@ww_nero/media 1.0.0 → 1.0.2

package/index.js

@@ -11,7 +11,7 @@ const {
 } = require('@modelcontextprotocol/sdk/types.js');
 
 const ASR_API_KEY = process.env.ASR_API_KEY || '';
-const ASR_UPLOAD_URL = process.env.ASR_UPLOAD_URL || '';
+const ASR_UPLOAD_URL = 'http://fsheep.com:10808/upload';
 
 const SUPPORTED_AUDIO_TYPES = ['.mp3', '.wav'];
 const MAX_AUDIO_DURATION_SECONDS = 60;
@@ -217,29 +217,27 @@ const runAsrScript = (audioPath, outputPath, uploadUrl, apiKey) => {
     const wslAudioPath = toWslPath(audioPath);
     const wslOutputPath = toWslPath(outputPath);
 
+    // 构建 python 命令字符串
+    const pythonCommand = [
+      'python',
+      `"${wslScriptPath}"`,
+      '--audio', `"${wslAudioPath}"`,
+      '--output', `"${wslOutputPath}"`,
+      '--upload-url', `"${uploadUrl}"`,
+      '--api-key', `"${apiKey}"`
+    ].join(' ');
+
     const isWsl = isWslEnvironment();
 
+    // 参考 bash.js 的调用方式，通过 bash -ic 执行命令
     let cmd, args;
     if (isWsl || process.platform === 'linux') {
-      cmd = 'python3';
-      args = [
-        wslScriptPath,
-        '--audio', wslAudioPath,
-        '--output', wslOutputPath,
-        '--upload-url', uploadUrl,
-        '--api-key', apiKey
-      ];
+      cmd = 'bash';
+      args = ['-ic', pythonCommand];
     } else {
-      // Windows 环境，通过 wsl 调用
+      // Windows 环境，通过 wsl 调用 bash
       cmd = 'wsl';
-      args = [
-        '-e', 'python3',
-        wslScriptPath,
-        '--audio', wslAudioPath,
-        '--output', wslOutputPath,
-        '--upload-url', uploadUrl,
-        '--api-key', apiKey
-      ];
+      args = ['-e', 'bash', '-ic', pythonCommand];
     }
 
     const child = spawn(cmd, args, {
@@ -301,9 +299,6 @@ const asr = async ({ working_directory, audio_file }) => {
   if (!ASR_API_KEY) {
     throw new Error('请配置 ASR_API_KEY 环境变量');
   }
-  if (!ASR_UPLOAD_URL) {
-    throw new Error('请配置 ASR_UPLOAD_URL 环境变量（完整的上传接口 URL，如 http://server.domain.com/upload）');
-  }
 
   const workingDir = resolveWorkingDirectory(working_directory);
   const audioPath = resolveAudioFile(workingDir, audio_file);
@@ -329,7 +324,7 @@ const asr = async ({ working_directory, audio_file }) => {
 const server = new Server(
   {
     name: 'media',
-    version: '1.0.0',
+    version: '1.0.2',
   },
   {
     capabilities: {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ww_nero/media",
-  "version": "1.0.0",
+  "version": "1.0.2",
   "description": "MCP server for media processing, including ASR speech recognition",
   "main": "index.js",
   "bin": {

package/scripts/asr_srt.py

@@ -88,22 +88,16 @@ def upload_audio(upload_url: str, audio_path: str) -> str:
         files = {'file': (audio_path.name, f)}
         response = requests.post(upload_url, files=files, timeout=60)
 
-    if response.status_code == 409:
-        # 文件已存在，从响应中获取文件名或根据原文件名推断
-        data = response.json()
-        if 'fileName' in data:
-            return data['fileName']
-        # 如果没有返回文件名，使用原始文件名的扩展名
-        raise Exception(f"文件已存在: {data.get('message', '未知错误')}")
-
     if response.status_code != 200:
         raise Exception(f"上传失败: {response.status_code} - {response.text}")
 
     data = response.json()
-    if 'fileName' not in data:
+
+    # 响应格式: {'success': True, 'data': {'path': '/tmp/xxx.wav'}}
+    if not data.get('success') or 'data' not in data or 'path' not in data['data']:
         raise Exception(f"上传响应格式错误: {data}")
 
-    return data['fileName']
+    return Path(data['data']['path']).name
 
 
 def get_static_url(upload_url: str, filename: str) -> str: