GraphemeSplitter

A C# implementation of the Unicode grapheme cluster breaking algorithm.

Notes

• This library uses Unicode 10.0 version of grepheme boundary algorithm.
• In .NET 5.0, StringInfo.GetTextElementEnumerator can enumerate graphemes correctly with Unicode 13.0 algorithm.

NuGet package

https://www.nuget.org/packages/GraphemeSplitter/

Install-Package GraphemeSplitter

Sample

using GraphemeSplitter;
using static System.Console;
using static System.String;

public partial class Program
{
    static string Split(string s) => Join(", ", s.GetGraphemes());

    static void Main()
    {
        WriteLine(Split("👨‍👨‍👧‍👦👩‍👩‍👧‍👦👨‍👨‍👧‍👦")); // 👨‍👨‍👧‍👦, 👩‍👩‍👧‍👦, 👨‍👨‍👧‍👦
    }
}

Web Sample:

Implementation

This library basically implements http://unicode.org/reports/tr29/.

Expample:

type	text	split result
diacritical marks	à̡̠́ḅ̢̂̃c̣̤̃̄d̥̦̅̆	"à̡̠́", "ḅ̢̂̃", "c̣̤̃̄", "d̥̦̅̆"
variation selector	葛葛󠄀葛󠄁	"葛", "葛󠄀", "葛󠄁"
asian syllable	안녕하세요	"안", "녕", "하", "세", "요"
family emoji	👨‍👨‍👧‍👦👩‍👩‍👧‍👦👨‍👨‍👧‍👦	"👨‍👨‍👧‍👦", "👩‍👩‍👧‍👦", "👨‍👨‍👧‍👦"
emoji skin tone	👩🏻👱🏼👧🏽👦🏾	"👩🏻", "👱🏼", "👧🏽", "👦🏾"

but slacks out the GB10, GB12, and GB13 rules for simplification.

original:

• GB10 … (E_Base | EBG) Extend* × E_Modifier
• GB12 … sot (RI RI)* RI × RI
• GB13 … [^RI] (RI RI)* RI × RI

implemented:

• GB10 … (E_Base | EBG) × Extend
• GB10 … (E_Base | EBG | Extend) × E_Modifier
• GB12/GB13 … RI × RI

Difference is:

sequence	original	implemented
à🏻‍ (U+61, U+300, U+1F3FB)	× ÷	× ×
🇯🇵🇺🇸 (U+1F1EF, U+1F1F5, U+1F1FA, U+1F1F8)	× ÷ ×	× × ×

(where ÷ and × means boundary and no bounadry respectively.)

Acknowledgements

This library is influenced by

• https://github.com/devongovett/grapheme-breaker
• https://github.com/orling/grapheme-splitter
• https://github.com/unicode-rs/unicode-segmentation