@youhaozhao/cninfo-mcp 1.0.6 → 1.1.0

package/README.md

@@ -2,7 +2,7 @@
 
 [![npm version](https://img.shields.io/npm/v/@youhaozhao/cninfo-mcp)](https://www.npmjs.com/package/@youhaozhao/cninfo-mcp)
 
-通过 MCP 协议查询和下载巨潮资讯网上市公司年报的工具，适用于 Claude Desktop。
+通过 MCP 协议查询和下载巨潮资讯网上市公司年报 PDF 的工具，适用于 Claude Desktop。
 
 ## 使用方法
 
@@ -50,3 +50,4 @@
 
 爬虫逻辑基于 [gaodechen/cninfo_process](https://github.com/gaodechen/cninfo_process)。
 
+

package/bin/cninfo-mcp.js

@@ -5,22 +5,45 @@
  * 自动检测 Python 并安装依赖，然后启动 Python MCP 服务器。
  */
 
-const { spawn } = require('child_process');
-const path = require('path');
-const fs = require('fs');
+const { spawn } = require("child_process");
+const path = require("path");
+const fs = require("fs");
+const os = require("os");
 
 // 配置路径
-const PYTHON_SCRIPT = path.join(__dirname, '..', 'python', 'mcp_server.py');
-const PYTHON_REQUIREMENTS = path.join(__dirname, '..', 'python', 'requirements.txt');
+const PYTHON_SCRIPT = path.join(__dirname, "..", "python", "mcp_server.py");
+const PYTHON_REQUIREMENTS = path.join(
+  __dirname,
+  "..",
+  "python",
+  "requirements.txt",
+);
+
+// 虚拟环境目录，放在用户目录下保证跨 npx 调用持久化
+const VENV_DIR = path.join(os.homedir(), ".cninfo-mcp", "venv");
+
+// 获取虚拟环境中的 Python 可执行文件路径
+function getVenvPython() {
+  if (process.platform === "win32") {
+    return path.join(VENV_DIR, "Scripts", "python.exe");
+  }
+  return path.join(VENV_DIR, "bin", "python3");
+}
 
-// 查找可用的 Python 可执行文件
+// 查找可用的系统 Python 可执行文件（仅用于创建 venv）
 async function findPython() {
-  const pythonCommands = ['python3', 'python', 'python3.12', 'python3.11', 'python3.10'];
+  const pythonCommands = [
+    "python3",
+    "python",
+    "python3.12",
+    "python3.11",
+    "python3.10",
+  ];
 
   for (const cmd of pythonCommands) {
     try {
-      const result = await spawnAsync(cmd, ['--version']);
-      if (result.stdout && result.stdout.includes('Python')) {
+      const result = await spawnAsync(cmd, ["--version"]);
+      if (result.stdout && result.stdout.includes("Python")) {
         return cmd;
       }
     } catch (error) {
@@ -29,38 +52,57 @@ async function findPython() {
   }
 
   throw new Error(
-    'Python not found. Please install Python 3.10+ from https://python.org\n' +
-    'After installation, restart your terminal and try again.'
+    "Python not found. Please install Python 3.10+ from https://python.org\n" +
+      "After installation, restart your terminal and try again.",
   );
 }
 
-// 检查并安装 Python 依赖
-async function ensureDependencies(pythonCmd) {
+// 创建虚拟环境（如果不存在）
+async function ensureVenv(systemPythonCmd) {
+  const venvPython = getVenvPython();
+  if (fs.existsSync(venvPython)) {
+    return venvPython;
+  }
+
+  console.error("Creating Python virtual environment...");
+  fs.mkdirSync(path.dirname(VENV_DIR), { recursive: true });
+  await spawnAsync(systemPythonCmd, ["-m", "venv", VENV_DIR], {
+    stdio: "inherit",
+  });
+  console.error("Virtual environment created\n");
+  return venvPython;
+}
+
+// 检查并安装 Python 依赖（使用 venv 中的 python）
+async function ensureDependencies(venvPython) {
   const requirementsPath = PYTHON_REQUIREMENTS;
 
   if (!fs.existsSync(requirementsPath)) {
-    console.error('Error: requirements.txt not found at', requirementsPath);
+    console.error("Error: requirements.txt not found at", requirementsPath);
     process.exit(1);
   }
 
