Tests4Py

Tests4Py是由亥姆霍兹信息安全中心和柏林洪堡大学联合创建的数据集，专注于系统测试。该数据集包含79个bug，来源于七个真实世界的Python应用和四个示例程序。每个bug都配备了用于验证功能正确性的oracle，并支持系统测试和单元测试的生成。Tests4Py的设计强调测试多样性，旨在促进更广泛和严格的评估，适用于研究测试生成、调试和自动程序修复。

Tests4Py is a dataset co-developed by the Helmholtz Center for Information Security and Humboldt-Universität zu Berlin, focusing on system testing. This dataset contains 79 bugs sourced from seven real-world Python applications and four example programs. Each bug is equipped with an oracle for verifying functional correctness, and supports the generation of both system tests and unit tests. Tests4Py’s design emphasizes test diversity, aiming to facilitate broader and more rigorous evaluations, and is applicable to research on test generation, debugging, and automatic program repair.