Update app.py

app.py

@@ -2,7 +2,6 @@ from fastapi import FastAPI, Query
 from fastapi.middleware.cors import CORSMiddleware
 from datasets import load_dataset
 import os
-import time
 
 os.environ["HF_HOME"] = "/tmp/hf"
 
@@ -26,24 +25,50 @@ def root():
     }
 
 @app.get("/search")
-def search_uid(uid: str = Query(..., description="Telegram UID")):
-    """Poore dataset mein UID dhoondho (25 second timeout ke saath)"""
+def search_uid(
+    uid: str = Query(..., description="Telegram UID"),
+    offset: int = Query(0, ge=0, description="Kahan se start karna hai (default 0)")
+):
+    """
+    Poore dataset mein UID dhoondho.
+    - offset=0 se start karo
+    - Agar nahi mila toh next_offset milega
+    - Woh next_offset daal ke dobara call karo
+    - Aise karte karte poora dataset cover hoga
+    """
     try:
         ds = load_dataset(DATASET, split="train", streaming=True)
         
-        start_time = time.time()
+        # Offset tak skip karo
+        iterator = iter(ds)
+        skipped = 0
+        for _ in range(offset):
+            try:
+                next(iterator)
+                skipped += 1
+            except StopIteration:
+                return {
+                    "found": False,
+                    "uid": uid,
+                    "offset": offset,
+                    "scanned": 0,
+                    "message": "End of dataset",
+                    "next_offset": None
+                }
+        
+        # Ab 5 lakh rows scan karo (HF Space 30 second timeout ke liye safe)
         scanned = 0
-        max_seconds = 25  # HF Space timeout se pehle ruk jao
+        max_scan = 500000
         
-        for row in ds:
+        for row in iterator:
             scanned += 1
-            
-            # UID match karo
             if uid == str(row.get("user_id", "")).strip():
                 return {
                     "found": True,
                     "uid": uid,
+                    "offset": offset,
                     "scanned": scanned,
+                    "total_scanned": offset + scanned,
                     "data": {
                         "user_id": row.get("user_id"),
                         "username": row.get("username"),
@@ -58,23 +83,16 @@ def search_uid(uid: str = Query(..., description="Telegram UID")):
                     }
                 }
             
-            # Har 5000 rows ke baad time check
-            if scanned % 5000 == 0:
-                if time.time() - start_time > max_seconds:
-                    return {
-                        "found": False,
-                        "uid": uid,
-                        "scanned": scanned,
-                        "time_taken": round(time.time() - start_time, 1),
-                        "message": f"{scanned} rows scan kiye, UID nahi mila. Phir try karo."
-                    }
+            if scanned >= max_scan:
+                break
         
         return {
             "found": False,
             "uid": uid,
+            "offset": offset,
             "scanned": scanned,
-            "time_taken": round(time.time() - start_time, 1),
-            "message": "Poore dataset mein nahi mila"
+            "next_offset": offset + scanned,
+            "message": f"Is chunk mein nahi mila. Ab offset={offset + scanned} try karo."
         }
     
     except Exception as e: