How to initialise a string from NSData in Swift

Solution 1:

This is the implemented code needed:

in Swift 3.0:

var dataString = String(data: fooData, encoding: String.Encoding.utf8)

or just

var dataString = String(data: fooData, encoding: .utf8)

Older swift version:

in Swift 2.0:

import Foundation

var dataString = String(data: fooData, encoding: NSUTF8StringEncoding)

in Swift 1.0:

var dataString = NSString(data: fooData, encoding:NSUTF8StringEncoding)

Solution 2:

This is how you should initialize the NSString:

Swift 2.X or older

let datastring = NSString(data: fooData, encoding: NSUTF8StringEncoding)

Swift 3 or newer:

let datastring = NSString(data: fooData, encoding: String.Encoding.utf8.rawValue)

This doc explains the syntax.

Solution 3:

Swift 2.0

It seems that Swift 2.0 has actually introduced the String(data:encoding:) as an String extension when you import Foundation. I haven't found any place where this is documented, weirdly enough.

(pre Swift 2.0) Lightweight extension

Here's a copy-pasteable little extension without using NSString, let's cut the middle-man.

import Foundation

extension NSData
{
    var byteBuffer : UnsafeBufferPointer<UInt8> { get { return UnsafeBufferPointer<UInt8>(start: UnsafeMutablePointer<UInt8>(self.bytes), count: self.length) }}
}

extension String
{
    init?(data : NSData, encoding : NSStringEncoding)
    {
        self.init(bytes: data.byteBuffer, encoding: encoding)
    }
}

// Playground test
let original = "Nymphs blitz quick vex dwarf jog"
let encoding = NSASCIIStringEncoding

if let data = original.dataUsingEncoding(encoding)
{
    String(data: data, encoding: encoding)
}

This also give you access to data.byteBuffer which is a sequence type, so all those cool operations you can do with sequences also work, like doing a reduce { $0 &+ $1 } for a checksum.

Notes

In my previous edit, I used this method:

var buffer = Array<UInt8>(count: data.length, repeatedValue: 0x00)
data.getBytes(&buffer, length: data.length)
self.init(bytes: buffer, encoding: encoding)

The problem with this approach, is that I'm creating a copy of the information into a new array, thus, I'm duplicating the amount of bytes (specifically: encoding size * data.length)