[browser][non-icu] HybridGlobalization checking for prefix/suffix (#…

…85093)

* Implementation.

* HG does not belong to legacy code.

* No need to create new instance when existing one is exported.

* TextEncoder's behavior varies between hosts.

* Nit

* Cutting prevents us from using IgnoreSymbols.

* Fixed asserts.

* Fix.

* Match platform with behavior.

* Missing changes to prev commit.

docs/design/features/hybrid-globalization.md

@@ -181,3 +181,20 @@ hiraganaBig.localeCompare(katakanaSmall, "en-US", { sensitivity: "base" }) // 0;
 `IgnoreKanaType | IgnoreWidth | IgnoreSymbols | IgnoreNonSpace`
 
 `IgnoreKanaType | IgnoreWidth | IgnoreSymbols | IgnoreNonSpace | IgnoreCase`
+
+
+**String starts with / ends with**
+
+Affected public APIs:
+- CompareInfo.IsPrefix
+- CompareInfo.IsSuffix
+- String.StartsWith
+- String.EndsWith
+
+Web API does not expose locale-sensitive endsWith/startsWith function. As a workaround, both strings get normalized and weightless characters are removed. Resulting strings are cut to the same length and comparison is performed. This approach, beyond having the same compare option limitations as described under **String comparison**, has additional limitations connected with the workaround used. Because we are normalizing strings to be able to cut them, we cannot calculate the match length on the original strings. Methods that calculate this information throw PlatformNotSupported exception:
+
+- [CompareInfo.IsPrefix](https://learn.microsoft.com/en-us/dotnet/api/system.globalization.compareinfo.isprefix?view=net-8.0#system-globalization-compareinfo-isprefix(system-readonlyspan((system-char))-system-readonlyspan((system-char))-system-globalization-compareoptions-system-int32@))
+- [CompareInfo.IsSuffix](https://learn.microsoft.com/en-us/dotnet/api/system.globalization.compareinfo.issuffix?view=net-8.0#system-globalization-compareinfo-issuffix(system-readonlyspan((system-char))-system-readonlyspan((system-char))-system-globalization-compareoptions-system-int32@))
+
+- `IgnoreSymbols`
+Only comparisons that do not skip character types are allowed. E.g. `IgnoreSymbols` skips symbol-chars in comparison/indexing. All `CompareOptions` combinations that include `IgnoreSymbols` throw `PlatformNotSupportedException`.

src/libraries/Common/src/Interop/Browser/Interop.CompareInfo.cs

@@ -9,5 +9,11 @@ internal static unsafe partial class JsGlobalization
     {
         [MethodImplAttribute(MethodImplOptions.InternalCall)]
         internal static extern unsafe int CompareString(out string exceptionMessage, in string culture, char* str1, int str1Len, char* str2, int str2Len, global::System.Globalization.CompareOptions options);
+
+        [MethodImplAttribute(MethodImplOptions.InternalCall)]
+        internal static extern unsafe bool StartsWith(out string exceptionMessage, in string culture, char* str1, int str1Len, char* str2, int str2Len, global::System.Globalization.CompareOptions options);
+
+        [MethodImplAttribute(MethodImplOptions.InternalCall)]
+        internal static extern unsafe bool EndsWith(out string exceptionMessage, in string culture, char* str1, int str1Len, char* str2, int str2Len, global::System.Globalization.CompareOptions options);
     }
 }

src/libraries/System.Globalization/tests/CompareInfo/CompareInfoTests.IsPrefix.cs

@@ -7,14 +7,8 @@
 
 namespace System.Globalization.Tests
 {
-    public class CompareInfoIsPrefixTests
+    public class CompareInfoIsPrefixTests : CompareInfoTestsBase
     {
-        private static CompareInfo s_invariantCompare = CultureInfo.InvariantCulture.CompareInfo;
-        private static CompareInfo s_germanCompare = new CultureInfo("de-DE").CompareInfo;
-        private static CompareInfo s_hungarianCompare = new CultureInfo("hu-HU").CompareInfo;
-        private static CompareInfo s_turkishCompare = new CultureInfo("tr-TR").CompareInfo;
-        private static CompareInfo s_frenchCompare = new CultureInfo("fr-FR").CompareInfo;
-
         public static IEnumerable<object[]> IsPrefix_TestData()
         {
             // Empty strings
@@ -31,7 +25,8 @@ public static IEnumerable<object[]> IsPrefix_TestData()
             yield return new object[] { s_invariantCompare, "dzsdzsfoobar", "ddzsf", CompareOptions.Ordinal, false, 0 };
             yield return new object[] { s_hungarianCompare, "dzsdzsfoobar", "ddzsf", CompareOptions.Ordinal, false, 0 };
             yield return new object[] { s_invariantCompare, "dz", "d", CompareOptions.None, true, 1 };
-            yield return new object[] { s_hungarianCompare, "dz", "d", CompareOptions.None, false, 0 };
+            if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+                yield return new object[] { s_hungarianCompare, "dz", "d", CompareOptions.None, false, 0 };
             yield return new object[] { s_hungarianCompare, "dz", "d", CompareOptions.Ordinal, true, 1 };
 
