apps/api/app/api/routes/voices.py

@@ -9,7 +9,7 @@ from typing import Literal, Optional
 
 from app.api.routes._helpers import get_user_plan
 from app.auth import AuthenticatedUser, get_current_user
-from app.dependencies import get_audio_url_signer, get_db_session, get_user_repository
+from app.dependencies import get_audio_url_signer, get_cosyvoice_service, get_db_session, get_user_repository
 from app.schemas.voice import (
     PresetVoiceItemResponse,
     PresetVoiceListResponse,
@@ -27,6 +27,7 @@ from sqlalchemy.orm import Session
 
 from packages.adapters.sqlalchemy_impl.voice_clone_profile_repository import SQLAlchemyVoiceCloneProfileRepository
 from packages.adapters.sqlalchemy_impl.voice_library_repository import SQLAlchemyVoiceLibraryRepository
+from packages.application.cosyvoice_service import CosyVoiceError, CosyVoiceService
 from packages.application.voice_library.commands import CreateVoiceLibraryCommand, UpdateVoiceLibraryCommand
 from packages.application.voice_library.use_cases import (
     CreateVoiceLibraryUseCase,
@@ -37,11 +38,18 @@ from packages.application.voice_library.use_cases import (
     QuotaExceededError,
     UpdateVoiceLibraryUseCase,
 )
-from packages.domain.preset_voices import PRESET_VOICES
+from packages.domain.preset_voices import PRESET_VOICES, get_preset_voice_by_id
 from packages.ports.user_repository import UserRepository
 
 router = APIRouter()
 
+# 预置音色试听音频缓存（内存缓存，减少重复TTS调用）
+# key: voice_id, value: (audio_url, timestamp)
+_preset_preview_cache: dict[str, tuple[str, float]] = {}
+PREVIEW_CACHE_TTL = 7 * 24 * 3600  # 7天TTL
+# 每个预置音色的默认试听文本
+PREVIEW_TEMPLATE = "你好，我是{name}，很高兴认识你。"
+
 
 def _get_voice_repository(session: Session = Depends(get_db_session)) -> SQLAlchemyVoiceLibraryRepository:
     return SQLAlchemyVoiceLibraryRepository(session)
@@ -216,6 +224,58 @@ def list_preset_voices() -> PresetVoiceListResponse:
     return PresetVoiceListResponse(items=items, total=len(items))
 
 
+@router.get("/presets/{voice_id}/preview")
+def get_preset_voice_preview(
+    voice_id: str,
+    text: str = Query("", description="自定义试听文本，为空则使用默认示例"),
+    cosyvoice: CosyVoiceService = Depends(get_cosyvoice_service),
+) -> dict:
+    """获取预置音色试听音频（实时 TTS 合成）。
+
+    - 首次调用会合成并缓存7天
+    - 相同 voice_id 重复调用直接返回缓存的音频URL
+    - 可传入自定义 text 参数试听不同文本
+    """
+    import time
+
+    preset = get_preset_voice_by_id(voice_id)
+    if preset is None:
+        raise HTTPException(status_code=404, detail=f"预置音色不存在: {voice_id}")
+
+    # 有自定义文本时不缓存
+    use_cache = not text.strip()
+
+    if use_cache and voice_id in _preset_preview_cache:
+        audio_url, cached_at = _preset_preview_cache[voice_id]
+        if time.time() - cached_at < PREVIEW_CACHE_TTL:
+            return {"voice_id": voice_id, "audio_url": audio_url, "cached": True}
+
+    # 合成试听音频
+    preview_text = text.strip() or PREVIEW_TEMPLATE.format(name=preset.name)
+    try:
+        result = cosyvoice.synthesize_speech(
+            text=preview_text,
+            voice_id=preset.voice_id,
+            format="mp3",
+            speed=1.0,
+        )
+    except CosyVoiceError as e:
+        raise HTTPException(status_code=502, detail=f"TTS 合成失败: {e}") from e
+
+    audio_url = result.audio_url
+
+    # 缓存（仅默认试听文本）
+    if use_cache:
+        _preset_preview_cache[voice_id] = (audio_url, time.time())
+
+    return {
+        "voice_id": voice_id,
+        "audio_url": audio_url,
+        "text": preview_text,
+        "cached": False,
+    }
+
+
 # ==================== 原有 CRUD 端点（保持向后兼容）====================