Using `Vec::extend(std::iter::repeat_n(item, N))` allows to use the more
natural number of elements to add `N`, as is probably done in the
original loop, instead of computing the difference between the existing
number of elements and the wanted one.
Before MSRV 1.82, the older suggestion to use `Vec::resize()` is still
issued.
Inspired by #6156 (which predates `repeat_n()`).
changelog: [`same_item_push`]: recommend using `Vec::extend()` to extend
a vector
67e6bf33fe