speed optim

backend/app/services/llm_cache.py

@@ -0,0 +1,24 @@
+import os
+import threading
+from typing import Any, Dict, List, Optional
+
+from app.api.llm import DATA_FILE, _load_data
+
+_cache_lock = threading.RLock()
+_cache_mtime: float = -1.0
+_cache_data: List[Dict[str, Any]] = []
+
+
+def get_llm_configs() -> List[Dict[str, Any]]:
+    global _cache_mtime, _cache_data
+    current_mtime = os.path.getmtime(DATA_FILE) if os.path.exists(DATA_FILE) else -1.0
+    with _cache_lock:
+        if current_mtime != _cache_mtime:
+            _cache_data = _load_data()
+            _cache_mtime = current_mtime
+        return list(_cache_data)
+
+
+def get_active_llm_config() -> Optional[Dict[str, Any]]:
+    configs = get_llm_configs()
+    return next((c for c in configs if c.get("is_active")), None)