Update main.py

main.py

@@ -301,63 +301,65 @@ def get_spending_overview():
         
         # Get transactions for the user.
         transactions_ref = db.reference('transactions')
-        transactions = transactions_ref.order_by_child('uid').equal_to(uid).get()
-
-        if transactions is None:
-            return jsonify({
-                'daily_spending': [],
-                'recent_transactions': []
-            })
+        transactions = transactions_ref.order_by_child('uid').equal_to(uid).get() or {}
         
-        # Create a DataFrame from the transactions.
-        df = pd.DataFrame.from_dict(transactions, orient='index')
-        if df.empty:
+        # Process transactions directly (similar to the admin overview approach)
+        transactions_list = [{'id': tid, **data} for tid, data in transactions.items()]
+        
+        if not transactions_list:
             return jsonify({
                 'daily_spending': [],
                 'recent_transactions': []
             })
         
-        # Drop rows missing essential fields.
-        df.dropna(subset=['uid', 'total', 'items', 'date', 'receipt_number', 'timestamp', 'hash'], inplace=True)
+        # Create a DataFrame
+        df = pd.DataFrame(transactions_list)
         
-        # IMPORTANT FIX: Removed 'manual_entry' from the essential fields list
-        # This was causing manual_entry transactions to be filtered out if a user's first transaction was manual
+        # Handle date formatting (only if the date column exists)
+        if 'date' in df.columns:
+            # Convert to datetime, but keep non-convertible values
+            df['date'] = pd.to_datetime(df['date'], errors='coerce')
+            # Filter out rows with invalid dates
+            df = df.dropna(subset=['date'])
+            
+            if not df.empty:
+                # Convert datetime to ISO format
+                df['date'] = df['date'].apply(lambda d: d.isoformat())
+                # Extract date part for grouping
+                df['date_only'] = df['date'].apply(lambda d: d.split("T")[0])
         
+        # Return empty response if we don't have valid data after processing
         if df.empty:
             return jsonify({
                 'daily_spending': [],
                 'recent_transactions': []
             })
         
-        # Parse the date column without a fixed format and drop invalid dates.
-        df['date'] = pd.to_datetime(df['date'], errors='coerce')
-        df.dropna(subset=['date'], inplace=True)
-        if df.empty:
-            return jsonify({
-                'daily_spending': [],
-                'recent_transactions': []
-            })
+        # For daily spending, group by date
+        if 'date_only' in df.columns and 'total' in df.columns:
+            daily_spending = df.groupby('date_only')['total'].sum().reset_index()
+            daily_spending.rename(columns={'date_only': 'date'}, inplace=True)
+        else:
+            daily_spending = pd.DataFrame(columns=['date', 'total'])
         
-        # Convert the date column to ISO formatted strings (e.g. "2025-01-02T00:00:00").
-        df['date'] = df['date'].apply(lambda d: d.isoformat())
+        # Sort the transactions by timestamp if available
+        if 'timestamp' in df.columns:
+            recent_transactions = df.sort_values(by='timestamp', ascending=False).head(10)
+        else:
+            recent_transactions = df.head(10)
         
-        # For daily spending, group by just the date part (YYYY-MM-DD).
-        # Extract the date part from the ISO string.
-        df['date_only'] = df['date'].apply(lambda d: d.split("T")[0])
-        daily_spending = df.groupby('date_only')['total'].sum().reset_index()
-        # Rename the column to "date" and ensure it's a string.
-        daily_spending.rename(columns={'date_only': 'date'}, inplace=True)
-        daily_spending['date'] = daily_spending['date'].astype(str)
+        # Replace NaN values with None for JSON serialization
         daily_spending = daily_spending.replace({np.nan: None})
-        
-        # Sort the transactions by timestamp (assuming it's in a sortable string format).
-        recent_transactions = df.sort_values(by='timestamp', ascending=False).head(10)
         recent_transactions = recent_transactions.replace({np.nan: None})
         
-        return jsonify({
+        # Prepare response
+        response_data = {
             'daily_spending': daily_spending.to_dict(orient='records'),
-            'recent_transactions': recent_transactions.drop(columns=['date_only']).to_dict(orient='records')
-        })
+            'recent_transactions': recent_transactions.drop(columns=['date_only'], errors='ignore').to_dict(orient='records')
+        }
+        
+        return jsonify(response_data)
+        
     except Exception as e:
         return jsonify({'error': str(e)}), 500