[iOS][globalization] Fix IndexOf on empty strings on iOS to return -1 (#111898)

Author: matouskozak

src/libraries/System.Runtime/tests/System.Globalization.Tests/CompareInfo/CompareInfoTests.IndexOf.cs

@@ -12,10 +12,17 @@ public class CompareInfoIndexOfTests : CompareInfoTestsBase
     {
         public static IEnumerable<object[]> IndexOf_TestData()
         {
-            // Empty string
+            // Empty string, invariant
             yield return new object[] { s_invariantCompare, "foo", "", 0, 3, CompareOptions.None, 0, 0 };
             yield return new object[] { s_invariantCompare, "foo", "", 2, 1, CompareOptions.None, 2, 0 };
             yield return new object[] { s_invariantCompare, "", "", 0, 0, CompareOptions.None, 0, 0 };
+            yield return new object[] { s_invariantCompare, "", "foo", 0, 0, CompareOptions.None, -1, 0 };
+
+            // Empty string, using non-invariant (s_germanCompare) CompareInfo to test the ICU path
+            yield return new object[] { s_germanCompare, "foo", "", 0, 3, CompareOptions.None, 0, 0 };
+            yield return new object[] { s_germanCompare, "foo", "", 2, 1, CompareOptions.None, 2, 0 };
+            yield return new object[] { s_germanCompare, "", "", 0, 0, CompareOptions.None, 0, 0 };
+            yield return new object[] { s_germanCompare, "", "foo", 0, 0, CompareOptions.None, -1, 0 };
 
             // OrdinalIgnoreCase
             yield return new object[] { s_invariantCompare, "Hello", "l", 0, 5, CompareOptions.OrdinalIgnoreCase, 2, 1 };
@@ -138,8 +145,8 @@ public static IEnumerable<object[]> IndexOf_TestData()
             }
 
             // Inputs where matched length does not equal value string length
-                yield return new object[] { s_germanCompare, "abc Strasse Strasse xyz", "stra\u00DFe", 0, 23, CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, 4, 7 };
-                yield return new object[] { s_germanCompare, "abc stra\u00DFe stra\u00DFe xyz", "Strasse", 0, 21, CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, 4, 6 };
+            yield return new object[] { s_germanCompare, "abc Strasse Strasse xyz", "stra\u00DFe", 0, 23, CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, 4, 7 };
+            yield return new object[] { s_germanCompare, "abc stra\u00DFe stra\u00DFe xyz", "Strasse", 0, 21, CompareOptions.IgnoreCase | CompareOptions.IgnoreNonSpace, 4, 6 };
             if (PlatformDetection.IsNotHybridGlobalizationOnApplePlatform)
             {
                 yield return new object[] { s_invariantCompare, "abcdzxyz", "\u01F3", 0, 8, CompareOptions.IgnoreNonSpace, 3, 2 };
@@ -153,7 +160,7 @@ public static IEnumerable<object[]> IndexOf_Aesc_Ligature_TestData()
         {
             bool useNls = PlatformDetection.IsNlsGlobalization;
             // Searches for the ligature \u00C6
-            string source1 = "Is AE or ae the same as \u00C6 or \u00E6?"; // 3 failures here
+            string source1 = "Is AE or ae the same as \u00C6 or \u00E6?";
             yield return new object[] { s_invariantCompare, source1, "AE", 8, 18, CompareOptions.None, useNls ? 24 : -1, useNls ? 1 : 0};
             yield return new object[] { s_invariantCompare, source1, "ae", 8, 18, CompareOptions.None, 9 , 2};
             yield return new object[] { s_invariantCompare, source1, "\u00C6", 8, 18, CompareOptions.None, 24, 1 };
@@ -171,7 +178,7 @@ public static IEnumerable<object[]> IndexOf_Aesc_Ligature_TestData()
         public static IEnumerable<object[]> IndexOf_U_WithDiaeresis_TestData()
         {
             // Searches for the combining character sequence Latin capital letter U with diaeresis or Latin small letter u with diaeresis.
-            string source = "Is \u0055\u0308 or \u0075\u0308 the same as \u00DC or \u00FC?"; // 7 failures here
+            string source = "Is \u0055\u0308 or \u0075\u0308 the same as \u00DC or \u00FC?";
             yield return new object[] { s_invariantCompare, source, "U\u0308", 8, 18, CompareOptions.None, 24, 1 };
             yield return new object[] { s_invariantCompare, source, "u\u0308", 8, 18, CompareOptions.None, 9, 2 };
             yield return new object[] { s_invariantCompare, source, "\u00DC", 8, 18, CompareOptions.None, 24, 1 };

