IntersectMBO/plutus

Add golden tests for the validation benchmarks

Open

#6,828 建立於 2025年2月10日

在 GitHub 查看
 (1 留言) (0 反應) (0 負責人)Haskell (1,637 star) (511 fork)batch import
BenchmarksGood first issueInternalTeststatus: triagedtech debt

描述

Currently the validation benchmarks don't seem to have any golden tests ensuring that what scripts evaluate to doesn't change. Unlike e.g. nofib or bitwise benchmarks (and lists ones have some property tests). We should fix that, otherwise an apparent optimization might as well turn out to be a bug.

Same about marlowe?

貢獻者指南

技術棧
無
領域
testingperformance
議題類型
test
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
3
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
1-2 days
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
fresh
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
clear
前置要求
HaskellPlutus basicsgolden testing
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
60
研究方向
First, examine how golden tests are implemented for the 'nofib' and 'bitwise' benchmarks in the same repository (e.g., look at the test files for those benchmarks). Understand the pattern: likely there is a script that outputs expected values and a test framework that compares against them. Then, apply the same pattern to the 'validation' benchmarks directory. The issue also mentions 'marlowe' so decide if it should be included. After writing the golden tests, run them to ensure they pass. If there are any failing tests, investigate whether the benchmarks produce deterministic output.