Add comprehensive column parsing test suite and unified analyzer

- Add test_column_parsing.sql with 20+ test scenarios covering basic columns,
  schema-qualified columns, alias chains, complex expressions, nested struct
  fields, and various SQL contexts (WHERE, GROUP BY, ORDER BY, etc.)
- Add test_column_parsing_core.sql for essential functionality verification
- Add unified_analyzer_v2.sql integrating column analysis with existing
  function and table parsing using correct DuckDB table function syntax
- Unified analyzer now provides complete SQL dependency analysis across
  functions, tables, and columns with existence checking and suggestions

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>

Author: teaguesterling

test_column_parsing.sql

@@ -0,0 +1,164 @@
+-- Test suite for column parsing functionality
+-- Load the extension first
+LOAD parser_tools;
+
+-- Create test tables
+CREATE TABLE users (id INT, age INT, name VARCHAR, email VARCHAR);
+CREATE TABLE orders (id INT, user_id INT, total DECIMAL, status VARCHAR);
+CREATE TABLE profiles (user_id INT, first_name VARCHAR, last_name VARCHAR, address STRUCT(street VARCHAR, city VARCHAR, zip VARCHAR));
+
+-- Test 1: Basic column parsing
+SELECT '=== Test 1: Basic column parsing ===' as test_section;
+SELECT * FROM parse_columns('SELECT name, age FROM users');
+
+-- Test 2: Schema-qualified columns
+SELECT '=== Test 2: Schema-qualified columns ===' as test_section;
+SELECT * FROM parse_columns('SELECT main.users.name, main.users.age FROM main.users');
+
+-- Test 3: Complex expressions with multiple columns
+SELECT '=== Test 3: Complex expressions ===' as test_section;
+SELECT * FROM parse_columns('SELECT u.name, o.total, u.age + o.total AS summary FROM users u JOIN orders o ON u.id = o.user_id');
+
+-- Test 4: Alias chain scenario (from our discussion)
+SELECT '=== Test 4: Alias chains ===' as test_section;
+SELECT * FROM parse_columns('SELECT 1 AS a, users.age AS b, a+b AS c, b AS d FROM users');
+
+-- Test 5: Nested struct field access
+SELECT '=== Test 5: Nested struct fields ===' as test_section;
+SELECT * FROM parse_columns('SELECT profiles.address.street, profiles.address.city FROM profiles');
+
+-- Test 6: Deeply nested struct with schema qualification
+SELECT '=== Test 6: Deep nested struct with schema ===' as test_section;
+SELECT * FROM parse_columns('SELECT main.profiles.address.city FROM main.profiles');
+
+-- Test 7: WHERE clause columns
+SELECT '=== Test 7: WHERE clause columns ===' as test_section;
+SELECT * FROM parse_columns('SELECT name FROM users WHERE age > 18 AND email LIKE ''%@gmail.com''');
+
+-- Test 8: GROUP BY and HAVING columns
+SELECT '=== Test 8: GROUP BY and HAVING columns ===' as test_section;
+SELECT * FROM parse_columns('SELECT status, COUNT(*) FROM orders GROUP BY status HAVING COUNT(*) > 5');
+
+-- Test 9: ORDER BY columns
+SELECT '=== Test 9: ORDER BY columns ===' as test_section;
+SELECT * FROM parse_columns('SELECT name, age FROM users ORDER BY age DESC, name ASC');
+
+-- Test 10: Function arguments with columns
+SELECT '=== Test 10: Function arguments ===' as test_section;
+SELECT * FROM parse_columns('SELECT UPPER(name), LENGTH(email), CONCAT(first_name, '' '', last_name) FROM users');
+
+-- Test 11: Window functions
+SELECT '=== Test 11: Window functions ===' as test_section;
+SELECT * FROM parse_columns('SELECT name, ROW_NUMBER() OVER (PARTITION BY age ORDER BY name) FROM users');
+
+-- Test 12: Complex query with joins, subqueries, and functions
+SELECT '=== Test 12: Complex query ===' as test_section;
+SELECT * FROM parse_columns('
+    WITH user_stats AS (
+        SELECT u.id, u.name, COUNT(o.id) as order_count
+        FROM users u
+        LEFT JOIN orders o ON u.id = o.user_id
+        GROUP BY u.id, u.name
+    )
+    SELECT 
+        us.name,
+        us.order_count,
+        CASE 
+            WHEN us.order_count > 5 THEN ''high''
+            WHEN us.order_count > 1 THEN ''medium''
+            ELSE ''low''
+        END as activity_level
+    FROM user_stats us
+    WHERE us.order_count > 0
+    ORDER BY us.order_count DESC
+');
+
+-- Test 13: Unqualified columns (aliases, literals)
+SELECT '=== Test 13: Unqualified columns and literals ===' as test_section;
+SELECT * FROM parse_columns('SELECT 42 AS answer, ''hello'' AS greeting, name FROM users');
+
+-- Test 14: Mixed qualified and unqualified references
+SELECT '=== Test 14: Mixed qualifications ===' as test_section;
+SELECT * FROM parse_columns('SELECT users.name, age, profiles.first_name FROM users JOIN profiles ON users.id = profiles.user_id');
+
+-- Test 15: CASE expressions with columns
+SELECT '=== Test 15: CASE expressions ===' as test_section;
+SELECT * FROM parse_columns('
+    SELECT 
+        name,
+        CASE 
+            WHEN age < 18 THEN ''minor''
+            WHEN age < 65 THEN ''adult''
+            ELSE ''senior''
+        END as age_group
+    FROM users
+');
+
+-- Test 16: Subquery column references
+SELECT '=== Test 16: Subquery columns ===' as test_section;
+SELECT * FROM parse_columns('
+    SELECT name, age 
+    FROM users 
+    WHERE id IN (SELECT user_id FROM orders WHERE total > 100)
+');
+
+-- Test 17: JOIN conditions
+SELECT '=== Test 17: JOIN conditions ===' as test_section;
+SELECT * FROM parse_columns('
+    SELECT u.name, o.total
+    FROM users u
+    INNER JOIN orders o ON u.id = o.user_id AND u.age > 18
+');
+
+-- Test 18: Multiple table aliases with same column names
+SELECT '=== Test 18: Multiple aliases, same column names ===' as test_section;
+SELECT * FROM parse_columns('
+    SELECT u.id as user_id, o.id as order_id, u.name, o.status
+    FROM users u
+    JOIN orders o ON u.id = o.user_id
+');
+
+-- Test 19: Column references in aggregates
+SELECT '=== Test 19: Aggregates with columns ===' as test_section;
+SELECT * FROM parse_columns('
+    SELECT 
+        COUNT(DISTINCT u.id) as unique_users,
+        AVG(o.total) as avg_order,
+        SUM(o.total) as total_revenue
+    FROM users u
+    JOIN orders o ON u.id = o.user_id
+');
+
+-- Test 20: Column with arithmetic operations
+SELECT '=== Test 20: Arithmetic operations ===' as test_section;
+SELECT * FROM parse_columns('SELECT age * 2 + 10 AS calculated_age, total / quantity AS unit_price FROM users JOIN orders ON users.id = orders.user_id');
+
+-- Summary report: Show unique contexts found
+SELECT '=== Summary: Column contexts found ===' as summary_section;
+SELECT DISTINCT context, COUNT(*) as count 
+FROM (
+    SELECT * FROM parse_columns('SELECT u.name, o.total FROM users u JOIN orders o ON u.id = o.user_id WHERE u.age > 18 ORDER BY o.total DESC')
+) 
+GROUP BY context 
+ORDER BY count DESC;
+
+-- Summary report: Show expression identifier patterns
+SELECT '=== Summary: Expression identifier patterns ===' as summary_section;
+SELECT 
+    CASE 
+        WHEN expression_identifiers LIKE '%[%[%,%' THEN 'Multiple identifiers'
+        WHEN expression_identifiers LIKE '%"%,"%,"%' THEN 'Three-part qualified'
+        WHEN expression_identifiers LIKE '%"%,"%' THEN 'Two-part qualified'
+        ELSE 'Single identifier'
+    END as pattern_type,
+    COUNT(*) as count
+FROM (
+    SELECT * FROM parse_columns('SELECT main.users.name, users.age, name, 1 AS const FROM main.users')
+)
+GROUP BY pattern_type
+ORDER BY count DESC;
+
+-- Cleanup
+DROP TABLE IF EXISTS users;
+DROP TABLE IF EXISTS orders;
+DROP TABLE IF EXISTS profiles;