src/libraries/System.Runtime/tests/System.Runtime.Tests/System/StringTests.cs

@@ -210,6 +210,7 @@ public static void Contains_Char(string s, char value, bool expected)
         [InlineData("Hello", 'e', StringComparison.CurrentCulture, true)]
         [InlineData("Hello", 'E', StringComparison.CurrentCulture, false)]
         [InlineData("", 'H', StringComparison.CurrentCulture, false)]
+        [InlineData("", '\u0301', StringComparison.CurrentCulture, false)] // Using non-ASCII character to test ICU path
         // CurrentCultureIgnoreCase
         [InlineData("Hello", 'H', StringComparison.CurrentCultureIgnoreCase, true)]
         [InlineData("Hello", 'Z', StringComparison.CurrentCultureIgnoreCase, false)]

src/native/libs/System.Globalization.Native/pal_collation.m

@@ -117,6 +117,11 @@ int32_t GlobalizationNative_CompareStringNative(const uint16_t* localeName, int3
     }
 }
 
+/**
+ * Removes zero-width and other weightless characters such as U+200B (Zero Width Space), 
+ * U+200C (Zero Width Non-Joiner), U+200D (Zero Width Joiner), U+FEFF (Zero Width No-Break Space), 
+ * and the NUL character from the specified string.
+ */
 static NSString* RemoveWeightlessCharacters(NSString* source)
 {
     NSError *error = nil;
@@ -143,10 +148,9 @@ static int32_t IsIndexFound(int32_t fromBeginning, int32_t foundLocation, int32_
 
 /*
 Function: IndexOf
-Find detailed explanation how this function works in https://github.com/dotnet/runtime/blob/main/docs/design/features/globalization-hybrid-mode.md
+Find detailed explanation how this function works in https://github.com/dotnet/runtime/blob/main/docs/design/features/globalization-hybrid-mode.md#string-indexing
 */
-Range GlobalizationNative_IndexOfNative(const uint16_t* localeName, int32_t lNameLength, const uint16_t* lpTarget, int32_t cwTargetLength,
-                                        const uint16_t* lpSource, int32_t cwSourceLength, int32_t comparisonOptions, int32_t fromBeginning)
+Range GlobalizationNative_IndexOfNative(const uint16_t* localeName, int32_t lNameLength, const uint16_t* lpTarget, int32_t cwTargetLength, const uint16_t* lpSource, int32_t cwSourceLength, int32_t comparisonOptions, int32_t fromBeginning)
 {
     @autoreleasepool
     {
@@ -158,6 +162,9 @@ Range GlobalizationNative_IndexOfNative(const uint16_t* localeName, int32_t lNam
             return result;
         }
         NSStringCompareOptions options = ConvertFromCompareOptionsToNSStringCompareOptions(comparisonOptions, true);
+        if (!fromBeginning) // LastIndexOf
+            options |= NSBackwardsSearch;
+
         NSString *searchString = [NSString stringWithCharacters: lpTarget length: (NSUInteger)cwTargetLength];
         NSString *searchStrCleaned = RemoveWeightlessCharacters(searchString);
         NSString *sourceString = [NSString stringWithCharacters: lpSource length: (NSUInteger)cwSourceLength];
@@ -168,7 +175,7 @@ Range GlobalizationNative_IndexOfNative(const uint16_t* localeName, int32_t lNam
             searchStrCleaned = ConvertToKatakana(searchStrCleaned);
         }
 
-        if (sourceStrCleaned.length == 0 || searchStrCleaned.length == 0)
+        if (searchStrCleaned.length == 0)
         {
             result.location = fromBeginning ? 0 : (int32_t)sourceString.length;
             return result;
@@ -178,9 +185,6 @@ Range GlobalizationNative_IndexOfNative(const uint16_t* localeName, int32_t lNam
         NSString *searchStrPrecomposed = searchStrCleaned.precomposedStringWithCanonicalMapping;
         NSString *sourceStrPrecomposed = sourceStrCleaned.precomposedStringWithCanonicalMapping;
 
-        // last index
-        if (!fromBeginning)
-            options |= NSBackwardsSearch;
 
         // check if there is a possible match and return -1 if not
         // doesn't matter which normalization form is used here
@@ -233,7 +237,7 @@ Range GlobalizationNative_IndexOfNative(const uint16_t* localeName, int32_t lNam
             result.location = (int32_t)precomposedRange.location;
             result.length = (int32_t)precomposedRange.length;
             if (!(comparisonOptions & IgnoreCase))
-            return result;
+                return result;
         }
 
         // check if sourceString has decomposed form of characters and searchString has precomposed form of characters