: {keywords}"
-    else:
-        return f"打开Chrome浏览器失败: {result}"
+    try:
+        url = f"https://kns.cnki.net/kns8s/search?q={quote(keywords)}"
+        logger.debug(f"[DEBUG] 打开URL: {url}")
+        
+        result = open_chrome(url)
+        
+        if isinstance(result, str) and "打开Chrome" in result:
+            logger.debug(f"[DEBUG] 直接打开Chrome结果: {result}")
+            
+            page_content = {
+                "count": 0,
+                "links": [],
+                "error": f"直接打开Chrome搜索: {result}"
+            }
+            
+        else:
+            logger.debug("[DEBUG] 直接打开Chrome成功")
+            
+            page_content = {
+                "count": 0,
+                "links": [],
+                "message": "已打开Chrome并搜索关键词，但无法自动获取链接。请安装playwright以获取完整功能。"
+            }
+        
+        return page_content
+    except Exception as e:
+        logger.debug(f"[DEBUG] search_with_direct_chrome出错: {str(e)}")
+        
+        page_content = {
+            "count": 0,
+            "links": [],
+            "error": f"使用Chrome搜索时出错: {str(e)}"
+        }
+        
+        return page_content
 
 def get_page_content():
     """获取当前页面内容（简化模拟）"""
@@ -515,15 +587,6 @@ async def handle_get_prompt(
 async def handle_list_tools() -> list[types.Tool]:
     """列出可用工具"""
     return [
-        types.Tool(
-            name="open-cnki",
-            description="打开中国知网搜索页面",
-            inputSchema={
-                "type": "object",
-                "properties": {},
-                "required": [],
-            },
-        ),
         types.Tool(
             name="search-keywords",
             description="在知网搜索关键词",
@@ -547,15 +610,6 @@ async def handle_list_tools() -> list[types.Tool]:
                 "required": ["name", "content"],
             },
         ),
-        types.Tool(
-            name="get-abstract-links",
-            description="获取最近一次搜索找到的论文摘要链接",
-            inputSchema={
-                "type": "object",
-                "properties": {},
-                "required": [],
-            },
-        )
     ]
 
 @server.call_tool()
@@ -565,25 +619,7 @@ async def handle_call_tool(
     """处理工具执行请求"""
     global current_url, page_content
     
-    if name == "open-cnki":
-        current_url = "https://kns.cnki.net/kns8s/search"
-        result = open_chrome(current_url)
-        if result is True:
-            return [
-                types.TextContent(
-                    type="text",
-                    text="已打开中国知网搜索页面。"
-                )
-            ]
-        else:
-            return [
-                types.TextContent(
-                    type="text",
-                    text=f"打开中国知网时出错: {result}"
-                )
-            ]
-    
-    elif name == "search-keywords":
+    if name == "search-keywords":
         if not arguments:
             raise ValueError("缺少参数")
         
@@ -593,24 +629,47 @@ async def handle_call_tool(
         
         # 优先使用playwright进行搜索
         if PLAYWRIGHT_AVAILABLE:
-            result = await search_with_playwright(keywords)
+            links_count = await search_with_playwright(keywords)
             current_url = "https://kns.cnki.net/kns8s/search"
             
-            return [
-                types.TextContent(
-                    type="text",
-                    text=result
-                )
-            ]
+            # 返回结构化数据
+            if isinstance(page_content, dict):
+                return [
+                    types.TextContent(
+                        type="text",
+                        text=json.dumps(page_content, ensure_ascii=False)
+                    )
+                ]
+            else:
+                # 如果page_content不是字典，返回空结构
+                error_result = {
+                    "count": 0,
+                    "links": [],
+                    "error": "搜索执行失败或结果格式异常"
+                }
+                return [
+                    types.TextContent(
+                        type="text",
+                        text=json.dumps(error_result, ensure_ascii=False)
+                    )
+                ]
         else:
             # 如果没有playwright，回退到传统方式
             result = search_with_direct_chrome(keywords)
             current_url = "https://kns.cnki.net/kns8s/search"
             
+            # 确保结果是字典格式
+            if not isinstance(result, dict):
+                result = {
+                    "count": 0,
+                    "links": [],
+                    "error": "需要安装playwright以获取链接: uv add playwright"
+                }
+                
             return [
                 types.TextContent(
                     type="text",
-                    text=f"{result}。如需自动搜索功能，请安装: uv add playwright"
+                    text=json.dumps(result, ensure_ascii=False)
                 )
             ]
     
@@ -637,26 +696,12 @@ async def handle_call_tool(
             )
         ]
     
-    elif name == "get-abstract-links":
-        if not page_content or "找到" not in page_content:
-            return [
-                types.TextContent(
-                    type="text",
-                    text="尚未执行搜索或未找到链接。请先使用search-keywords工具搜索。"
-                )
-            ]
-        
-        return [
-            types.TextContent(
-                type="text",
-                text=page_content
-            )
-        ]
-    
     raise ValueError(f"未知工具: {name}")
 
 async def find_and_count_abstract_links(page):
     """查找并统计包含article/abstract?v=的链接"""
+    global page_content
+    
     try:
         logger.debug("[DEBUG] 开始查找所有包含article/abstract?v=的链接")
         
