Ephyra:面向大语言模型时代的基于属性的测试框架
Ephyra: A Property-Based Testing Framework for the LLM Era
AI 导读
论文提出 Ephyra,一个用 Rust 编写的基于属性的测试(PBT)库,围绕可定制性、可观测性和生成器合成三个概念构建,允许用户自由组合生成、变异和 shrinking 组件来满足功能谓词。该框架已在 14 个开源系统上应用,发现 349 个 bug 并报告了 300 个,涉及 SQLite、PostgreSQL、DuckDB、Cedar 等数据库、策略引擎与编译器。
正文 · 原文
Abstract:Coding agents now write a growing share of software tests; however, property-based testing (PBT) libraries are designed with historical considerations. We present Ephyra, a PBT library in Rust built on three concepts: customizability, observability and generator synthesis. Ephyra lets users compose their own testing loops from generation, mutation and shrinking, and synthesize these components to satisfy functional predicates. Applying Ephyra to 14 open source systems, we have found 349 bugs and reported 300 of them, in database management systems (SQLite, Turso, PostgreSQL, DuckDB, DataFusion), the verified policy engine (Cedar), the SQL equivalence prover (QED), the URL parser used in this http URL (ada-url), reimplementations of existing open source systems in Rust (coreutils, brush, gitoxide), and compilers (binaryen, OCaml, and OxCaml).
| Subjects: | Software Engineering (cs.SE); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.23282 [cs.SE] |
| (or arXiv:2611.23282v1 [cs.SE] for this version) | |
| https://doi.org/10.48550/arXiv.2611.23282 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Alperen Keles [view email]
[v1]
Fri, 23 Oct 2026 21:07:17 UTC (34 KB)
来源:arXiv cs.AI · arxiv.org