Update app.py

app.py

@@ -1,8 +1,8 @@
-import os, shutil, base64, uuid, mimetypes, json
+import os, shutil, base64, uuid, mimetypes
 from pydub import AudioSegment
 from openai import OpenAI
 import gradio as gr
-from fastapi import FastAPI, Request, HTTPException
+from fastapi import Request
 from fastapi.responses import JSONResponse
 
 # ====== 基本設定 ======
@@ -15,15 +15,9 @@ print(f"APP_PASSWORD: {'✅ 已載入' if PASSWORD else '❌ 未載入'}")
 
 # ====== 工具:把 data:URL 轉成臨時檔 ======
 MIME_EXT = {
-    "audio/mp4": "m4a",
-    "audio/m4a": "m4a",
-    "audio/aac": "aac",
-    "audio/mpeg": "mp3",
-    "audio/wav": "wav",
-    "audio/x-wav": "wav",
-    "audio/ogg": "ogg",
-    "audio/webm": "webm",
-    "audio/opus": "opus",
+    "audio/mp4": "m4a", "audio/m4a": "m4a", "audio/aac": "aac",
+    "audio/mpeg": "mp3", "audio/wav": "wav", "audio/x-wav": "wav",
+    "audio/ogg": "ogg", "audio/webm": "webm", "audio/opus": "opus",
     "video/mp4": "mp4",
 }
 
@@ -31,72 +25,39 @@ def _dataurl_to_file(data_url: str, orig_name: str | None = None) -> str:
     try:
         header, b64 = data_url.split(",", 1)
     except ValueError:
-        raise ValueError("data URL format error (missing comma).")
+        raise ValueError("data URL format error")
     mime = header.split(";")[0].split(":", 1)[-1].strip()
     ext = MIME_EXT.get(mime) or (mimetypes.guess_extension(mime) or "m4a").lstrip(".")
     fname = orig_name if (orig_name and "." in orig_name) else f"upload_{uuid.uuid4().hex}.{ext}"
     with open(fname, "wb") as f:
         f.write(base64.b64decode(b64))
-    print(f"✅ Created file from data URL: {fname}, size: {os.path.getsize(fname)} bytes")
     return fname
 
 def _extract_effective_path(file_obj) -> str:
     """從各種格式中提取有效檔案路徑"""
-    print(f"\n🔍 Starting file parsing...")
-    print(f"📦 Received type: {type(file_obj)}")
-    print(f"📦 Content: {str(file_obj)[:200]}...")
-    
     # 字串模式
     if isinstance(file_obj, str):
         s = file_obj.strip().strip('"')
-        print(f"  → Mode A: String")
         if s.startswith("data:"):
-            print(f"  → Detected data URL, length: {len(s)}")
             return _dataurl_to_file(s, None)
         if os.path.isfile(s):
-            print(f"  → Found valid path: {s}")
             return s
-    
     # 字典模式
     if isinstance(file_obj, dict):
-        print(f"  → Mode B: Dictionary")
-        print(f"  → Keys: {list(file_obj.keys())}")
-        
-        # 優先 data URL
         data = file_obj.get("data")
         if isinstance(data, str) and data.startswith("data:"):
-            print(f"  → ✅ Found data URL! Length: {len(data)}")
-            orig_name = file_obj.get("orig_name")
-            return _dataurl_to_file(data, orig_name)
-        
-        # path
+            return _dataurl_to_file(data, file_obj.get("orig_name"))
         p = str(file_obj.get("path") or "").strip().strip('"')
         if p and os.path.isfile(p):
-            print(f"  → Found path: {p}")
             return p
-        
-        # url
-        u = str(file_obj.get("url") or "").strip().strip('"')
-        if u and os.path.isfile(u):
-            print(f"  → Found url: {u}")
-            return u
-    
     # 物件模式
-    print(f"  → Mode C: Object")
     for attr in ("name", "path"):
         p = getattr(file_obj, attr, None)
         if isinstance(p, str):
             s = p.strip().strip('"')
             if os.path.isfile(s):
-                print(f"  → Found object attr {attr}: {s}")
                 return s
-    
-    data = getattr(file_obj, "data", None)
-    if isinstance(data, str) and data.startswith("data:"):
-        print(f"  → Found object data URL")
-        return _dataurl_to_file(data, getattr(file_obj, "orig_name", None))
-    
-    raise FileNotFoundError(f"Cannot parse uploaded file. Type: {type(file_obj)}")
+    raise FileNotFoundError("Cannot parse uploaded file")
 
 # ====== 分段處理 ======
 def split_audio(path):
@@ -120,21 +81,18 @@ def transcribe_core(path, model="whisper-1"):
         try:
             shutil.copy(path, fixed)
             path = fixed
-        except Exception as e:
-            print(f"⚠️ mp4→m4a failed: {e}")
-
+        except:
+            pass
     chunks = split_audio(path)
     raw = []
     for c in chunks:
         with open(c, "rb") as af:
             txt = client.audio.transcriptions.create(
-                model=model,
-                file=af,
-                response_format="text"
+                model=model, file=af, response_format="text"
             )
             raw.append(txt)
     raw_txt = "\n".join(raw)
-
+    
     conv = client.chat.completions.create(
         model="gpt-4o-mini",
         messages=[
@@ -144,7 +102,7 @@ def transcribe_core(path, model="whisper-1"):
         temperature=0.0
     )
     trad = conv.choices[0].message.content.strip()
-
+    
     summ = client.chat.completions.create(
         model="gpt-4o-mini",
         messages=[
@@ -157,76 +115,80 @@ def transcribe_core(path, model="whisper-1"):
 
