A packed integer array is represented as a pair of an integer k and a string s. The integer k is the number of bits per integer that we use. The string s is just an array of bits, which is read in 8-bit chunks.

The OCaml programming language treats string literals and array literals in slightly different ways: the former are statically allocated, while the latter are dynamically allocated. (This is rather arbitrary.) In the context of Menhir's table-based back-end, where compact, immutable integer arrays are needed, ocaml strings are preferable to ocaml arrays.

pack a turns an array of integers into a packed integer array.

Because the sign bit is the most significant bit, the magnitude of any negative number is the word size. In other words, as soon as the array a contains any negative numbers, pack does not achieve any space savings.

get1 s i returns the integer stored in the packed array s at index i. It assumes (and does not check) that the array's bit width is 1. The packed array s is just a string; there is no bit width component. In other words, it is get, specialized to the bit width 1.

get2 is get, specialized to the bit width 2.

get4 is get, specialized to the bit width 4.

get8 is get, specialized to the bit width 8.

get16 is get, specialized to the bit width 16.

get32 is get, specialized to the bit width 32.