faster splitter

[qznc via Digitalmars-d]

On Friday, 27 May 2016 at 14:41:29 UTC, Chris wrote:
> The improved `std find` comes very close to the manual function 
> above now, sometimes it's even faster or at least as fast.
>
> std find    	took     12573666
> manual find 	took      7418454
> my std find 	took      6903854 <===
> findStringS 	took      7166720
> findStringS_  	took      6579529 <===

I just updated my benchmark. It now checks lots of different 
scenarios instead of you having to specify one. You can only set 
the number of iterations now.

It generates a random scenario (haystack length, needle length, 
alphabet, etc) and runs it once with all algorithms. Instead of 
recording the absolute runtime, it records the slowdown compared 
to the fastest of the three. Average those and also compute the 
absolute deviation (the second number).

Example:
std find:    153 ±33
manual find: 112 ±19
my std find: 119 ±16

So manual find is on average 12% slower than the fastest ±19%. I 
would interpret that as no significant difference between manual 
and improved find.