@@ -690,11 +735,11 @@ async def find_and_count_abstract_links(page):
         else:
             logger.debug(f"[DEBUG] 链接数量为{links_count}条，多于预期的50条")
         
-        # 存储结果 - 只包含编号和链接，不包含标题和连字符
-        global page_content
-        page_content = f"找到{links_count}条包含article/abstract?v=的链接\n\n" + "\n".join([
-            f"{link['index']}. {link['href']}" for link in links_info
-        ])
+        # 存储结果 - 使用字典结构而不是纯文本
+        page_content = {
+            "count": links_count,
+            "links": [{"index": link['index'], "url": link['href']} for link in links_info]
+        }
         
         return links_count
     except Exception as e:
@@ -710,7 +755,7 @@ async def main():
             write_stream,
             InitializationOptions(
                 server_name="cnks",
-                server_version="0.1.0",
+                server_version="0.1.1",
                 capabilities=server.get_capabilities(
                     notification_options=NotificationOptions(),
                     experimental_capabilities={},
@@ -726,27 +771,28 @@ def create_fastmcp_server():
         fast_mcp = FastMCP("知网搜索")
         
         @fast_mcp.tool()
-        def open_cnki_search():
-            """打开中国知网搜索页面"""
-            return open_chrome("https://kns.cnki.net/kns8s/search")
-        
-        @fast_mcp.tool()
-        async def search_keywords(keywords: str) -> str:
-            """在知网搜索关键词"""
+        async def search_keywords(keywords: str) -> dict:
+            """在知网搜索关键词并返回50个论文链接"""
             logger.debug("[DEBUG] 正在使用FastMCP的search_keywords函数")
             if PLAYWRIGHT_AVAILABLE:
-                result = await search_with_playwright(keywords)
-                return result
+                result_count = await search_with_playwright(keywords)
+                # 返回结构化数据
+                if isinstance(page_content, dict):
+                    return page_content
+                else:
+                    # 如果page_content不是字典，返回空结构
+                    return {
+                        "count": 0,
+                        "links": [],
+                        "error": "搜索执行失败或结果格式异常"
+                    }
             else:
-                result = search_with_direct_chrome(keywords)
-                return f"{result}。如需自动搜索功能，请安装: uv add playwright"
-        
-        @fast_mcp.tool()
-        def get_abstract_links() -> str:
-            """获取最近一次搜索找到的论文摘要链接"""
-            if not page_content or "找到" not in page_content:
-                return "尚未执行搜索或未找到链接。请先使用search_keywords工具搜索。"
-            return page_content
+                # 直接Chrome方式不会返回链接
+                return {
+                    "count": 0,
+                    "links": [],
+                    "error": "需要安装playwright以获取链接: uv add playwright"
+                }
         
         @fast_mcp.resource("webpage://current")
         def get_current_webpage() -> str:

{cnks-0.1.0.dist-info → cnks-0.1.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cnks
-Version: 0.1.0
+Version: 0.1.1
 Summary: CNKI Search tool with Chrome browser integration
 Author-email: bai-z-l <b@iziliang.com>
 Requires-Python: >=3.13

cnks-0.1.1.dist-info/RECORD

@@ -0,0 +1,6 @@
+cnks/__init__.py,sha256=Sdhdrl3THgmQ5s_eMFKkggxDSrxP8HUN8dj63ddE6xQ,1266
+cnks/server.py,sha256=R3l0SUfajuUBU_cSrOpIDSlaVf2Z4MtXMEeHJewrkqU,32356
+cnks-0.1.1.dist-info/METADATA,sha256=4-9Nf2YNCZp7x8mWRMKf8ot41LehU2Vsxi_qLd6Tivk,22797
+cnks-0.1.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+cnks-0.1.1.dist-info/entry_points.txt,sha256=OkIiD7Ctn1Fn5B5zY09ltFFjIA8mJd3lw5V20AGtyYg,35
+cnks-0.1.1.dist-info/RECORD,,

cnks-0.1.0.dist-info/RECORD

@@ -1,6 +0,0 @@
-cnks/__init__.py,sha256=Sdhdrl3THgmQ5s_eMFKkggxDSrxP8HUN8dj63ddE6xQ,1266
-cnks/server.py,sha256=m9jCDDBmjsYs9hE8okChGGyhkLC9k82njjHHaE-Jy7c,31534
-cnks-0.1.0.dist-info/METADATA,sha256=q-7GjkhKon6H_OPiilqBGkquGd3ZiYiyRZBYpep7S1I,22797
-cnks-0.1.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-cnks-0.1.0.dist-info/entry_points.txt,sha256=OkIiD7Ctn1Fn5B5zY09ltFFjIA8mJd3lw5V20AGtyYg,35
-cnks-0.1.0.dist-info/RECORD,,