swetry1
收藏资源简介:
该数据集包含与软件工程代码修复或测试生成任务相关的样本,每个样本代表一个来自GitHub仓库的实例,关联到拉取请求和问题跟踪。数据集定义了丰富的特征字段,包括:标识信息(如仓库名、拉取请求编号、实例ID、问题编号列表)、代码变更信息(如基础提交哈希、代码补丁、测试补丁)、任务描述(如问题陈述、提示文本、所有提示文本)、过程信息(如提交URL列表、创建时间、提交URL)、执行命令(如重建命令列表、测试命令列表、打印命令列表)、测试结果(如从失败到通过的测试列表、从通过到通过的测试列表)以及环境配置(如日志解析器、Docker镜像)。每个样本还包含一个难度评估结构体,记录涉及的文件数、代码块数和代码行数。数据集目前仅包含一个名为python的分割,共有9个样本,总数据集大小约为535.58 KB,下载大小约为517.57 KB,适用于代码自动修复、测试用例生成、程序分析或软件维护等相关的研究与开发任务。
This dataset contains samples related to software engineering code repair and test generation tasks. Each sample represents an instance sourced from a GitHub repository, linked to pull requests and issue trackers. The dataset defines a rich set of feature fields, including identification information (such as repository name, pull request number, instance ID, list of issue numbers), code change information (such as base commit hash, code patch, test patch), task description (such as problem statement, prompt text, all prompt texts), process information (such as list of commit URLs, creation time, commit URL), execution commands (such as list of rebuild commands, list of test commands, list of print commands), test results (such as list of tests that changed from failed to passed, list of tests that remained passed), and environment configuration (such as log parser, Docker image). Each sample also includes a difficulty assessment structure that records the number of involved files, number of code blocks, and lines of code. Currently, the dataset only contains one split named "python", with a total of 9 samples. The total dataset size is approximately 535.58 KB, and the download size is about 517.57 KB. This dataset is applicable to research and development tasks related to automatic code repair, test case generation, program analysis, or software maintenance.
数据集概述:lieeli/swetry1
- 数据集地址:https://huggingface.co/datasets/lieeli/swetry1
数据特征
该数据集包含以下字段:
repo:仓库名称(字符串)pull_number:拉取请求编号(字符串)instance_id:实例ID(字符串)issue_numbers:相关议题编号列表(字符串列表)base_commit:基础提交哈希(字符串)patch:补丁内容(字符串)test_patch:测试补丁内容(字符串)problem_statement:问题描述(字符串)hints_text:提示文本(字符串)all_hints_text:所有提示文本(字符串)commit_urls:提交URL列表(字符串列表)created_at:创建时间(字符串)commit_url:提交URL(字符串)rebuild_cmds:重建命令列表(字符串列表)test_cmds:测试命令列表(字符串列表)print_cmds:打印命令列表(字符串列表)log_parser:日志解析器(字符串)FAIL_TO_PASS:失败到通过的测试列表(字符串列表)PASS_TO_PASS:通过到通过的测试列表(字符串列表)docker_image:Docker镜像名称(字符串)difficulty:难度结构,包含:files:文件数量(整数)hunks:补丁块数量(整数)lines:代码行数(整数)
数据集划分
- 划分名称:
python - 样本数量:9条
- 数据大小:535,580字节(下载大小:517,571字节)
配置文件
- 配置名称:
default - 数据文件:
split: python,路径为data/python-*




