Emoji conversion to a specific string representation

Question

Currently I'm using a HashSet of Tuples called Emoji to replace Emoji to a string representation so that for example the emoji for bomb becomes U0001F4A3. The conversion's done via

Emoji.Aggregate(input, (current, pair) => current.Replace(pair.Item1, pair.Item2));

Works as expected.

However I'm trying to achieve the same thing without making use of predefined list of 2600+ items. Did anyone already achieve such a thing where the Emoji in a string are replaced with their counterpart without leading \?

For example:

"This string contains the unicode character bomb (💣)"

becomes

"This string contains the unicode character bomb (U0001F4A3)"

Answer 1

It sounds like you're happy to replace any character not in the basic multi-lingual plane with its hex representation. The code to do that is slightly longwinded, but it's pretty simple:

using System;
using System.Text;

class Test
{
    static void Main()
    {
        string text = "This string contains the unicode character bomb (\U0001F4A3)";
        Console.WriteLine(ReplaceNonBmpWithHex(text));
    }

    static string ReplaceNonBmpWithHex(string input)
    {
        // TODO: If most string don't have any non-BMP characters, consider
        // an optimization of checking for high/low surrogate characters first,
        // and return input if there aren't any.
        StringBuilder builder = new StringBuilder(input.Length);
        for (int i = 0; i < input.Length; i++)
        {
            char c = input[i];
            // A surrogate pair is a high surrogate followed by a low surrogate
            if (char.IsHighSurrogate(c))
            {
                if (i == input.Length -1)
                {
                    throw new ArgumentException($"High surrogate at end of string");
                }
                // Fetch the low surrogate, advancing our counter
                i++;
                char d = input[i];
                if (!char.IsLowSurrogate(d))
                {
                    throw new ArgumentException($"Unmatched low surrogate at index {i-1}");
                }
                uint highTranslated = (uint) ((c - 0xd800) * 0x400);
                uint lowTranslated = (uint) (d - 0xdc00);
                uint utf32 = (uint) (highTranslated + lowTranslated + 0x10000);
                builder.AppendFormat("U{0:X8}", utf32);
            }
            // We should never see a low surrogate on its own
            else if (char.IsLowSurrogate(c))
            {
                throw new ArgumentException($"Unmatched low surrogate at index {i}");
            }
            // Most common case: BMP character; just append it.
            else
            {
                builder.Append(c);
            }
        }
        return builder.ToString();
    }
}

Note that this does not attempt to handle the situation where multiple characters are used together, as per Yury's answer. It would replace each modifier/emoji/secondary-char as a separate UXXXXXXXX part.

Answer 2

I'm afraid you have one false assumption here. Emoji is not just a "special Unicode char". Actual length of particular emoji can be 4 or more chars in a row. For instance:

• emoji itself
• zero-width jointer
• secondary char (like graduation cap or microphone)
• gender modifier (man or woman)
• skin tone modifier (Fitzpatrick Scale)

So, you should take into consideration that variable length for sure.

Examples:

• https://emojipedia.org/female-health-worker
• https://emojipedia.org/male-farmer-type-1-2