   try {
     // 检查 mcp 包是否已安装
-    const checkResult = await spawnAsync(pythonCmd, ['-c', 'import mcp']);
+    await spawnAsync(venvPython, ["-c", "import mcp"]);
   } catch (error) {
     // 未安装，执行安装
-    console.error('Installing Python dependencies...');
-    const installResult = await spawnAsync(pythonCmd, ['-m', 'pip', 'install', '-r', requirementsPath], {
-      stdio: 'inherit'
-    });
-
-    if (installResult.code !== 0) {
-      console.error('\n❌ Failed to install Python dependencies');
-      console.error('Please run manually:');
-      console.error(`  ${pythonCmd} -m pip install -r ${requirementsPath}`);
+    console.error("Installing Python dependencies...");
+    try {
+      await spawnAsync(
+        venvPython,
+        ["-m", "pip", "install", "-r", requirementsPath],
+        {
+          stdio: "inherit",
+        },
+      );
+      console.error("Python dependencies installed successfully\n");
+    } catch (installError) {
+      console.error("\n❌ Failed to install Python dependencies");
+      console.error("Please run manually:");
+      console.error(`  ${venvPython} -m pip install -r ${requirementsPath}`);
       process.exit(1);
     }
-
-    console.error('✅ Python dependencies installed successfully\n');
   }
 }
 
@@ -68,28 +110,28 @@ async function ensureDependencies(pythonCmd) {
 function spawnAsync(command, args, options = {}) {
   return new Promise((resolve, reject) => {
     const child = spawn(command, args, {
-      stdio: options.stdio || 'pipe',
-      shell: process.platform === 'win32',
-      ...options
+      stdio: options.stdio || "pipe",
+      shell: process.platform === "win32",
+      ...options,
     });
 
-    let stdout = '';
-    let stderr = '';
+    let stdout = "";
+    let stderr = "";
     let code = null;
 
     if (child.stdout) {
-      child.stdout.on('data', (data) => {
+      child.stdout.on("data", (data) => {
         stdout += data.toString();
       });
     }
 
     if (child.stderr) {
-      child.stderr.on('data', (data) => {
+      child.stderr.on("data", (data) => {
         stderr += data.toString();
       });
     }
 
-    child.on('close', (exitCode) => {
+    child.on("close", (exitCode) => {
       code = exitCode;
       if (code === 0) {
         resolve({ stdout, stderr, code });
@@ -102,7 +144,7 @@ function spawnAsync(command, args, options = {}) {
       }
     });
 
-    child.on('error', (error) => {
+    child.on("error", (error) => {
       reject(error);
     });
   });
@@ -112,36 +154,36 @@ async function main() {
   try {
     // 检查 Python 脚本是否存在
     if (!fs.existsSync(PYTHON_SCRIPT)) {
-      console.error('Error: mcp_server.py not found at', PYTHON_SCRIPT);
+      console.error("Error: mcp_server.py not found at", PYTHON_SCRIPT);
       process.exit(1);
     }
 
-    const pythonCmd = await findPython();
-    await ensureDependencies(pythonCmd);
+    const systemPython = await findPython();
+    const venvPython = await ensureVenv(systemPython);
+    await ensureDependencies(venvPython);
 
     // 启动 MCP 服务器
-    console.error('巨潮资讯 MCP 服务器已启动，等待连接...');
-    const child = spawn(pythonCmd, [PYTHON_SCRIPT], {
-      stdio: 'inherit',
-      shell: process.platform === 'win32',
+    console.error("巨潮资讯 MCP 服务器已启动，等待连接...");
+    const child = spawn(venvPython, [PYTHON_SCRIPT], {
+      stdio: "inherit",
+      shell: process.platform === "win32",
       env: {
         ...process.env,
-        PYTHONPATH: path.join(__dirname, '..', 'python')
-      }
+        PYTHONPATH: path.join(__dirname, "..", "python"),
+      },
     });
 
     // 处理子进程退出
-    child.on('error', (error) => {
-      console.error('Failed to start MCP Server:', error.message);
+    child.on("error", (error) => {
+      console.error("Failed to start MCP Server:", error.message);
       process.exit(1);
     });
 
