Less method invalidations for OffsetInteger

[SimonDanisch]

Back in the day, I didn't have enough time to figure out why OffsetIntegers invalidate so many methods and how to fix it. The answer is, just let promotion do the hard work, and dont define unused methods :D

[ffreyer]

This made me check if OffsetIntegers are tested and they are, but only with an offset of 0. I guess that just covers the first method and not both?
Either way, it would probably be good to extend the test to include some nonzero offset

GeometryBasics.jl/test/runtests.jl

Lines 313 to 341 in fcb4edf

	@testset "Offsetintegers" begin
	x = 1
	@test GeometryBasics.raw(x) isa Int
	@test GeometryBasics.value(x) == x
	x = ZeroIndex(1)
	@test eltype(x) == Int
	x = OffsetInteger{0}(1)
	@test typeof(x) == OffsetInteger{0,Int}
	for x1 in [OffsetInteger{0}(2), 2, 0x02]
	@test Base.to_index(x1) == 2
	@test -(x1) == OffsetInteger{0,Int}(-2)
	@test abs(x1) == OffsetInteger{0,Int}(2)
	for x in [OffsetInteger{0}(1), 1, 0x01]
	@test +(x, x1) == OffsetInteger{0,Int}(3)
	@test *(x, x1) == OffsetInteger{0,Int}(2)
	@test -(x, x1) == OffsetInteger{0,Int}(-1)
	#test for /
	@test div(x, x1) == OffsetInteger{0,Int}(0)
	@test !==(x, x1)
	@test !>=(x, x1)
	@test <=(x, x1)
	@test !>(x, x1)
	@test <(x, x1)
	end
	end
	end