Convert UTF-8 (Bytes) Emoji Code to Emoji icon as a text

Question

I am getting this below string as a response from WS API when they send emoji as a string:

let strTemp = "Hii \\xF0\\x9F\\x98\\x81"

I want it to be converted to the emoji icon like this -> Hii 😁

I think so it is coming in UTF-8 Format as explained in the below Image: Image Unicode

I have tried decoding it Online using UTF-8 Decoder

And i got the emoticon Successfully decoded

Before Decoding:

After Decoding:

But the issue here is I do not know how to work with it in Swift.

I referred following link but it did not worked for me.

Swift Encode/decode emojis

Any help would be appreciated.

Thanks.

Answer 1

As you already given the link of converter tool which is clearly doing UTF-8 encoding and decoding. You have UTF-8 encoded string so here is an example of UTF8-Decoding.

Objective-C

const char *ch = [@"Hii \xF0\x9F\x98\x81" cStringUsingEncoding:NSUTF8StringEncoding];
NSString *decode_string = [NSString stringWithUTF8String:ch];
NSLog(@"%@",decode_string);

Output: Hii 😁

---

Swift

I'm able to convert \\xF0\\x9F\\x98\\x81 to 😁 in SWift. First I converted the hexa string into Data and then back to String using UTF-8 encoding.

var str = "\\xF0\\x9F\\x98\\x81"
if let data = data(fromHexaStr: str) {
     print(String(data: data, encoding: String.Encoding.utf8) ?? "")
}

Output: 😁

Below is the function I used to convert the hexa string into data. I followed this answer.

func data(fromHexaStr hexaStr: String) -> Data? {
    var data = Data(capacity: hexaStr.characters.count / 2)
    let regex = try! NSRegularExpression(pattern: "[0-9a-f]{1,2}", options: .caseInsensitive)
    regex.enumerateMatches(in: hexaStr, range: NSMakeRange(0, hexaStr.utf16.count)) { match, flags, stop in
        let byteString = (hexaStr as NSString).substring(with: match!.range)
        var num = UInt8(byteString, radix: 16)!
        data.append(&num, count: 1)
    }

    guard data.count > 0 else { return nil }

    return data
}

Note: Problem with above code is it converts hexa string only not combined strings.

FINAL WORKING SOLUTION: SWIFT

I have done this by using for loop instead of [0-9a-f]{1,2} regex because this will also scan 81, 9F, Any Two digits number which is wrong obviously.

For example: I have 81 INR \\xF0\\x9F\\x98\\x81.

/// This line will convert "F0" into hexa bytes
let byte = UInt8("F0", radix: 16)

I made a String extension in which I check upto every 4 characters if it has prefix \x and count 4 and last two characters are convertible into hexa bytes by using radix as mentioned above.

extension String {

    func hexaDecoededString() -> String {

        var newData = Data()
        var emojiStr: String = ""
        for char in self.characters {

            let str = String(char)
            if str == "\\" || str.lowercased() == "x" {
                emojiStr.append(str)
            }
            else if emojiStr.hasPrefix("\\x") || emojiStr.hasPrefix("\\X") {
                emojiStr.append(str)
                if emojiStr.count == 4 {
                    /// It can be a hexa value
                    let value = emojiStr.replacingOccurrences(of: "\\x", with: "")
                    if let byte = UInt8(value, radix: 16) {
                        newData.append(byte)
                    }
                    else {
                        newData.append(emojiStr.data(using: .utf8)!)
                    }
                    /// Reset emojiStr
                    emojiStr = ""
                }
            }
            else {
                /// Append the data as it is
                newData.append(str.data(using: .utf8)!)
            }
        }

        let decodedString = String(data: newData, encoding: String.Encoding.utf8)
        return decodedString ?? ""
    }
}

USAGE:

var hexaStr = "Hi \\xF0\\x9F\\x98\\x81 81"
print(hexaStr.hexaDecoededString())

Hi 😁 81

hexaStr = "Welcome to SP19!\\xF0\\x9f\\x98\\x81"
print(hexaStr.hexaDecoededString())

Welcome to SP19!😁

Answer 2

I fix your issue but it need more work to make it general , the problem here is that your Emijo is Represented by Hex Byte x9F , so we have to convert this Hex to utf8 then convert it to Data and at last convert data to String

Final result Hii 😁 Please read comment

 let strTemp = "Hii \\xF0\\x9F\\x98\\x81"


            let regex = try! NSRegularExpression(pattern: "[0-9a-f]{1,2}", options: .caseInsensitive)
            // get all matched hex  xF0 , x9f,..etc

            let matches = regex.matches(in: strTemp, options: [], range: NSMakeRange(0, strTemp.count))


            // Data that will hanlde convert hex to UTf8
            var emijoData = Data(capacity: strTemp.count / 2)

            matches.enumerated().forEach { (offset , check) in
                let byteString = (strTemp as NSString).substring(with: check.range)
                var num = UInt8(byteString, radix: 16)!
                emijoData.append(&num, count: 1)
            }

            let subStringEmijo = String.init(data: emijoData, encoding: String.Encoding.utf8)!
            //now we have your emijo text  😁 we can replace by its code from string using matched ranges `first` and `last`

            // All range range of  \\xF0\\x9F\\x98\\x81 in "Hii \\xF0\\x9F\\x98\\x81" to replce by your emijo

            if let start = matches.first?.range.location, let end = matches.last?.range.location  , let endLength = matches.last?.range.length {

                let startLocation = start  - 2
                let length = end - startLocation + endLength

                let sub = (strTemp as NSString).substring(with: NSRange.init(location: startLocation, length: length))

                print( strTemp.replacingOccurrences(of: sub, with: subStringEmijo))
              // Hii 😁

            }