-    child.on('exit', (code) => {
+    child.on("exit", (code) => {
       process.exit(code || 0);
     });
-
   } catch (error) {
-    console.error('Error:', error.message);
+    console.error("Error:", error.message);
     process.exit(1);
   }
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@youhaozhao/cninfo-mcp",
-  "version": "1.0.6",
+  "version": "1.1.0",
   "description": "MCP Server for querying and downloading Chinese listed companies' annual reports from CNINFO (巨潮资讯网)",
   "keywords": [
     "mcp",
@@ -50,4 +50,4 @@
     "README.md",
     "LICENSE"
   ]
-}
+}

package/python/bridge.py

@@ -0,0 +1,161 @@
+#!/usr/bin/env python3
+"""CLI bridge that exposes cninfo spider functions as JSON in/out for Node.js."""
+
+import json
+import os
+import sys
+import traceback
+from typing import Any, Optional
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+from spider import (  # noqa: E402
+    download_annual_reports,
+    download_prospectus,
+    query_annual_reports,
+    query_prospectus,
+    saving_path,
+)
+
+
+BASE_URL = "https://static.cninfo.com.cn/"
+
+
+def _format_reports(reports: list[dict]) -> list[dict]:
+    formatted = []
+    for report in reports:
+        adj = report.get("adjunctUrl", "")
+        formatted.append(
+            {
+                "announcementTitle": report.get("announcementTitle", ""),
+                "announcementTime": report.get("announcementTime", ""),
+                "secCode": report.get("secCode", ""),
+                "secName": report.get("secName", ""),
+                "adjunctUrl": BASE_URL + adj if adj else "",
+            }
+        )
+    return formatted
+
+
+def _require_stock_code(payload: dict) -> str:
+    stock_code = (payload.get("stock_code") or "").strip()
+    if not stock_code:
+        raise ValueError("stock_code is required")
+    return stock_code
+
+
+def _optional_year(payload: dict) -> Optional[int]:
+    year = payload.get("year")
+    if year is None:
+        return None
+    return int(year)
+
+
+def _resolve_save_path(payload: dict) -> str:
+    save_path = payload.get("save_path")
+    return save_path if save_path else saving_path
+
+
+def action_query_annual_reports(payload: dict) -> dict:
+    stock_code = _require_stock_code(payload)
+    year = _optional_year(payload)
+    reports = query_annual_reports(stock_code, year)
+    suffix = f" for year {year}" if year else ""
+    if not reports:
+        return {
+            "success": False,
+            "stock_code": stock_code,
+            "year": year,
+            "count": 0,
+            "reports": [],
+            "message": f"No annual reports found for stock {stock_code}{suffix}",
+        }
+    return {
+        "success": True,
+        "stock_code": stock_code,
+        "year": year,
+        "count": len(reports),
+        "reports": _format_reports(reports),
+        "message": f"Found {len(reports)} annual report(s){suffix}",
+    }
+
+
+def action_download_annual_reports(payload: dict) -> dict:
+    stock_code = _require_stock_code(payload)
+    year = _optional_year(payload)
+    output_dir = _resolve_save_path(payload)
+    os.makedirs(output_dir, exist_ok=True)
+    result = download_annual_reports(stock_code, year, save_path=output_dir)
+    result["stock_code"] = stock_code
+    result["year"] = year
+    return result
+
+
+def action_query_prospectus(payload: dict) -> dict:
+    stock_code = _require_stock_code(payload)
+    reports = query_prospectus(stock_code)
+    if not reports:
+        return {
+            "success": False,
+            "stock_code": stock_code,
+            "count": 0,
+            "reports": [],
+            "message": f"No prospectus found for stock {stock_code}",
+        }
+    return {
+        "success": True,
+        "stock_code": stock_code,
+        "count": len(reports),
+        "reports": _format_reports(reports),
+        "message": f"Found {len(reports)} prospectus document(s)",
+    }
+
+
+def action_download_prospectus(payload: dict) -> dict:
+    stock_code = _require_stock_code(payload)
+    output_dir = _resolve_save_path(payload)
+    os.makedirs(output_dir, exist_ok=True)
+    result = download_prospectus(stock_code, save_path=output_dir)
+    result["stock_code"] = stock_code
+    return result
+
+
+ACTIONS = {
+    "query_annual_reports": action_query_annual_reports,
+    "download_annual_reports": action_download_annual_reports,
+    "query_prospectus": action_query_prospectus,
+    "download_prospectus": action_download_prospectus,
+}
+
+
+def main() -> int:
+    if len(sys.argv) < 3:
+        sys.stderr.write("usage: bridge.py <action> <json-payload>\n")
+        return 2
+
+    action = sys.argv[1]
+    raw_payload = sys.argv[2]
+
+    handler = ACTIONS.get(action)
+    if handler is None:
+        sys.stdout.write(json.dumps({"ok": False, "error": f"Unknown action: {action}"}))
+        return 1
+
+    try:
+        payload = json.loads(raw_payload) if raw_payload else {}
+    except json.JSONDecodeError as exc:
+        sys.stdout.write(json.dumps({"ok": False, "error": f"Invalid JSON payload: {exc}"}))
+        return 1
+
+    try:
+        result: Any = handler(payload)
+        sys.stdout.write(json.dumps({"ok": True, "payload": result}, ensure_ascii=False))
+        return 0
+    except Exception as exc:
+        sys.stderr.write(traceback.format_exc())
+        sys.stdout.write(json.dumps({"ok": False, "error": str(exc)}))
+        return 1
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

