I did mention there are optimizations at play over a pure merge sort.
In fact, my testing shows that passing n sorted items, it only performs n comparisons. This is clearly not the behaviour of a pure merge sort.
But I was indeed wrong about the amount being constant.
In reply to Re^3: sort of misunderstanding of sort
by ikegami
in thread sort of misunderstanding of sort
by Discipulus
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |