AI测试评估体系

def calculate_test_case_quality(test_cases):
    """
    计算测试用例质量评分
    
    Args:
        test_cases: 测试用例列表
    
    Returns:
        float: 质量评分 (0-100)
    """
    total_score = 0
    total_weight = 0
    
    # 有效性得分 (权重: 40%)
    effective_cases = [tc for tc in test_cases if tc.has_found_defects]
    effectiveness = len(effective_cases) / len(test_cases) if test_cases else 0
    total_score += effectiveness * 100 * 0.4
    total_weight += 0.4
    
    # 稳定性得分 (权重: 30%)
    stable_cases = [tc for tc in test_cases if tc.stability_rate > 0.95]
    stability = len(stable_cases) / len(test_cases) if test_cases else 0
    total_score += stability * 100 * 0.3
    total_weight += 0.3
    
    # 可维护性得分 (权重: 20%)
    maintainable_cases = [tc for tc in test_cases if tc.modification_count < 3]
    maintainability = len(maintainable_cases) / len(test_cases) if test_cases else 0
    total_score += maintainability * 100 * 0.2
    total_weight += 0.2
    
    # 复用性得分 (权重: 10%)
    reusable_cases = [tc for tc in test_cases if tc.reuse_count > 1]
    reusability = len(reusable_cases) / len(test_cases) if test_cases else 0
    total_score += reusability * 100 * 0.1
    total_weight += 0.1
    
    return total_score / total_weight if total_weight > 0 else 0

def compare_defect_detection(ai_defects, manual_defects):
    """
    对比AI和人工的缺陷发现能力
    
    Args:
        ai_defects: AI发现的缺陷列表
        manual_defects: 人工发现的缺陷列表
    
    Returns:
        dict: 对比结果
    """
    ai_set = set(d.id for d in ai_defects)
    manual_set = set(d.id for d in manual_defects)
    
    # 独有缺陷
    ai_unique = ai_set - manual_set
    manual_unique = manual_set - ai_set
    
    # 共同缺陷
    common = ai_set & manual_set
    
    # 发现率对比
    total_defects = len(ai_set | manual_set)
    ai_rate = len(ai_set) / total_defects if total_defects > 0 else 0
    manual_rate = len(manual_set) / total_defects if total_defects > 0 else 0
    
    return {
        "ai_unique_count": len(ai_unique),
        "manual_unique_count": len(manual_unique),
        "common_count": len(common),
        "ai_discovery_rate": ai_rate,
        "manual_discovery_rate": manual_rate,
        "ai_advantage": len(ai_unique) > len(manual_unique)
    }