package/python/spider.py

@@ -18,6 +18,13 @@ _saving_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "pdf")
 saving_path = _saving_path + "/"
 logger = logging.getLogger(__name__)
 
+# 巨潮资讯历史公告的实际下限约为 2001 会计年度，再往前查询无数据返回。
+EARLIEST_DATE = "2001-01-01"
+# 接口单页最大返回条数
+PAGE_SIZE = 30
+# 翻页安全上限，防止异常情况下无限循环
+MAX_PAGES = 100
+
 User_Agent = [
     "Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Win64; x64; Trident/5.0; .NET CLR 3.5.30729; .NET CLR 3.0.30729; .NET CLR 2.0.50727; Media Center PC 6.0)",
     "Mozilla/5.0 (compatible; MSIE 8.0; Windows NT 6.0; Trident/4.0; WOW64; Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729; .NET CLR 1.0.3705; .NET CLR 1.1.4322)",
@@ -55,6 +62,26 @@ def _date_range(start_date: str) -> str:
     return f"{start_date}~{today}"
 
 
+def _paginate(fetch_fn, stock):
+    """
+    对单页查询函数翻页，汇总所有页的公告。
+
+    巨潮接口单页最多返回 PAGE_SIZE 条，放开时间区间后历史年报会跨越多页，
+    必须翻页才能取全。以“返回数量不足一页”作为终止条件，并设安全上限。
+    """
+    all_items = []
+    for page in range(1, MAX_PAGES + 1):
+        items = fetch_fn(page, stock)
+        if not items:
+            break
+        all_items.extend(items)
+        if len(items) < PAGE_SIZE:  # 不足一页说明已到最后一页
+            break
+    else:
+        logger.warning("翻页达到上限 %s，结果可能被截断（%s）", MAX_PAGES, stock)
+    return all_items
+
+
 def _is_annual_report_title(
     title: str, year_filter: Optional[Union[int, str]] = None
 ) -> bool:
@@ -101,7 +128,7 @@ def szseAnnual(page, stock):
     query_path = "http://www.cninfo.com.cn/new/hisAnnouncement/query"
     query = {
         "pageNum": page,  # 页码
-        "pageSize": 30,
+        "pageSize": PAGE_SIZE,
         "tabName": "fulltext",
         "column": "szse",  # 深交所
         "stock": "",
@@ -110,7 +137,7 @@ def szseAnnual(page, stock):
         "plate": "sz",
         "category": "category_ndbg_szsh",  # 年度报告
         "trade": "",
-        "seDate": _date_range("2020-01-01"),  # 时间区间
+        "seDate": _date_range(EARLIEST_DATE),  # 时间区间
     }
 
     namelist = requests.post(
@@ -127,7 +154,7 @@ def sseAnnual(page, stock):
     query_path = "http://www.cninfo.com.cn/new/hisAnnouncement/query"
     query = {
         "pageNum": page,  # 页码
-        "pageSize": 30,
+        "pageSize": PAGE_SIZE,
         "tabName": "fulltext",
         "column": "sse",
         "stock": "",
@@ -136,7 +163,7 @@ def sseAnnual(page, stock):
         "plate": "sh",
         "category": "category_ndbg_szsh",  # 年度报告
         "trade": "",
-        "seDate": _date_range("2020-01-01"),  # 时间区间
+        "seDate": _date_range(EARLIEST_DATE),  # 时间区间
     }
 
     namelist = requests.post(
@@ -153,7 +180,7 @@ def szseStock(page, stock):
     query_path = "http://www.cninfo.com.cn/new/hisAnnouncement/query"
     query = {
         "pageNum": page,  # 页码
-        "pageSize": 30,
+        "pageSize": PAGE_SIZE,
         "tabName": "fulltext",
         "column": "szse",
         "stock": "",
@@ -162,7 +189,7 @@ def szseStock(page, stock):
         "plate": "sz",
         "category": "",
         "trade": "",
-        "seDate": _date_range("2015-01-01"),  # 时间区间
+        "seDate": _date_range(EARLIEST_DATE),  # 时间区间
     }
 
