Expand Locale parser to support three-character language codes (#6124)

Author: baronfel

src/NuGet.Core/NuGet.Packaging/ContentModel/ManagedCodeConventions.cs

@@ -160,7 +160,7 @@ private static object CodeLanguage_Parser(ReadOnlyMemory<char> name, PatternTabl
         /// If matchOnly is true, then an empty string may be returned as a performance optimization.
         /// If matchOnly is false, the parsed result will be returned.
         /// </summary>
-        private static object Locale_Parser(ReadOnlyMemory<char> name, PatternTable table, bool matchOnly)
+        internal static object Locale_Parser(ReadOnlyMemory<char> name, PatternTable table, bool matchOnly)
         {
             if (table != null)
             {
@@ -171,22 +171,40 @@ private static object Locale_Parser(ReadOnlyMemory<char> name, PatternTable tabl
                 }
             }
 
-            if (name.Length == 2)
+            // We use a heuristic here for common locale codes. Locale codes are often
+            // * two characters for the language: en, es, fr, de
+            // * three characters for the language: agq
+            if (name.Length == 2 || name.Length == 3)
             {
                 if (matchOnly)
                 {
                     return string.Empty;
                 }
                 return name.ToString();
             }
-            else if (name.Length >= 4 && name.Span[2] == '-')
+
+            // * a language portion that is two or three characters followed by a '-' and a country code
+            else if (name.Length >= 4 && name.Span[2] == '-') // e.g. en-US
             {
                 if (matchOnly)
                 {
                     return string.Empty;
                 }
                 return name.ToString();
             }
+            else if (name.Length >= 5 && name.Span[3] == '-') // e.g agq-CM
+            {
+                if (matchOnly)
+                {
+                    return string.Empty;
+                }
+                return name.ToString();
+            }
+
+            // there are other variations, but this heuristic doesn't cover them all. A future-proof implementation would make
+            // use of the .NET CultureInfo APIs to compare the locale against the underlying system ICU database. This would
+            // be correct, but potentially more expensive because the CultureInfo APIs are lazily-loaded and throw if an
+            // invalid/unknown locale is used.
 
             return null;
         }

test/NuGet.Core.Tests/NuGet.Packaging.Test/ContentModelTests/ContentModelResourceTests.cs

@@ -0,0 +1,28 @@
+// Copyright (c) .NET Foundation. All rights reserved.
+// Licensed under the Apache License, Version 2.0. See License.txt in the project root for license information.
+
+using System;
+using System.Collections.Generic;
+using System.Globalization;
+using System.Linq;
+using Xunit;
+
+namespace NuGet.Client.Test
+{
+    public class ContentModelResourceTests
+    {
+        [MemberData(nameof(AllCultures))]
+        [Theory]
+        public void CanParseEverySystemKnownCultureResource(CultureInfo culture)
+        {
+            var result = ManagedCodeConventions.Locale_Parser(culture.Name.AsMemory(), null, false);
+            Assert.Equal(culture.Name, result as string);
+        }
+
+        public static IEnumerable<object[]> AllCultures()
+        {
+            return CultureInfo.GetCultures(CultureTypes.AllCultures).Where(c => !string.IsNullOrEmpty(c.Name)).Select(culture => new[] { culture });
+        }
+    }
+}
+