saved_models/codesearch_simp/gen_label.py

import json
import torch
import os

# file_path = '/home/yiming/ContrastDebugger/EXP/codesearch_query/Model/Epoch_1/index.json'

# # 打开并读取 JSON 文件
# with open(file_path, 'r') as file:
#     json_data = json.load(file)

# testset_label = None
# for i in range(len(json_data)):
#     if testset_label != None:
#         testset_label = torch.cat((testset_label, torch.tensor([0])), 0)
#     else:
#         testset_label = torch.tensor([0])

# torch.save(testset_label, "/home/yiming/ContrastDebugger/EXP/codesearch_query/Training_data/training_dataset_label.pth")

input_file = "/home/yiming/ContrastDebugger/EXP/codesearch/Model/label_list.json"
output_file = "/home/yiming/ContrastDebugger/EXP/codesearch/Model/new_label_list.json"  # 替换为输出文件的路径

# 读取输入文件
with open(input_file, "r") as f:
    data = json.load(f)

# 提取每个数据的前十个字符
processed_data = [item[:30] for item in data]

# 保存到新的 JSON 文件
with open(output_file, "w") as f:
    json.dump(processed_data, f)