test_column_parsing_core.sql

@@ -0,0 +1,39 @@
+-- Core column parsing tests
+-- This file tests the essential functionality without dependencies on complex table structures
+
+-- Test 1: Basic unqualified columns
+SELECT 'Test 1: Basic columns' as test_name;
+SELECT * FROM parse_columns('SELECT name, age FROM users');
+
+-- Test 2: Schema-qualified columns  
+SELECT 'Test 2: Schema-qualified' as test_name;
+SELECT * FROM parse_columns('SELECT main.users.name FROM main.users');
+
+-- Test 3: Alias chain (our key scenario)
+SELECT 'Test 3: Alias chains' as test_name;
+SELECT * FROM parse_columns('SELECT 1 AS a, users.age AS b, a+b AS c, b AS d FROM users');
+
+-- Test 4: Complex expression with multiple identifiers
+SELECT 'Test 4: Complex expressions' as test_name;
+SELECT * FROM parse_columns('SELECT u.name, o.total, u.age + o.total AS summary FROM users u JOIN orders o ON u.id = o.user_id');
+
+-- Test 5: WHERE clause columns
+SELECT 'Test 5: WHERE clause' as test_name;
+SELECT * FROM parse_columns('SELECT name FROM users WHERE age > 18 AND email LIKE ''test''');
+
+-- Test 6: Function arguments
+SELECT 'Test 6: Function arguments' as test_name;
+SELECT * FROM parse_columns('SELECT UPPER(name), CONCAT(first_name, last_name) FROM users');
+
+-- Test 7: Nested struct field (simulated)
+SELECT 'Test 7: Nested struct' as test_name;
+SELECT * FROM parse_columns('SELECT users.profile.address.city FROM users');
+
+-- Test 8: Output validation - check NULL handling
+SELECT 'Test 8: NULL handling verification' as test_name;
+SELECT 
+    CASE WHEN table_schema IS NULL THEN 'NULL' ELSE table_schema END as schema_check,
+    CASE WHEN table_name IS NULL THEN 'NULL' ELSE table_name END as table_check,
+    CASE WHEN selected_name IS NULL THEN 'NULL' ELSE selected_name END as selected_check
+FROM parse_columns('SELECT 1 AS a, users.age AS b FROM users')
+LIMIT 3;