Update app.py

app.py

@@ -6,6 +6,7 @@ from filtered_search_engine import SmartRecommender
 from reranker import Reranker
 from intent_classifier import IntentClassifier
 from keyword_boosting_layer import apply_keyword_boost
+from cache_manager import CacheManager
 
 # ------------------------------
 # Initialize App
@@ -27,14 +28,22 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
-# Mount images folder to serve static files
-app.mount("/images", StaticFiles(directory="images"), name="images")
+# Mount images folder to serve static files with Cache-Control
+class CachedStaticFiles(StaticFiles):
+    def file_response(self, *args, **kwargs):
+        response = super().file_response(*args, **kwargs)
+        response.headers["Cache-Control"] = "public, max-age=31536000"
+        return response
+
+app.mount("/images", CachedStaticFiles(directory="images"), name="images")
 
 # ------------------------------
 # Load Core Components Once
 # ------------------------------
 print("📌 Loading dataset...")
 df = pd.read_csv("salahkar_enhanced.csv")
+# Optimization: Create O(1) lookup map
+name_to_row = df.set_index("name").to_dict('index')
 
 print("📌 Loading smart recommendation engine...")
 engine = SmartRecommender()
@@ -45,6 +54,9 @@ reranker = Reranker()
 print("📌 Loading intent recognizer...")
 intent_detector = IntentClassifier()
 
+print("📌 Initializing Cache Manager...")
+cache = CacheManager(capacity=200, ttl_seconds=3600)
+
 print("🚀 Salahkar AI Ready!")
 
 
@@ -65,13 +77,19 @@ def get_recommendation(query: str = Query(..., description="User's search text")
 
     print(f"\n🔍 User Query: {query}")
 
+    # 0️⃣ Check Cache
+    cached_response = cache.get(query)
+    if cached_response:
+        return cached_response
+
     # 1️⃣ Detect intent
     detected_intent = intent_detector.predict_intent(query)
     print(f"🧠 Intent Detected: {detected_intent}")
 
     # 2️⃣ FAISS + Filter Search
     # engine.recommend returns (results_list, intent)
-    rec_results, _ = engine.recommend(query, k=k)
+    # Optimization: Pass detected_intent to avoid re-running classification
+    rec_results, _ = engine.recommend(query, k=k, intent=detected_intent)
 
     # 3️⃣ Prepare results for reranker
     prepared = []
@@ -81,7 +99,12 @@ def get_recommendation(query: str = Query(..., description="User's search text")
         category = item["category"]
         region = item["region"]
         score = item["score"]
-        row = df[df["name"] == name].iloc[0]
+        
+        # Optimization: O(1) lookup instead of O(N) dataframe filter
+        row = name_to_row.get(name)
+        if not row:
+            continue
+            
         prepared.append({
             "name": name,
             "domain": domain,
@@ -111,12 +134,17 @@ def get_recommendation(query: str = Query(..., description="User's search text")
         for item in final_results[:k]
     ]
 
-    return {
+    final_response = {
         "query": query,
         "intent": detected_intent,
         "results": response
     }
 
+    # 7️⃣ Save to Cache
+    cache.set(query, final_response)
+
+    return final_response
+
 
 # -------------------------------------------
 # Run (Ignored by HuggingFace — needed only for local testing)