feat: support both JSON and form-urlencoded body; change default port to 3333

Dockerfile

@@ -12,6 +12,6 @@ COPY app/ .
 
 RUN mkdir -p /app/data /app/audio
 
-EXPOSE 17200
+EXPOSE 3333
 
-CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "17200"]
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "3333"]

app/config.py

@@ -12,7 +12,7 @@ MIMO_VOICE = os.environ.get("MIMO_VOICE", "mimo_default")
 
 # 服务器配置
 SERVER_HOST = os.environ.get("SERVER_HOST", "0.0.0.0")
-SERVER_PORT = int(os.environ.get("SERVER_PORT", "17200"))
+SERVER_PORT = int(os.environ.get("SERVER_PORT", "3333"))
 
 # 音频存储目录
 AUDIO_DIR = os.environ.get("AUDIO_DIR", os.path.join(BASE_DIR, "audio"))

app/main.py

@@ -219,27 +219,43 @@ async def get_chapter_audio(book_id: str, chapter_id: str):
     return FileResponse(chapter.audio_file, media_type="audio/mpeg", filename=f"{chapter_id}.mp3")
 
 
-# ── 实时 TTS 接口（兼容百度/阿里云风格）──────────────────────────────────
+# ── 实时 TTS 接口（兼容听书 App 格式）─────────────────────────────────────
 
 from fastapi.responses import Response
+from urllib.parse import parse_qs
 
 @app.post("/api/tts")
 async def realtime_tts(request: Request):
     """
     实时 TTS 生成接口
-    请求体 (JSON):
+    兼容两种 App 发送格式:
-      - text: 要合成的文本（必填）
+      1. JSON body:  {"text": "内容", "style": "开心"}
-      - style: 风格（可选，如"开心"、"东北话"）
+      2. Form body:  tex=内容&spd=5 (百度风格)
-      - speed: 语速调整（可选，暂未使用，MiMo 通过 style 控制语速）
+    返回: MP3 音频二进制流 (audio/mpeg)，失败返回 JSON
-    返回: MP3 音频二进制流 (audio/mpeg)
     """
-    data = await request.json()
+    text = ""
-    text = data.get("text", "").strip()
+    style = ""
-    style = data.get("style", "").strip()
+    content_type = request.headers.get("content-type", "")
+
+    try:
+        if "json" in content_type:
+            data = await request.json()
+            text = data.get("text", "").strip()
+            style = data.get("style", "").strip()
+        else:
+            # form-urlencoded (百度风格)
+            body_bytes = await request.body()
+            params = parse_qs(body_bytes.decode("utf-8"))
+            text = (params.get("tex", [""])[0]).strip()
+            # URL 解码（百度会 double-encode）
+            from urllib.parse import unquote
+            text = unquote(unquote(text))
+    except Exception:
+        pass
 
     if not text:
         return Response(
-            content=json.dumps({"status": 40000001, "message": "text 不能为空"}, ensure_ascii=False),
+            content=json.dumps({"status": 40000001, "message": "text/tex 不能为空"}, ensure_ascii=False),
             media_type="application/json",
             status_code=400,
         )
@@ -248,7 +264,7 @@ async def realtime_tts(request: Request):
         # MiMo TTS 生成 WAV
         wav_bytes = await call_mimo_tts(text, style)
 
-        # WAV → MP3（内存中完成）
+        # WAV → MP3（临时文件）
         tmp_dir = Path(config.AUDIO_DIR) / "_tmp"
         tmp_dir.mkdir(parents=True, exist_ok=True)
         tmp_id = uuid.uuid4().hex
@@ -270,12 +286,6 @@ async def realtime_tts(request: Request):
 
         return Response(content=mp3_bytes, media_type="audio/mpeg")
 
-    except HTTPException:
-        return Response(
-            content=json.dumps({"status": 50000001, "message": "TTS 生成失败"}, ensure_ascii=False),
-            media_type="application/json",
-            status_code=502,
-        )
     except Exception as e:
         return Response(
             content=json.dumps({"status": 50000002, "message": str(e)[:300]}, ensure_ascii=False),

docker-compose.yml

@@ -5,13 +5,13 @@ services:
     build: .
     container_name: tts-book-service
     ports:
-      - "17200:17200"
+      - "3333:3333"
     environment:
       - MIMO_API_KEY=${MIMO_API_KEY}
       - MIMO_API_ENDPOINT=${MIMO_API_ENDPOINT:-https://api.xiaomimimo.com/v1/chat/completions}
       - MIMO_TTS_MODEL=${MIMO_TTS_MODEL:-mimo-v2-audio-tts}
       - MIMO_VOICE=${MIMO_VOICE:-mimo_default}
-      - SERVER_PORT=17200
+      - SERVER_PORT=3333
     volumes:
       - tts-data:/app/data
       - tts-audio:/app/audio