Add IsEmoji method

Emoji validation is performed by using a regex which is generated from Twemoji. Source: https://raw.githubusercontent.com/twitter/twemoji-parser/master/src/lib/regex.js

CHANGELOG.md

@@ -24,6 +24,9 @@
 - Added `IEnumerable<T>.ForEach` (#50)
 - Added `IEnumerable<T>.DisposeAll`
 - Added `IEnumerable<T>.DisposeAllAsync`
+- Added `char.IsEmoji`
+- Added `Rune.IsEmoji`
+- Added `string.IsEmoji`
 
 ## [3.0.0]
 

X10D.SourceGenerator/EmojiRegexGenerator.cs

@@ -0,0 +1,68 @@
+using System.Text;
+using System.Text.RegularExpressions;
+using Microsoft.CodeAnalysis;
+using Microsoft.CodeAnalysis.Text;
+
+namespace X10D.SourceGenerator;
+
+[Generator]
+internal sealed class EmojiRegexGenerator : ISourceGenerator
+{
+    // ReSharper disable once IdentifierTypo
+    private const string TwemojiRegexUrl = "https://raw.githubusercontent.com/twitter/twemoji-parser/master/src/lib/regex.js";
+    private static readonly HttpClient HttpClient = new();
+    private string _emojiRegex = string.Empty;
+
+    /// <inheritdoc />
+    public void Initialize(GeneratorInitializationContext context)
+    {
+        string response = HttpClient.GetStringAsync(TwemojiRegexUrl).GetAwaiter().GetResult();
+        using var reader = new StringReader(response);
+
+        while (reader.ReadLine() is { } line)
+        {
+            if (!line.StartsWith("export default /"))
+            {
+                continue;
+            }
+
+            Match match = Regex.Match(line, @"export default /(?<regex>.+)/g;");
+            if (!match.Success)
+            {
+                continue;
+            }
+
+            _emojiRegex = $"^{match.Groups["regex"].Value}$";
+            break;
+        }
+    }
+
+    /// <inheritdoc />
+    public void Execute(GeneratorExecutionContext context)
+    {
+        if (string.IsNullOrEmpty(_emojiRegex))
+        {
+            return;
+        }
+
+        var builder = new StringBuilder();
+        builder.AppendLine("// This file was auto-generated by X10D.SourceGenerator");
+        builder.AppendLine("// Do not edit this file manually");
+        builder.AppendLine();
+
+        builder.AppendLine("using System.Text.RegularExpressions;");
+        builder.AppendLine();
+        builder.AppendLine("namespace X10D.Text;");
+        builder.AppendLine();
+        builder.AppendLine("internal static class EmojiRegex");
+        builder.AppendLine("{");
+        builder.AppendLine("    internal static readonly Regex Value = new Regex(");
+        builder.AppendLine($"        @\"{_emojiRegex}\",");
+        // ReSharper disable once StringLiteralTypo
+        builder.AppendLine("        RegexOptions.Compiled | RegexOptions.CultureInvariant | RegexOptions.Singleline");
+        builder.AppendLine("    );");
+        builder.AppendLine("}");
+
+        context.AddSource("EmojiRegex.g.cs", SourceText.From(builder.ToString(), Encoding.UTF8));
+    }
+}

X10D.SourceGenerator/X10D.SourceGenerator.csproj

@@ -0,0 +1,18 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <TargetFramework>netstandard2.0</TargetFramework>
+        <LangVersion>10.0</LangVersion>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <Nullable>enable</Nullable>
+    </PropertyGroup>
+
+    <ItemGroup>
+        <PackageReference Include="Microsoft.CodeAnalysis.CSharp" Version="4.0.1" PrivateAssets="all"/>
+        <PackageReference Include="Microsoft.CodeAnalysis.Analyzers" Version="3.3.3">
+            <PrivateAssets>all</PrivateAssets>
+            <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
+        </PackageReference>
+    </ItemGroup>
+
+</Project>

