A variant of @OOPer's solution would be to use a conditionally binding while loop and index(_:offsetBy:limitedBy:) in order to iterate over the 8 character substrings, taking advantage of the fact that index(_:offsetBy:limitedBy:) returns nil when you try to advance past the limit.
let binaryBits = "010010000110010101111001"
var result = ""
var index = binaryBits.startIndex
while let next = binaryBits.index(index, offsetBy: 8, limitedBy: binaryBits.endIndex) {
let asciiCode = UInt8(binaryBits[index..<next], radix: 2)!
result.append(Character(UnicodeScalar(asciiCode)))
index = next
}
print(result) // Hey
Note that we're going via Character rather than String in the intermediate step – this is simply to take advantage of the fact that Character is specially optimised for cases where the UTF-8 representation fits into 63 bytes, which is the case here. This saves heap-allocating an intermediate buffer for each character.
Purely for the fun of it, another approach could be to use sequence(state:next:) in order to create a sequence of the start and end indices of each substring, and then reduce in order to concatenate the resultant characters together into a string:
let binaryBits = "010010000110010101111001"
// returns a lazily evaluated sequence of the start and end indices for each substring
// of 8 characters.
let indices = sequence(state: binaryBits.startIndex, next: {
index -> (index: String.Index, nextIndex: String.Index)? in
let previousIndex = index
// Advance the current index – if it didn't go past the limit, then return the
// current index along with the advanced index as a new element of the sequence.
return binaryBits.characters.formIndex(&index, offsetBy: 8, limitedBy: binaryBits.endIndex) ? (previousIndex, index) : nil
})
// iterate over the indices, concatenating the resultant characters together.
let result = indices.reduce("") {
$0 + String(UnicodeScalar(UInt8(binaryBits[$1.index..<$1.nextIndex], radix: 2)!))
}
print(result) // Hey
On the face of it, this appears to be much less efficient than the first solution (due to the fact that reduce should copy the string at each iteration) – however it appears the compiler is able to perform some optimisations to make it not much slower than the first solution.