Features

• Add support for Luau

Fixes

• Fix fallback to default os.clock when higher precision clocks are unvailable

Full Changelog

https://github.com/jeffzi/luamark/compare/v0.6.0...v0.7.0

Features

• Rank results by median instead of mean

Full Changelog

https://github.com/jeffzi/luamark/compare/v0...v0.6.0

Features

• Replace print_summary with summarize which returns a string instead of printing directly. The __tostring metamethod on the results table make it easier to print with just print(results).
• Tweak result formatting: center and re-order headers
• Convert to human-readable units. The results table still uses seconds and kilobytes for maximum flexibility.

Fixes

• Sort ranks as numbers instead of lexical orders.

Full Changelog

https://github.com/jeffzi/luamark/compare/v0...v0.5.0

Features

• Remove measure_time and measure_memory from public api. timeit and memit cover the same usage.

Fixes

• Execute setup and teardown functions every time the measured function is called, even when calibrating the number of iterations and rounds.

Full Changelog

https://github.com/jeffzi/luamark/compare/v0.3.1...v0.4.0

Features

• Add setup and teardown arguments to timeit and memit.

Fixes

• Remove duplicate total in stats table

Full Changelog

https://github.com/jeffzi/luamark/compare/v0.3.0...v0.3.1

Restructured the library to focus on user-friendliness while simultaneously enhancing accuracy. This is achieved by supporting multiple Lua modules that provide clock time.

Full Changelog

https://github.com/jeffzi/luamark/compare/v0.2.0...v0.3.0

Features

• auto-discovery of runs arg
• return stats instead of samples

Full Changelog

https://github.com/jeffzi/luamark/compare/v0.1.0...v0.2.0

Initial Release

LuaMark is a portable microbenchmarking library for Lua, offering precise measurement of execution time and memory usage.