             // Turkish
@@ -56,7 +51,7 @@ public static IEnumerable<object[]> IsPrefix_TestData()
             yield return new object[] { s_invariantCompare, "\u00C0nimal", "a\u0300", CompareOptions.Ordinal, false, 0 };
             yield return new object[] { s_invariantCompare, "\u00C0nimal", "a\u0300", CompareOptions.OrdinalIgnoreCase, false, 0 };
             yield return new object[] { s_invariantCompare, "FooBar", "Foo\u0400Bar", CompareOptions.Ordinal, false, 0 };
-            yield return new object[] { s_invariantCompare, "FooBA\u0300R", "FooB\u00C0R", CompareOptions.IgnoreNonSpace, true, 7 };
+            yield return new object[] { s_invariantCompare, "FooBA\u0300R", "FooB\u00C0R", supportedIgnoreNonSpaceOption, true, 7 };
             yield return new object[] { s_invariantCompare, "o\u0308", "o", CompareOptions.None, false, 0 };
             yield return new object[] { s_invariantCompare, "o\u0308", "o", CompareOptions.Ordinal, true, 1 };
             yield return new object[] { s_invariantCompare, "o\u0000\u0308", "o", CompareOptions.None, true, 1 };
@@ -76,42 +71,57 @@ public static IEnumerable<object[]> IsPrefix_TestData()
             yield return new object[] { s_invariantCompare, "\uD800\uD800", "\uD800\uD800", CompareOptions.None, true, 2 };
 
             // Ignore symbols
-            yield return new object[] { s_invariantCompare, "Test's can be interesting", "Tests", CompareOptions.IgnoreSymbols, true, 6 };
-            yield return new object[] { s_invariantCompare, "Test's can be interesting", "Tests", CompareOptions.None, false, 0 };
+            if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+            {
+                yield return new object[] { s_invariantCompare, "Test's can be interesting", "Tests", CompareOptions.IgnoreSymbols, true, 6 };
+                yield return new object[] { s_invariantCompare, "Test's can be interesting", "Tests", CompareOptions.None, false, 0 };
+            }
 
             // Platform differences
-            bool useNls = PlatformDetection.IsNlsGlobalization;
-            if (useNls)
+            // in HybridGlobalization on Browser we use TextEncoder that is not supported for v8 and the manual decoding works like NLS
+            bool behavesLikeNls = PlatformDetection.IsNlsGlobalization || 
+                (PlatformDetection.IsHybridGlobalizationOnBrowser && !PlatformDetection.IsBrowserDomSupportedOrNodeJS);
+            if (behavesLikeNls)
             {
-                yield return new object[] { s_hungarianCompare, "dzsdzsfoobar", "ddzsf", CompareOptions.None, true, 7 };
-                yield return new object[] { s_invariantCompare, "''Tests", "Tests", CompareOptions.IgnoreSymbols, true, 7 };
-                yield return new object[] { s_frenchCompare, "\u0153", "oe", CompareOptions.None, true, 1 };
+                if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+                {
+                    yield return new object[] { s_hungarianCompare, "dzsdzsfoobar", "ddzsf", CompareOptions.None, true, 7 };
+                    yield return new object[] { s_invariantCompare, "''Tests", "Tests", CompareOptions.IgnoreSymbols, true, 7 };
+                    yield return new object[] { s_frenchCompare, "\u0153", "oe", CompareOptions.None, true, 1 };
+                }
                 yield return new object[] { s_invariantCompare, "\uD800\uDC00", "\uD800", CompareOptions.None, true, 1 };
                 yield return new object[] { s_invariantCompare, "\uD800\uDC00", "\uD800", CompareOptions.IgnoreCase, true, 1 };
             }
             else
             {
                 yield return new object[] { s_hungarianCompare, "dzsdzsfoobar", "ddzsf", CompareOptions.None, false, 0 };
-                yield return new object[] { s_invariantCompare, "''Tests", "Tests", CompareOptions.IgnoreSymbols, false, 0 };
+                if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+                    yield return new object[] { s_invariantCompare, "''Tests", "Tests", CompareOptions.IgnoreSymbols, false, 0 };
                 yield return new object[] { s_frenchCompare, "\u0153", "oe", CompareOptions.None, false, 0 };
                 yield return new object[] { s_invariantCompare, "\uD800\uDC00", "\uD800", CompareOptions.None, false, 0 };
                 yield return new object[] { s_invariantCompare, "\uD800\uDC00", "\uD800", CompareOptions.IgnoreCase, false, 0 };
             }
 
