import json
import os
import uuid

def create_exam_cases():
    # 读取process_3.json文件
    with open('temp_analysis/process_3.json', 'r', encoding='utf-8') as f:
        data = json.load(f)
    
    exam_cases = []
    num = 0
    
    # 遍历每个URL条目
    for url, url_data in data.items():
        # 获取title作为question
        question = url_data.get('title', '').strip()
        
        # 遍历所有最短路径的元数据
        for path_meta in url_data.get('shortestPathsMeta', []):
            chain_ids = path_meta.get('chainIDs', [])
            chain_urls = path_meta.get('chainUrls', [])
            chain_ax_tree_ids = path_meta.get('chainAxTreeID', [])
            chain_texts = path_meta.get('chainTexts', [])

            trajectory_id = str(uuid.uuid4())
            trajectory_step_num = 0
            
            # 确保所有列表长度一致
            min_length = min(len(chain_ids), len(chain_urls), len(chain_ax_tree_ids), len(chain_texts))
            
            # 为每个步骤创建一个测试用例
            for i in range(min_length):
                num +=1
                trajectory_step_num += 1
                case = {
                    "num": num,
                    "id": str(uuid.uuid4()),
                    "trajectory_id": trajectory_id,
                    "trajectory_step_num": trajectory_step_num,
                    "page_id": chain_ids[i] if i < len(chain_ids) else "",
                    "url": chain_urls[i] if i < len(chain_urls) else "",
                    "question": question,
                    "answer": str(chain_ax_tree_ids[i]) if i < len(chain_ax_tree_ids) else "",
                    "answer_text": chain_texts[i] if i < len(chain_texts) else ""
                }
                exam_cases.append(case)
    
    # 将结果保存到exam.json
    with open('temp_analysis/exam.json', 'w', encoding='utf-8') as f:
        json.dump(exam_cases, f, ensure_ascii=False, indent=2)
    
    print(f"已生成 {len(exam_cases)} 个测试用例并保存到 temp_analysis/exam.json")

if __name__ == "__main__":
    create_exam_cases()