X10D.Tests/src/Text/CharTests.cs

@@ -1,4 +1,5 @@
-using Microsoft.VisualStudio.TestTools.UnitTesting;
+using System.Text;
+using Microsoft.VisualStudio.TestTools.UnitTesting;
 using X10D.Text;
 
 namespace X10D.Tests.Text;
@@ -6,6 +7,25 @@ namespace X10D.Tests.Text;
 [TestClass]
 public class CharTests
 {
+    [TestMethod]
+    public void IsEmoji_ShouldReturnTrue_GivenBasicEmoji()
+    {
+        Assert.IsTrue('✂'.IsEmoji());
+        Assert.IsTrue('✅'.IsEmoji());
+        Assert.IsTrue('❎'.IsEmoji());
+        Assert.IsTrue('➕'.IsEmoji());
+        Assert.IsTrue('➖'.IsEmoji());
+    }
+
+    [TestMethod]
+    public void IsEmoji_ShouldReturnFalse_GivenNonEmoji()
+    {
+        for (var letter = 'A'; letter <= 'Z'; letter++)
+        {
+            Assert.IsFalse(letter.IsEmoji());
+        }
+    }
+
     [TestMethod]
     public void RepeatShouldBeCorrect()
     {

X10D.Tests/src/Text/RuneTests.cs

@@ -8,6 +8,25 @@ namespace X10D.Tests.Text;
 [TestClass]
 public class RuneTests
 {
+    [TestMethod]
+    public void IsEmoji_ShouldReturnTrue_GivenBasicEmoji()
+    {
+        Assert.IsTrue(new Rune('✂').IsEmoji());
+        Assert.IsTrue(new Rune('✅').IsEmoji());
+        Assert.IsTrue(new Rune('❎').IsEmoji());
+        Assert.IsTrue(new Rune('➕').IsEmoji());
+        Assert.IsTrue(new Rune('➖').IsEmoji());
+    }
+
+    [TestMethod]
+    public void IsEmoji_ShouldReturnFalse_GivenNonEmoji()
+    {
+        for (var letter = 'A'; letter <= 'Z'; letter++)
+        {
+            Assert.IsFalse(new Rune(letter).IsEmoji());
+        }
+    }
+
     [TestMethod]
     public void RepeatShouldBeCorrect()
     {

X10D.Tests/src/Text/StringTests.cs

@@ -1,4 +1,4 @@
-using System.Text;
+using System.Text;
 #if NET5_0_OR_GREATER
 using System.Text.Json.Serialization;
 #endif
@@ -199,6 +199,44 @@ public class StringTests
         Assert.ThrowsException<ArgumentNullException>(() => "Hello World".GetBytes(null!));
     }
 
+    [TestMethod]
+    public void IsEmoji_ShouldReturnTrue_GivenBasicEmoji()
+    {
+        Assert.IsTrue("😀".IsEmoji());
+        Assert.IsTrue("🤓".IsEmoji());
+        Assert.IsTrue("🟦".IsEmoji());
+        Assert.IsTrue("🟧".IsEmoji());
+        Assert.IsTrue("🟨".IsEmoji());
+        Assert.IsTrue("🟩".IsEmoji());
+        Assert.IsTrue("🟪".IsEmoji());
+        Assert.IsTrue("🟫".IsEmoji());
+        Assert.IsTrue("📱".IsEmoji());
+        Assert.IsTrue("🎨".IsEmoji());
+    }
+
+    [TestMethod]
+    public void IsEmoji_ShouldReturnTrue_GivenMultiByteEmoji()
+    {
+        string[] regionalIndicatorCodes = Enumerable.Range(0, 26)
+            .Select(i => Encoding.Unicode.GetString(new byte[] {0x3C, 0xD8, (byte)(0xE6 + i), 0xDD}))
+            .ToArray();
+
+        for (var i = 0; i < 26; i++)
+        for (var j = 0; j < 26; j++)
+        {
+            string flag = (regionalIndicatorCodes[i] + regionalIndicatorCodes[j]);
+            Assert.IsTrue(flag.IsEmoji());
+        }
+    }
+
+    [TestMethod]
+    public void IsEmoji_ShouldReturnFalse_GivenNonEmoji()
+    {
+        Assert.IsFalse("Hello World".IsEmoji());
+        Assert.IsFalse("Hello".IsEmoji());
+        Assert.IsFalse("World".IsEmoji());
+    }
+
     [TestMethod]
     public void IsLower_ShouldReturnTrue_GivenLowercaseString()
     {

