On Sunday, 8 May 2016 at 23:38:31 UTC, Jon D wrote:
* Performance varied widely between compilers, and for DMD,
whether the -inline flag was included. The performance delta
between asLowerCase and the mapAsLowerCase replacement was very
dependent on these choices. Similarly, the delta between
inclusion and exclusion of auto-decoding was highly dependent
on these selections.
This makes me suspicious, since you did not measure standard
deviation or variance. A quick local test with your code looks
ok, though.
One of my students recently made a benchmarking tool.
If you really want to do this properly, you could try that:
https://github.com/parttimenerd/temci
I would not require better statistics here, since the speedups
seem to be stable. :)