Reputation: 1841
We know that String.utf16 provides the codeunits or String.unicodeScalars provides the scalars.
If we manipulate the codeunits and unicodeScales by removing some elements etc. is there a way to construct back the resulting string?
Upvotes: 10
Views: 6583
Reputation: 7623
The answer is as simple as:
/// An array of the UTF-16 for "Hello, world!".
let a: [UTF16.CodeUnit] = Array("Hello, world!".utf16)
/// A string representation of a, interpreted as UTF-16
let s = String(decoding: a, as: UTF16.self) // <=== The API you want
print(s)
Upvotes: 1
Reputation: 539805
Update for Swift 2.1:
You can create a String
from an array of UTF-16 characters
with the
public init(utf16CodeUnits: UnsafePointer<unichar>, count: Int)
initializer. Example:
let str = "H€llo 😄"
// String to UTF16 array:
let utf16array = Array(str.utf16)
print(utf16array)
// Output: [72, 8364, 108, 108, 111, 32, 55357, 56836]
// UTF16 array to string:
let str2 = String(utf16CodeUnits: utf16array, count: utf16array.count)
print(str2)
// H€llo 😄
Previous answer:
There is nothing "built-in" (as far as I know), but you can use the UTF16
struct
which provides a decode()
method:
extension String {
init?(utf16chars:[UInt16]) {
var str = ""
var generator = utf16chars.generate()
var utf16 : UTF16 = UTF16()
var done = false
while !done {
let r = utf16.decode(&generator)
switch (r) {
case .EmptyInput:
done = true
case let .Result(val):
str.append(Character(val))
case .Error:
return nil
}
}
self = str
}
}
Example:
let str = "H€llo 😄"
// String to UTF16 array:
let utf16array = Array(str.utf16)
print(utf16array)
// Output: [72, 8364, 108, 108, 111, 32, 55357, 56836]
// UTF16 array to string:
if let str2 = String(utf16chars: utf16array) {
print(str2)
// Output: H€llo 😄
}
Slightly more generic, you could define a method that creates a string from an array (or any sequence) of code points, using a given codec:
extension String {
init?<S : SequenceType, C : UnicodeCodecType where S.Generator.Element == C.CodeUnit>
(codeUnits : S, var codec : C) {
var str = ""
var generator = codeUnits.generate()
var done = false
while !done {
let r = codec.decode(&generator)
switch (r) {
case .EmptyInput:
done = true
case let .Result(val):
str.append(Character(val))
case .Error:
return nil
}
}
self = str
}
}
Then the conversion from UTF16 is done as
if let str2a = String(codeUnits: utf16array, codec: UTF16()) {
print(str2a)
}
Here is another possible solution. While the previous methods are "pure Swift", this one uses the Foundation framework and the automatic
bridging between NSString
and Swift String
:
extension String {
init?(utf16chars:[UInt16]) {
let data = NSData(bytes: utf16chars, length: utf16chars.count * sizeof(UInt16))
if let ns = NSString(data: data, encoding: NSUTF16LittleEndianStringEncoding) {
self = ns as String
} else {
return nil
}
}
}
Upvotes: 9
Reputation: 1637
Here it is.
extension String {
static func fromUTF16Chars(utf16s:UInt16[]) -> String {
var str = ""
for var i = 0; i < utf16s.count; i++ {
let hi = Int(utf16s[i])
switch hi {
case 0xD800...0xDBFF:
let lo = Int(utf16s[++i])
let us = 0x10000
+ (hi - 0xD800)*0x400 + (lo - 0xDC00)
str += Character(UnicodeScalar(us))
default:
str += Character(UnicodeScalar(hi))
}
}
return str
}
}
let str = "aαあ🐣aαあ🐣"
var utf16cs = UInt16[]()
for utf16c in str.utf16 {
utf16cs += utf16c
}
let str2 = String.fromUTF16Chars(utf16cs)
assert(str2 == str)
println(str2)
Upvotes: 0