 # ====== Gradio UI 函式 ======
 def transcribe_ui(password, file):
-    print("\n" + "="*50)
-    print("🎯 New transcription request (UI)")
-    print(f"🔑 Password: {password[:2]}*** (length: {len(password)})")
-    print(f"📁 File: {type(file)}")
-    print("="*50)
-    
-    if password.strip() != PASSWORD:
+    print(f"\n🎯 Web UI Request | Password: {password[:2] if password else ''}***")
+    if not password or password.strip() != PASSWORD:
         return "❌ Password incorrect", "", ""
     if not file:
         return "⚠️ No file uploaded", "", ""
-
     try:
         path = _extract_effective_path(file)
         text, summary = transcribe_core(path)
         return "✅ Transcription completed", text, summary
     except Exception as e:
-        import traceback
-        print(traceback.format_exc())
+        print(f"❌ Error: {e}")
         return f"❌ Error: {e}", "", ""
 
-# ====== Gradio UI ======
-with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("## 🎧 LINE Audio Transcription & Summary")
-    pw = gr.Textbox(label="Password", type="password")
-    f = gr.File(label="Upload Audio File")
-    run = gr.Button("Start Transcription 🚀")
-    s = gr.Textbox(label="Status", interactive=False)
-    t = gr.Textbox(label="Transcription Result", lines=10)
-    su = gr.Textbox(label="AI Summary", lines=8)
-    run.click(transcribe_ui, [pw, f], [s, t, su], queue=False)
-
-# ====== 🔴 新增:自訂 API 端點(繞過 Gradio validation) ======
-app = FastAPI()
-
-@app.post("/api/transcribe")
-async def transcribe_api(request: Request):
-    """
-    直接接受 Base64 data URL 的自訂 API
+# ====== Gradio 介面 ======
+with gr.Blocks(theme=gr.themes.Soft(), title="LINE Audio Transcription") as demo:
+    gr.Markdown("# 🎧 LINE Audio Transcription & Summary")
     
-    請求格式:
-    {
-      "password": "chou",
-      "file": {
-        "data": "data:audio/m4a;base64,UklGR...",
-        "orig_name": "recording.m4a"
-      }
-    }
-    """
+    with gr.Row():
+        with gr.Column(scale=1):
+            pw_ui = gr.Textbox(label="Password", type="password", placeholder="Enter password")
+            file_ui = gr.File(label="Upload Audio File", file_types=["audio"])
+            btn_ui = gr.Button("Start Transcription 🚀", variant="primary", size="lg")
+        
+        with gr.Column(scale=2):
+            status_ui = gr.Textbox(label="Status", interactive=False)
+            transcript_ui = gr.Textbox(label="Transcription Result", lines=10)
+            summary_ui = gr.Textbox(label="AI Summary", lines=6)
+    
+    btn_ui.click(transcribe_ui, [pw_ui, file_ui], [status_ui, transcript_ui, summary_ui])
+    
+    with gr.Accordion("📱 API Documentation (iPhone Shortcut)", open=False):
+        gr.Markdown("""
+        ### API Endpoint
+        **POST** `/api/transcribe`
+        
+        ### Request Format (JSON)
+        ```json
+        {
+          "password": "your_password",
+          "file": {
+            "data": "data:audio/m4a;base64,UklGR...",
+            "orig_name": "recording.m4a"
+          }
+        }
+        ```
+        
+        ### Response Format
+        ```json
+        {
+          "status": "success",
+          "transcription": "轉錄內容...",
+          "summary": "摘要內容..."
+        }
+        ```
+        
+        💡 **Tip**: Use this endpoint in iPhone Shortcuts for automated transcription
+        """)
+
+# ====== 自訂 API 路由 ======
+@demo.fastapi_app.post("/api/transcribe")
+async def api_transcribe(request: Request):
+    """iPhone 捷徑專用的 API 端點"""
     try:
         body = await request.json()
-        print("\n" + "="*50)
-        print("🎯 New transcription request (Custom API)")
-        print(f"📦 Request body keys: {list(body.keys())}")
-        print("="*50)
+        print(f"\n🎯 API Request | Keys: {list(body.keys())}")
         
         password = body.get("password", "")
         if password.strip() != PASSWORD:
-            return JSONResponse(
-                status_code=401,
-                content={"error": "Password incorrect"}
-            )
+            return JSONResponse(status_code=401, content={"error": "Password incorrect"})
         
         file_obj = body.get("file")
         if not file_obj:
-            return JSONResponse(
-                status_code=400,
-                content={"error": "No file provided"}
-            )
+            return JSONResponse(status_code=400, content={"error": "No file provided"})
         
-        # 直接處理 Base64
         path = _extract_effective_path(file_obj)
         text, summary = transcribe_core(path)
         
@@ -235,19 +197,11 @@ async def transcribe_api(request: Request):
             "transcription": text,
             "summary": summary
         })
-        
     except Exception as e:
         import traceback
-        error_detail = traceback.format_exc()
-        print(f"❌ API Error:\n{error_detail}")
-        return JSONResponse(
-            status_code=500,
-            content={"error": str(e), "detail": error_detail}
-        )
-
-# 掛載 Gradio 到 FastAPI
-app = gr.mount_gradio_app(app, demo, path="/")
+        print(f"❌ API Error:\n{traceback.format_exc()}")
+        return JSONResponse(status_code=500, content={"error": str(e)})
 
+# ====== 啟動 ======
 if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    demo.launch(server_name="0.0.0.0", server_port=7860)