Binary multisearch

Description:
	Given a sorted list of unique items, find in another
	sorted list their position or if missing where they
	ought to be inserted. Both lists must allow random access.

Complexity:
	n = number of elements in target list
	m = number of elements to locate

	Time complexity: O(S(n) + m)
		S = time complexity function of underlying search algorithm,
			e.g. S(n) = O(log n) in average case for binary search given O(1) comparison.

	Auxiliary space complexity:
		List return version:	O(m)
		Online version:     	O(log m)

Notes:
	In online version, positions for elements are returned in
	arbitrary (deterministic) order, rather than sequential.

I wrote this algorithm back in February 2013.