def process_answers()

in extract_answers.py [0:0]

• 34 lines of code
• 8 McCabe index (conditional complexity)

def process_answers(df: pd.DataFrame) -> pd.DataFrame:
    """Process each answer through the sympy extraction workflow."""
    results = []
    
    # Set up extraction config and get regexes
    extraction_target = (ExprExtractionConfig(), LatexExtractionConfig())
    for _, row in df.iterrows():
        try:
            # Extract answer using regexes
            extracted = parse(row['answer'], extraction_config=extraction_target)
            feedback = None
            extracted_answer = None
            if len(extracted) == 2:
                extracted_answer = extracted[0]
                feedback = extracted[1]
            elif len(extracted) == 1:
                extracted_answer = extracted[0]
            else:
                feedback = "No valid extraction found"
                extracted_answer = None


            
            result = {
                'original_answer': row['answer'],
                'extracted_answer': serialize_sympy_object(extracted_answer),
                'extracted_feedback': feedback,
                'extraction_success': extracted_answer is not None
            }
            
            # Copy any other columns from input
            for col in df.columns:
                if col != 'answer':
                    result[col] = row[col]
                    
            results.append(result)
            
        except Exception as e:
            results.append({
                'original_answer': row['answer'],
                'extracted_answer': '',
                'extraction_success': False,
                'error': str(e)
            })
    
    return pd.DataFrame(results)