     namelist = requests.post(
@@ -179,7 +206,7 @@ def sseStock(page, stock):
     query_path = "http://www.cninfo.com.cn/new/hisAnnouncement/query"
     query = {
         "pageNum": page,  # 页码
-        "pageSize": 30,
+        "pageSize": PAGE_SIZE,
         "tabName": "fulltext",
         "column": "sse",
         "stock": "",
@@ -188,7 +215,7 @@ def sseStock(page, stock):
         "plate": "sh",
         "category": "",
         "trade": "",
-        "seDate": _date_range("2015-01-01"),  # 时间区间
+        "seDate": _date_range(EARLIEST_DATE),  # 时间区间
     }
 
     namelist = requests.post(
@@ -271,13 +298,13 @@ def query_prospectus(stock_code):
     all_announcements = []
 
     try:
-        announcements_sse = sseStock(1, stock_code)
+        announcements_sse = _paginate(sseStock, stock_code)
         all_announcements.extend(announcements_sse)
     except Exception as e:
         logger.warning("沪市招股书查询失败: %s", e)
 
     try:
-        announcements_szse = szseStock(1, stock_code)
+        announcements_szse = _paginate(szseStock, stock_code)
         all_announcements.extend(announcements_szse)
     except Exception as e:
         logger.warning("深市招股书查询失败: %s", e)
@@ -323,14 +350,14 @@ def query_annual_reports(stock_code, year=None):
 
     # 查询沪市
     try:
-        announcements_sse = sseAnnual(1, stock_code)
+        announcements_sse = _paginate(sseAnnual, stock_code)
         all_announcements.extend(announcements_sse)
     except Exception as e:
         logger.warning("沪市年报查询失败: %s", e)
 
     # 查询深市
     try:
-        announcements_szse = szseAnnual(1, stock_code)
+        announcements_szse = _paginate(szseAnnual, stock_code)
         all_announcements.extend(announcements_szse)
     except Exception as e:
         logger.warning("深市年报查询失败: %s", e)

package/scripts/install-python-deps.js

@@ -7,6 +7,7 @@
 const { spawn } = require("child_process");
 const fs = require("fs");
 const path = require("path");
+const os = require("os");
 
 const REQUIREMENTS_FILE = path.join(
   __dirname,
@@ -15,6 +16,15 @@ const REQUIREMENTS_FILE = path.join(
   "requirements.txt",
 );
 
+const VENV_DIR = path.join(os.homedir(), ".cninfo-mcp", "venv");
+
+function getVenvPython() {
+  if (process.platform === "win32") {
+    return path.join(VENV_DIR, "Scripts", "python.exe");
+  }
+  return path.join(VENV_DIR, "bin", "python3");
+}
+
 async function findPython() {
   const pythonCommands = [
     "python3",
@@ -75,16 +85,31 @@ async function main() {
     return;
   }
 
+  // 创建虚拟环境（如果不存在）
+  const venvPython = getVenvPython();
+  if (!fs.existsSync(venvPython)) {
+    console.log("Creating Python virtual environment...");
+    try {
+      fs.mkdirSync(path.dirname(VENV_DIR), { recursive: true });
+      await spawnCommand(pythonCmd, ["-m", "venv", VENV_DIR]);
+      console.log("Virtual environment created");
+    } catch (venvError) {
+      console.warn("  Failed to create virtual environment during npm install");
+      console.warn("  It will be created automatically on first run");
+      return;
+    }
+  }
+
   try {
-    // 检查 mcp 是否已安装
-    await spawnCommand(pythonCmd, ["-c", "import mcp"]);
+    // 检查 mcp 是否已安装（用 venv 的 python）
+    await spawnCommand(venvPython, ["-c", "import mcp"]);
     console.log("✅ Python dependencies already installed");
   } catch (error) {
-    // 执行安装
+    // 执行安装（用 venv 的 pip）
     console.log("📦 Installing Python dependencies...");
     try {
       await spawnCommand(
-        pythonCmd,
+        venvPython,
         ["-m", "pip", "install", "-r", REQUIREMENTS_FILE],
         {
           stdio: "inherit",