See the newest version of my benchmark tool. Once I upped the # of iterations to 10000, CommonMarkSharp experienced a huge drop in performance for small inputs (as large as 5 sec vs 500ms). Of course it is still ~0.5ms per iteration but its 10x faster when the number of iterations is 2000 so something strange is going on...
empty.text 12 B (10000 iterations)
Library Total Each vs Baseline
--------------------------------------------------
Baseline 529 0 100%
CommonMark.NET 243 0 46%
CommonMarkSharp 5467 1 1033%
MarkdownSharp 582 0 110%
MarkdownDeep 269 0 51%
See the newest version of my benchmark tool. Once I upped the # of iterations to 10000, CommonMarkSharp experienced a huge drop in performance for small inputs (as large as 5 sec vs 500ms). Of course it is still ~0.5ms per iteration but its 10x faster when the number of iterations is 2000 so something strange is going on...