X10D/X10D.csproj

@@ -78,4 +78,10 @@
         </EmbeddedResource>
     </ItemGroup>
 
+    <ItemGroup>
+        <ProjectReference Include="..\X10D.SourceGenerator\X10D.SourceGenerator.csproj"
+                          OutputItemType="Analyzer"
+                          ReferenceOutputAssembly="false"/>
+    </ItemGroup>
+
 </Project>

X10D/src/Text/CharExtensions.cs

@@ -8,6 +8,22 @@ namespace X10D.Text;
 /// </summary>
 public static class CharExtensions
 {
+    /// <summary>
+    ///     Returns a value indicating whether this character constitutes an emoji.
+    /// </summary>
+    /// <param name="value">The character to check.</param>
+    /// <returns><see langword="true" /> if this character is an emoji; otherwise, <see langword="false" />.</returns>
+    [Pure]
+#if NETSTANDARD2_1
+    [MethodImpl(MethodImplOptions.AggressiveInlining)]
+#else
+    [MethodImpl(MethodImplOptions.AggressiveInlining | MethodImplOptions.AggressiveOptimization)]
+#endif
+    public static bool IsEmoji(this char value)
+    {
+        return value.ToString().IsEmoji();
+    }
+
     /// <summary>
     ///     Returns a string composed of the current character repeated a specified number of times.
     /// </summary>

X10D/src/Text/RuneExtensions.cs

@@ -10,6 +10,18 @@ namespace X10D.Text;
 /// </summary>
 public static class RuneExtensions
 {
+    /// <summary>
+    ///     Returns a value indicating whether this rune constitutes an emoji.
+    /// </summary>
+    /// <param name="value">The rune to check.</param>
+    /// <returns><see langword="true" /> if this rune is an emoji; otherwise, <see langword="false" />.</returns>
+    [Pure]
+    [MethodImpl(MethodImplOptions.AggressiveInlining | MethodImplOptions.AggressiveOptimization)]
+    public static bool IsEmoji(this Rune value)
+    {
+        return value.ToString().IsEmoji();
+    }
+
     /// <summary>
     ///     Returns a string composed of the current rune repeated a specified number of times.
     /// </summary>

X10D/src/Text/StringExtensions.cs

@@ -295,6 +295,31 @@ public static class StringExtensions
         return encoding.GetBytes(value);
     }
 
+    /// <summary>
+    ///     Returns a value indicating whether this string constitutes an emoji.
+    /// </summary>
+    /// <param name="value">The input string.</param>
+    /// <returns><see langword="true" /> if this string is an emoji; otherwise, <see langword="false" />.</returns>
+    [Pure]
+#if NETSTANDARD2_1
+    [MethodImpl(MethodImplOptions.AggressiveInlining)]
+#else
+    [MethodImpl(MethodImplOptions.AggressiveInlining | MethodImplOptions.AggressiveOptimization)]
+#endif
+    public static bool IsEmoji(this string value)
+    {
+#if NET6_0_OR_GREATER
+        ArgumentNullException.ThrowIfNull(value);
+#else
+        if (value is null)
+        {
+            throw new ArgumentNullException(nameof(value));
+        }
+#endif
+
+        return EmojiRegex.Value.IsMatch(value);
+    }
+
     /// <summary>
     ///     Determines if all alpha characters in this string are considered lowercase.
     /// </summary>