Several recently proposed program logics have incorporated notions of \emph{underapproximation} into their design, enabling them to reason about reachability rather than safety. In this paper, we explore how similar ideas can be integrated into an expressive type and effect system. We use the resulting underapproximate type specifications to guide the synthesis of test generators that probe the behavior of effectful black-box systems. A key novelty of our type language is its ability to capture underapproximate behaviors of \emph{effectful} operations using symbolic traces that expose latent data and control dependencies, constraints that must be preserved by the test sequences the generator outputs. We implement this approach in a tool called \name{}, and evaluate it on a diverse range of applications by integrating \name{}'s synthesized generators into property-based testing frameworks like QCheck and model-checking tools like P. In both settings, the generators synthesized by \name{} are significantly more effective than the default testing strategy, and are competitive with state-of-the-art, handwritten solutions.
Thu 18 JunDisplayed time zone: Mountain Time (US & Canada) change
14:00 - 15:20 | Generative Testing and Program SynthesisPLDI Research Papers at Flatirons 2 Chair(s): Stefan K. Muller University of Connecticut | ||
14:00 20mTalk | [SIGPLAN OOPSLA’25] Scalable and Accurate Application-Level Crash-Consistency Testing via Representative Testing PLDI Research Papers Yile Gu University of Washington, Ian Neal University of Michigan and Veridise, Jiexiao Xu University of Washington, Shaun Christopher Lee University of Washington, Ayman Said University of Michigan, Musa Haydar University of Michigan, Jacob Van Geffen , Andrew Quinn University of California at Santa Cruz, Baris Kasikci University of Washington | ||
14:20 20mTalk | Trace-Guided Synthesis of Effectful Test Generators PLDI Research Papers Zhe Zhou Purdue University, Ankush Desai Snowflake, Benjamin Delaware Purdue University, Suresh Jagannathan Purdue University DOI | ||
14:40 20mTalk | Semantic Reification: A New Paradigm for Random Program Generation PLDI Research Papers Kavya Chopra ETH Zurich, Cong Li ETH Zurich, Thodoris Sotiropoulos ETH Zurich, Zhendong Su ETH Zurich DOI Pre-print | ||
15:00 20mTalk | The Search for Constrained Random Generators PLDI Research Papers Harrison Goldstein SUNY Buffalo, Hila Peleg Technion, Cassia Torczon University of Pennsylvania, Daniel Sainati University of Pennsylvania, Leonidas Lampropoulos University of Maryland at College Park, Benjamin C. Pierce University of Pennsylvania DOI | ||