@ww_nero/media 1.0.2 → 1.0.4

package/index.js

@@ -324,7 +324,7 @@ const asr = async ({ working_directory, audio_file }) => {
 const server = new Server(
   {
     name: 'media',
-    version: '1.0.2',
+    version: '1.0.4',
   },
   {
     capabilities: {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ww_nero/media",
-  "version": "1.0.2",
+  "version": "1.0.4",
   "description": "MCP server for media processing, including ASR speech recognition",
   "main": "index.js",
   "bin": {

package/scripts/asr_srt.py

@@ -97,26 +97,28 @@ def upload_audio(upload_url: str, audio_path: str) -> str:
     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 Path(data['data']['path']).name
+    # 返回完整的相对路径，用于拼接到服务器地址后面
+    return data['data']['path']
 
 
-def get_static_url(upload_url: str, filename: str) -> str:
+def get_static_url(upload_url: str, file_path: str) -> str:
     """
-    根据上传接口 URL 和文件名构建静态资源 URL
+    根据上传接口 URL 和文件路径构建静态资源 URL
 
     Args:
         upload_url: 上传接口的完整 URL (如 http://server.domain.com/upload)
-        filename: 上传后的文件名
+        file_path: 上传后返回的文件路径 (如 /tmp/xxx.wav)
 
     Returns:
         静态资源的完整 URL
     """
     # 从上传 URL 中提取基础 URL
-    # 例如: http://server.domain.com/upload -> http://server.domain.com/
+    # 例如: http://server.domain.com/upload -> http://server.domain.com
     from urllib.parse import urlparse, urlunparse
     parsed = urlparse(upload_url)
-    base_url = urlunparse((parsed.scheme, parsed.netloc, '/', '', '', ''))
-    return f"{base_url.rstrip('/')}/{filename}"
+    base_url = urlunparse((parsed.scheme, parsed.netloc, '', '', '', ''))
+    # file_path 已经是以 / 开头的完整相对路径，直接拼接
+    return f"{base_url}{file_path}"
 
 
 def transcribe_audio(audio_url: str, api_key: str) -> list:
@@ -193,11 +195,11 @@ def main():
     try:
         # 1. 上传音频文件
         print(f"正在上传音频文件: {args.audio}")
-        filename = upload_audio(args.upload_url, args.audio)
-        print(f"上传成功: {filename}")
+        file_path = upload_audio(args.upload_url, args.audio)
+        print(f"上传成功: {file_path}")
 
         # 2. 构建静态资源 URL
-        audio_url = get_static_url(args.upload_url, filename)
+        audio_url = get_static_url(args.upload_url, file_path)
         print(f"音频 URL: {audio_url}")
 
         # 3. 调用 ASR 识别