Sporq: An Interactive Environment for Exploring Code Using Query-by-Example
summary
Sporq is an interactive VS Code extension that converts a codebase with CodeQL into relational program representations, synthesizes Datalog queries from highlighted examples, and iterates with users via labels. On 17 benchmark patterns it solves 14 within 10s and the rest within a minute, typically needing under five rounds (12/17) and under five labels (11/17). A 5‑participant study finds more patterns found than with Semmle or SemGrep, helped by uniqueness constraints and generalized feedback.
AI-written summary · about
chapters (6)
- 0:00Introduction to Sporq
- 1:56How Sporq works
- 3:03Demo finding casts
- 6:43Query synthesis results
- 7:31Evaluation of Sporq
- 9:30Conclusion and summary
related talks
57mVariant analysis
08mSupercharging Developer Productivity With Advanced Code Search
20mDevX: History of Code Search
37mEnhancing VS Code extensions with GitHub Copilot
38mCode reading
10mTo search or not to search: it depends on the question
36mCode as Data: How data insights on legacy codebases fill the knowledge gap in modernization projects
32mSynthesizing Code Quality Rules from Examples