Add searching test cases for multi-byte encoding

4 files changed, 61 insertions, 11 deletions

diff --git a/test/unit/UnitTester.vcxproj b/test/unit/UnitTester.vcxproj
index 358302321..2e130ec3d 100644
--- a/test/unit/UnitTester.vcxproj
+++ b/test/unit/UnitTester.vcxproj
@@ -150,6 +150,8 @@
     </Link>

   </ItemDefinitionGroup>

   <ItemGroup>

+    <ClCompile Include="..\..\src\CaseConvert.cxx" />

+    <ClCompile Include="..\..\src\CaseFolder.cxx" />

     <ClCompile Include="..\..\src\CellBuffer.cxx" />

     <ClCompile Include="..\..\src\CharacterCategoryMap.cxx" />

     <ClCompile Include="..\..\src\CharClassify.cxx" />

diff --git a/test/unit/makefile b/test/unit/makefile
index ed06e5f93..1cde6319e 100644
--- a/test/unit/makefile
+++ b/test/unit/makefile
@@ -49,6 +49,8 @@ CXXFLAGS += -Wall -Wextra
 TESTSRC=test*.cxx
 # Files being tested from scintilla/src directory
 TESTEDSRC=\
+ ../../src/CaseConvert.cxx \
+ ../../src/CaseFolder.cxx \
  ../../src/CellBuffer.cxx \
  ../../src/CharacterCategoryMap.cxx \
  ../../src/CharClassify.cxx \
diff --git a/test/unit/test.mak b/test/unit/test.mak
index e9f71619c..116c6c0d5 100644
--- a/test/unit/test.mak
+++ b/test/unit/test.mak
@@ -12,6 +12,8 @@ CXXFLAGS = /EHsc /std:c++17 /D_HAS_AUTO_PTR_ETC=1 /wd 4805 $(INCLUDEDIRS)
 TESTSRC=test*.cxx
 # Files being tested from scintilla/src directory
 TESTEDSRC=\
+ ../../src/CaseConvert.cxx \
+ ../../src/CaseFolder.cxx \
  ../../src/CellBuffer.cxx \
  ../../src/CharacterCategoryMap.cxx \
  ../../src/CharClassify.cxx \
diff --git a/test/unit/testDocument.cxx b/test/unit/testDocument.cxx
index 983074daf..14d5fc22f 100644
--- a/test/unit/testDocument.cxx
+++ b/test/unit/testDocument.cxx
@@ -36,24 +36,68 @@ using namespace Scintilla::Internal;
 
 // Test Document.
 
+struct DocPlus {
+	Document document;
+
+	DocPlus(std::string_view svInitial, int codePage) : document(DocumentOption::Default) {
+		document.SetDBCSCodePage(codePage);
+		if (codePage == CpUtf8) {
+			document.SetCaseFolder(std::make_unique<CaseFolderUnicode>());
+		} else {
+			// This case folder will not handle many DBCS cases. Scintilla uses platform-specific code for DBCS
+			// case folding which can not easily be inserted in platform-independent tests.
+			std::unique_ptr<CaseFolderTable> pcft = std::make_unique<CaseFolderTable>();
+			pcft->StandardASCII();
+			document.SetCaseFolder(std::move(pcft));
+		}
+		document.InsertString(0, svInitial.data(), svInitial.length());
+	}
+
+};
+
 TEST_CASE("Document") {
 
 	const char sText[] = "Scintilla";
 	const Sci::Position sLength = static_cast<Sci::Position>(strlen(sText));
 
-	Document doc(DocumentOption::Default);
-
 	SECTION("InsertOneLine") {
-		const Sci::Position length = doc.InsertString(0, sText, sLength);
-		REQUIRE(sLength == doc.Length());
+		DocPlus doc("", 0);
+		const Sci::Position length = doc.document.InsertString(0, sText, sLength);
+		REQUIRE(sLength == doc.document.Length());
 		REQUIRE(length == sLength);
-		REQUIRE(1 == doc.LinesTotal());
-		REQUIRE(0 == doc.LineStart(0));
-		REQUIRE(0 == doc.LineFromPosition(0));
-		REQUIRE(sLength == doc.LineStart(1));
-		REQUIRE(0 == doc.LineFromPosition(static_cast<int>(sLength)));
-		REQUIRE(doc.CanUndo());
-		REQUIRE(!doc.CanRedo());
+		REQUIRE(1 == doc.document.LinesTotal());
+		REQUIRE(0 == doc.document.LineStart(0));
+		REQUIRE(0 == doc.document.LineFromPosition(0));
+		REQUIRE(sLength == doc.document.LineStart(1));
+		REQUIRE(0 == doc.document.LineFromPosition(static_cast<int>(sLength)));
+		REQUIRE(doc.document.CanUndo());
+		REQUIRE(!doc.document.CanRedo());
+	}
+
+	SECTION("SearchInUTF8") {
+		DocPlus doc("ab\xCE\x93" "d", CpUtf8);	// a b gamma d
+		std::string finding = "b";
+		Sci::Position lengthFinding = finding.length();
+		Sci::Position location = doc.document.FindText(0, doc.document.Length(), finding.c_str(), FindOption::MatchCase, &lengthFinding);
+		REQUIRE(location == 1);
+		location = doc.document.FindText(doc.document.Length(), 0, finding.c_str(), FindOption::MatchCase, &lengthFinding);
+		REQUIRE(location == 1);
+	}
+
+	SECTION("SearchInShiftJIS") {
+		// {CJK UNIFIED IDEOGRAPH-9955} is two bytes: {0xE9, 'b'} in Shift-JIS
+		// The 'b' can be incorrectly matched by the search string 'b' when the search
+		// does not iterate the text correctly.
+		DocPlus doc("ab\xe9" "b ", 932);	// a b {CJK UNIFIED IDEOGRAPH-9955} {space}
+		std::string finding = "b";
+		// Search forwards
+		Sci::Position lengthFinding = finding.length();
+		Sci::Position location = doc.document.FindText(0, doc.document.Length(), finding.c_str(), FindOption::MatchCase, &lengthFinding);
+		REQUIRE(location == 1);
+		// Search backwards
+		lengthFinding = finding.length();
+		location = doc.document.FindText(doc.document.Length(), 0, finding.c_str(), FindOption::MatchCase, &lengthFinding);
+		REQUIRE(location == 1);
 	}
 
 }