             // ICU bugs
             // UInt16 overflow: https://unicode-org.atlassian.net/browse/ICU-20832 fixed in https://github.com/unicode-org/icu/pull/840 (ICU 65)
-            if (useNls || PlatformDetection.ICUVersion.Major >= 65)
+            if (PlatformDetection.IsNlsGlobalization || PlatformDetection.ICUVersion.Major >= 65)
             {
                 yield return new object[] { s_frenchCompare, "b", new string('a', UInt16.MaxValue + 1), CompareOptions.None, false, 0 };
             }
 
             // Prefixes where matched length does not equal value string length
-            yield return new object[] { s_invariantCompare, "dzxyz", "\u01F3", CompareOptions.IgnoreNonSpace, true, 2 };
-            yield return new object[] { s_invariantCompare, "\u01F3xyz", "dz", CompareOptions.IgnoreNonSpace, true, 1 };
-            yield return new object[] { s_germanCompare, "Strasse xyz", "stra\u00DFe", CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, true, 7 };
-            yield return new object[] { s_germanCompare, "Strasse xyz", "xtra\u00DFe", CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, false, 0 };
-            yield return new object[] { s_germanCompare, "stra\u00DFe xyz", "Strasse", CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, true, 6 };
-            yield return new object[] { s_germanCompare, "stra\u00DFe xyz", "Xtrasse", CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, false, 0 };
+            if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+            {
+                yield return new object[] { s_invariantCompare, "dzxyz", "\u01F3", supportedIgnoreNonSpaceOption, true, 2 };
+                yield return new object[] { s_invariantCompare, "\u01F3xyz", "dz", supportedIgnoreNonSpaceOption, true, 1 };
+            }
+            if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+            {
+                yield return new object[] { s_germanCompare, "Strasse xyz", "stra\u00DFe", supportedIgnoreCaseIgnoreNonSpaceOptions, true, 7 };
+                yield return new object[] { s_germanCompare, "stra\u00DFe xyz", "Strasse", supportedIgnoreCaseIgnoreNonSpaceOptions, true, 6 };
+            }
+            yield return new object[] { s_germanCompare, "Strasse xyz", "xtra\u00DFe", supportedIgnoreCaseIgnoreNonSpaceOptions, false, 0 };
+            yield return new object[] { s_germanCompare, "stra\u00DFe xyz", "Xtrasse", supportedIgnoreCaseIgnoreNonSpaceOptions, false, 0 };
         }
 
         [Theory]
@@ -140,8 +150,11 @@ public void IsPrefix(CompareInfo compareInfo, string source, string value, Compa
             valueBoundedMemory.MakeReadonly();
 
             Assert.Equal(expected, compareInfo.IsPrefix(sourceBoundedMemory.Span, valueBoundedMemory.Span, options));
-            Assert.Equal(expected, compareInfo.IsPrefix(sourceBoundedMemory.Span, valueBoundedMemory.Span, options, out int actualMatchLength));
-            Assert.Equal(expectedMatchLength, actualMatchLength);
+            if (!PlatformDetection.IsHybridGlobalizationOnBrowser)
+            {
+                Assert.Equal(expected, compareInfo.IsPrefix(sourceBoundedMemory.Span, valueBoundedMemory.Span, options, out int actualMatchLength));
+                Assert.Equal(expectedMatchLength, actualMatchLength);
+            }
         }
 
         [Fact]
@@ -150,7 +163,7 @@ public void IsPrefix_UnassignedUnicode()
             bool result = PlatformDetection.IsNlsGlobalization ? true : false;
             int expectedMatchLength = (result) ? 6 : 0;
             IsPrefix(s_invariantCompare, "FooBar", "Foo\uFFFFBar", CompareOptions.None, result, expectedMatchLength);
-            IsPrefix(s_invariantCompare, "FooBar", "Foo\uFFFFBar", CompareOptions.IgnoreNonSpace, result, expectedMatchLength);
+            IsPrefix(s_invariantCompare, "FooBar", "Foo\uFFFFBar", supportedIgnoreNonSpaceOption, result, expectedMatchLength);
         }
 
         [Fact]