`stri_rand_strings()` should allow `length = integer()`

[DavisVaughan]

I've got some fuzzing code that generates random strings like this

n_strings <- sample(0:10000, size = 1)
string_size <- sample(0:100, n_strings, TRUE)
stringi::stri_rand_strings(n_strings, string_size)

Notice how the length of string_size is equal to the number of strings, i.e. n_strings, so that it can be provided to stri_rand_strings().

This fails when we get n_strings = 0:

stringi::stri_rand_strings(0, integer())

I'd argue this should probably be allowed?

[gagolews]

Good point, anyone keen on fixing this border case?