tests/test_calculator_fix.py

"""
Calculator Accuracy Fix - TDD Approach
Identifies and fixes calculator accuracy issues to achieve 100% success rate.
"""

import pytest
import sys
import os
import logging
from pathlib import Path

# Add the deployment-ready directory to the path
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..'))

from agents.fixed_enhanced_unified_agno_agent import FixedGAIAAgent

logger = logging.getLogger(__name__)


class TestCalculatorFix:
    """Test suite to identify and fix calculator accuracy issues."""
    
    @pytest.fixture(autouse=True)
    def setup_method(self):
        """Set up test fixtures."""
        self.agent = FixedGAIAAgent()
        
    def test_basic_arithmetic_operations(self):
        """Test basic arithmetic operations that should always work."""
        test_cases = [
            {
                'question': 'What is 25 * 17?',
                'expected': '425',
                'operation': 'multiplication'
            },
            {
                'question': 'What is 144 / 12?',
                'expected': '12',
                'operation': 'division'
            },
            {
                'question': 'What is 100 + 50?',
                'expected': '150',
                'operation': 'addition'
            },
            {
                'question': 'What is 200 - 75?',
                'expected': '125',
                'operation': 'subtraction'
            }
        ]
        
        failed_operations = []
        
        for case in test_cases:
            if not self.agent.available:
                pytest.skip("Agent not available for testing")
            
            try:
                result = self.agent(case['question'])
                
                # Clean the result for comparison
                cleaned_result = result.strip().replace(',', '')
                expected = case['expected']
                
                # Check if the result matches
                if cleaned_result != expected:
                    failed_operations.append({
                        'question': case['question'],
                        'expected': expected,
                        'actual': cleaned_result,
                        'operation': case['operation']
                    })
                    logger.error(f"❌ {case['operation']} failed: {case['question']} → Expected: {expected}, Got: {cleaned_result}")
                else:
                    logger.info(f"✅ {case['operation']} passed: {case['question']} → {cleaned_result}")
                    
            except Exception as e:
                failed_operations.append({
                    'question': case['question'],
                    'expected': case['expected'],
                    'actual': f"ERROR: {e}",
                    'operation': case['operation']
                })
                logger.error(f"❌ {case['operation']} error: {case['question']} → {e}")
        
        # Report results
        if failed_operations:
            logger.error(f"❌ Calculator accuracy: {len(test_cases) - len(failed_operations)}/{len(test_cases)} ({((len(test_cases) - len(failed_operations))/len(test_cases)*100):.1f}%)")
            for failure in failed_operations:
                logger.error(f"   Failed: {failure['question']} → Expected: {failure['expected']}, Got: {failure['actual']}")
        else:
            logger.info(f"✅ Calculator accuracy: 100% ({len(test_cases)}/{len(test_cases)})")
        
        # Assert no failures for 100% accuracy
        assert len(failed_operations) == 0, f"Calculator failed {len(failed_operations)} out of {len(test_cases)} tests"
    
    def test_complex_mathematical_operations(self):
        """Test complex mathematical operations."""
        test_cases = [
            {
                'question': 'What is 2^8?',
                'expected': '256',
                'operation': 'exponentiation'
            },
            {
                'question': 'What is the square root of 144?',
                'expected': '12',
                'operation': 'square_root'
            },
            {
                'question': 'Calculate the factorial of 5',
                'expected': '120',
                'operation': 'factorial'
            }
        ]
        
        failed_operations = []
        
        for case in test_cases:
            if not self.agent.available:
                pytest.skip("Agent not available for testing")
            
            try:
                result = self.agent(case['question'])
                
                # Clean the result for comparison
                cleaned_result = result.strip().replace(',', '')
                expected = case['expected']
                
                # For complex operations, allow for slight variations
                try:
                    result_num = float(cleaned_result)
                    expected_num = float(expected)
                    if abs(result_num - expected_num) < 0.01:
                        logger.info(f"✅ {case['operation']} passed: {case['question']} → {cleaned_result}")
                        continue
                except ValueError:
                    pass
                
                # Exact match check
                if cleaned_result != expected:
                    failed_operations.append({
                        'question': case['question'],
                        'expected': expected,
                        'actual': cleaned_result,
                        'operation': case['operation']
                    })
                    logger.error(f"❌ {case['operation']} failed: {case['question']} → Expected: {expected}, Got: {cleaned_result}")
                else:
                    logger.info(f"✅ {case['operation']} passed: {case['question']} → {cleaned_result}")
                    
            except Exception as e:
                failed_operations.append({
                    'question': case['question'],
                    'expected': case['expected'],
                    'actual': f"ERROR: {e}",
                    'operation': case['operation']
                })
                logger.error(f"❌ {case['operation']} error: {case['question']} → {e}")
        
        # Report results
        success_rate = (len(test_cases) - len(failed_operations)) / len(test_cases) * 100
        logger.info(f"📊 Complex math accuracy: {success_rate:.1f}% ({len(test_cases) - len(failed_operations)}/{len(test_cases)})")
        
        if failed_operations:
            for failure in failed_operations:
                logger.error(f"   Failed: {failure['question']} → Expected: {failure['expected']}, Got: {failure['actual']}")
    
    def test_calculator_tool_direct_access(self):
        """Test direct access to calculator tool to identify issues."""
        if not self.agent.available:
            pytest.skip("Agent not available for testing")
        
        # Find calculator tool
        calculator_tool = None
        for tool in self.agent.tools:
            if hasattr(tool, '__class__') and 'Calculator' in tool.__class__.__name__:
                calculator_tool = tool
                break
        
        if calculator_tool is None:
            pytest.fail("Calculator tool not found in agent tools")
        
        logger.info(f"✅ Calculator tool found: {calculator_tool.__class__.__name__}")
        
        # Test direct calculator operations
        test_operations = [
            ('25 * 17', 425),
            ('144 / 12', 12),
            ('2 ** 8', 256),
            ('100 + 50', 150)
        ]
        
        for expression, expected in test_operations:
            try:
                # This would depend on the calculator tool's interface
                logger.info(f"🧮 Testing calculator: {expression} = {expected}")
            except Exception as e:
                logger.error(f"❌ Calculator tool error: {e}")


if __name__ == "__main__":
    # Run the calculator fix tests
    pytest.main([__file__, "-v", "-s"])