Update app.py

app.py

@@ -21,13 +21,74 @@ IMPORTANT REQUIREMENTS:
 - Always filter by user_id = $1 for security
 - The current year is 2025. You are working in this year!
 - CRITICAL: Use ONLY the dates provided in the input parameters. Do NOT infer or change dates on your own!
-- Single date provided: {date_info if date_info else '[none provided]'}
 - If date range is provided, use DATE(created_at) BETWEEN 'startDate' AND 'endDate'
 - If single date is provided, use DATE(created_at) = 'YYYY-MM-DD'
 - NEVER use hardcoded years like 2024 - always use the provided dates exactly as given
-- Generate ONLY simple SELECT, INSERT, UPDATE, DELETE statements.
-- NO WITH clauses, NO CTEs, NO complex subqueries.
-- Use RETURNING * for INSERT, UPDATE, DELETE when applicable.
+- Use proper SQL syntax
+- CRITICAL: Generate ONLY simple SQL statements - NO WITH clauses, NO CTEs, NO complex subqueries
+- Use direct SELECT, INSERT, UPDATE, DELETE statements only
+- Keep queries simple and straightforward
+- For INSERT statements, use the RETURNING clause
+- For INSERT statements with specific date:
+  * Use INSERT INTO expenses (user_id, amount, category, note, created_at) VALUES ($1, amount_value, 'category_name', 'note_text', 'YYYY-MM-DD HH:MM:SS') RETURNING *
+  * If date is provided, use that specific date for created_at instead of NOW()
+  * Format the date as 'YYYY-MM-DD 00:00:00' for the specific date
+- For SELECT statements:
+  * Use SIMPLE SELECT statements - NO WITH clauses, NO CTEs, NO complex subqueries
+  * If category is provided: SELECT * FROM expenses WHERE user_id = $1 AND category = 'exact_category_name'
+  * If keywords are provided: SELECT * FROM expenses WHERE user_id = $1 AND LOWER(note) LIKE '%exact_keyword%'
+  * If single date is provided: SELECT * FROM expenses WHERE user_id = $1 AND DATE(created_at) = 'YYYY-MM-DD'
+  * If date range is provided: SELECT * FROM expenses WHERE user_id = $1 AND DATE(created_at) BETWEEN 'startDate' AND 'endDate'
+  * If both category and keywords: SELECT * FROM expenses WHERE user_id = $1 AND category = 'exact_category_name' AND LOWER(note) LIKE '%exact_keyword%'
+  * If both category and single date: SELECT * FROM expenses WHERE user_id = $1 AND category = 'exact_category_name' AND DATE(created_at) = 'YYYY-MM-DD'
+  * If both category and date range: SELECT * FROM expenses WHERE user_id = $1 AND category = 'exact_category_name' AND DATE(created_at) BETWEEN 'startDate' AND 'endDate'
+  * If both keywords and single date: SELECT * FROM expenses WHERE user_id = $1 AND LOWER(note) LIKE '%exact_keyword%' AND DATE(created_at) = 'YYYY-MM-DD'
+  * If both keywords and date range: SELECT * FROM expenses WHERE user_id = $1 AND LOWER(note) LIKE '%exact_keyword%' AND DATE(created_at) BETWEEN 'startDate' AND 'endDate'
+  * If all three (category, keywords, date): SELECT * FROM expenses WHERE user_id = $1 AND category = 'exact_category_name' AND LOWER(note) LIKE '%exact_keyword%' AND DATE(created_at) = 'YYYY-MM-DD'
+  * If all three (category, keywords, date range): SELECT * FROM expenses WHERE user_id = $1 AND category = 'exact_category_name' AND LOWER(note) LIKE '%exact_keyword%' AND DATE(created_at) BETWEEN 'startDate' AND 'endDate'
+  * If none provided: SELECT * FROM expenses WHERE user_id = $1
+  * Return all relevant columns
+  * CRITICAL: Use ONLY simple SELECT statements, NO WITH clauses, NO CTEs
