import json
import csv

def trainjson2csv(json_file_path = 'output_results.json' ,csv_file_path = 'output.csv'):
    # 打开CSV文件以写入
    with open(csv_file_path, 'w', newline='', encoding='utf-8') as csv_file:
        writer = None  # 初始化写入器

        # 逐行读取JSON文件
        with open(json_file_path, 'r', encoding='utf-8') as json_file:
            for line in json_file:
                # 解析JSON行
                data = json.loads(line.strip())

                # 如果writer未初始化，则根据JSON键设置CSV的列头
                if writer is None:
                    writer = csv.DictWriter(csv_file, fieldnames=data.keys())
                    writer.writeheader()

                # 写入当前行数据
                writer.writerow(data)

    print(f"JSON 数据已成功逐行转换为 CSV 文件，并保存为 {csv_file_path}")

def origin2json(originfilepath):
    # 打开原始 JSON 文件
    with open(originfilepath, 'r', encoding='utf-8') as file:
        datas = file.readlines()
        datas = json.loads(datas[0])

    results = []

    # 遍历数据并处理
    for data in datas:
        for key, value in data['mapping'].items():
            # 检查 'message' 和 'content' 是否存在并且不为 None
            if value.get('message') and value['message'].get('content'):
                input_text = value['message']['content'].get('parts', [None])[0]
            else:
                input_text = None

            # 提取子节点（输出）
            if value.get('children'):
                child_id = value['children'][0]
                if data['mapping'].get(child_id) and data['mapping'][child_id].get('message') and data['mapping'][child_id]['message'].get('content'):
                    output_text = str(data['mapping'][child_id]['message']['content'].get('parts', [None])[0]).replace("\n","")
                else:
                    output_text = None
            else:
                output_text = None

            # 仅在 input 和 output 都存在的情况下构建字典并添加到结果列表
            if input_text and output_text:
                result = {
                    "title": data['title'],
                    # "id": value['id'],
                    "input": input_text,
                    "output": output_text
                }

                results.append(result)

    # 将每个结果的 dict 写入新的 JSON 文件，每行一个 JSON 对象
    with open('output_results.json', 'w', encoding='utf-8') as outfile:
        for result in results:
            json.dump(result, outfile, ensure_ascii=False)
            outfile.write('\n')
    print("数据已成功保存到 output_results.json 文件中")

if __name__ == '__main__':
    originfilepath = '7239cd96cb1ff7cc6db225c7bed1c1c4216352d844513d89e3d767f9a6db5c71-2024-08-09-06-34-56/conversations.json' #替换为你导出的原始文件中conversations.json的位置
    json_file_path = 'output_results.json'  # 你需要替换为你的JSON文件路径
    csv_file_path = 'output.csv'            # 你需要替换为你的csv文件路径
    origin2json(originfilepath)
    trainjson2csv(json_file_path,csv_file_path)