+  * CRITICAL: For date filtering, use DATE(created_at) = 'YYYY-MM-DD' for single dates, DATE(created_at) BETWEEN 'startDate' AND 'endDate' for date ranges
+  * CRITICAL: Use the EXACT dates provided in the input parameters. Do NOT infer or override dates from the user query wording.
+  * EXAMPLE: If date range is '2025-01-13' to '2025-01-19', use DATE(created_at) BETWEEN '2025-01-13' AND '2025-01-19'
+- For UPDATE statements: 
+  * Use UPDATE expenses SET amount = new_amount WHERE user_id = $1 AND LOWER(note) LIKE '%keyword%'
+  * If category is provided, also add AND category = 'category_name'
+  * Use RETURNING clause to return the updated record
+- For DELETE statements:
+  * If category is provided: DELETE FROM expenses WHERE user_id = $1 AND category = 'exact_category_name'
+  * If keywords are provided: DELETE FROM expenses WHERE user_id = $1 AND LOWER(note) LIKE '%exact_keyword%'
+  * If both category and keywords: DELETE FROM expenses WHERE user_id = $1 AND category = 'exact_category_name' AND LOWER(note) LIKE '%exact_keyword%'
+  * If no category and no keywords (general delete): DELETE FROM expenses WHERE user_id = $1
+  * Use RETURNING clause to return the deleted records
+- Handle the user_id parameter safely
+- If the query mentions a specific month (like "June"), filter by that month using EXTRACT(MONTH FROM created_at)
+- If the query mentions a specific year, filter by that year using EXTRACT(YEAR FROM created_at)
+- CRITICAL: If a specific date is provided (like "26th june 2025"), use exact date filtering: DATE(created_at) = 'YYYY-MM-DD'
+- For date filtering, use proper PostgreSQL date functions
+- Use EXACT category names from the list above
+- For keyword searches in UPDATE/DELETE, use LOWER(note) LIKE '%exact_keyword%' pattern
+- For category searches, use category = 'exact_category_name'
+- When both category and keywords are provided for SELECT, prioritize category filtering
+- CRITICAL: If the intent is UPDATE, generate an UPDATE query, NOT a SELECT query
+- CRITICAL: If the intent is DELETE, generate a DELETE query, NOT a SELECT query
+- NEVER use placeholder text like '%keyword%' or 'category_name' - use the actual values provided
+
+CORRECT SQL EXAMPLES:
+- SELECT * FROM expenses WHERE user_id = $1 AND DATE(created_at) = '2025-06-26';
+- SELECT SUM(amount) FROM expenses WHERE user_id = $1 AND DATE(created_at) = '2025-06-26';
+- SELECT * FROM expenses WHERE user_id = $1 AND DATE(created_at) BETWEEN '2025-01-13' AND '2025-01-19';
+- SELECT SUM(amount) FROM expenses WHERE user_id = $1 AND DATE(created_at) BETWEEN '2025-01-13' AND '2025-01-19';
+- SELECT * FROM expenses WHERE user_id = $1 AND category = 'Food & Dining';
+- SELECT * FROM expenses WHERE user_id = $1 AND LOWER(note) LIKE '%lunch%';
+- INSERT INTO expenses (user_id, amount, category, note, created_at) VALUES ($1, 500.00, 'Health & Fitness', 'Spent 500 rupees yesterday on medicines', '2025-01-19 00:00:00') RETURNING *;
+- EXAMPLE: If date range is '2025-01-13' to '2025-01-19', use DATE(created_at) BETWEEN '2025-01-13' AND '2025-01-19'
+
+INCORRECT SQL EXAMPLES (DO NOT USE):
+- WITH filtered_date AS (SELECT '2025-06-24' AS target_date) SELECT * FROM expenses WHERE user_id = $1 AND DATE(created_at) = (SELECT target_date FROM filtered_date);
+- WITH filtered_expenses AS (SELECT * FROM expenses WHERE user_id = $1) SELECT * FROM filtered_expenses;
+
 
 SQL Query:"""