perf(java): optimize Java parser performance

chapi-ast-java/src/main/antlr/JavaLexer.g4

@@ -141,6 +141,9 @@ SEMI   : ';';
 COMMA  : ',';
 DOT    : '.';
 
+// Java 22+ (preview, continued into later releases): unnamed variables / patterns
+UNDERSCORE : '_';
+
 // Operators
 
 ASSIGN   : '=';

chapi-ast-java/src/main/antlr/JavaParser.g4

@@ -255,7 +255,13 @@ variableDeclarator
     ;
 
 variableDeclaratorId
-    : identifier ('[' ']')*
+    : (identifier | UNDERSCORE) ('[' ']')*
+    ;
+
+// Java 22+ (preview, continued into later releases): unnamed variables / patterns
+variableName
+    : identifier
+    | UNDERSCORE
     ;
 
 variableInitializer
@@ -310,7 +316,7 @@ lambdaLVTIList
     ;
 
 lambdaLVTIParameter
-    : variableModifier* VAR identifier
+    : variableModifier* VAR (identifier | UNDERSCORE)
     ;
 
 qualifiedName
@@ -474,7 +480,7 @@ blockStatement
     ;
 
 localVariableDeclaration
-    : variableModifier* (VAR identifier '=' expression | typeType variableDeclarators)
+    : variableModifier* (VAR (identifier | UNDERSCORE) '=' expression | typeType variableDeclarators)
     ;
 
 identifier
@@ -539,7 +545,7 @@ statement
     ;
 
 catchClause
-    : CATCH '(' variableModifier* catchType identifier ')' block
+    : CATCH '(' variableModifier* catchType (identifier | UNDERSCORE) ')' block
     ;
 
 catchType
@@ -559,7 +565,7 @@ resources
     ;
 
 resource
-    : variableModifier* (classOrInterfaceType variableDeclaratorId | VAR identifier) '=' expression
+    : variableModifier* (classOrInterfaceType variableDeclaratorId | VAR (identifier | UNDERSCORE)) '=' expression
     | qualifiedName
     ;
 
@@ -574,7 +580,7 @@ switchLabel
     : CASE (
         constantExpression = expression
         | enumConstantName = IDENTIFIER
-        | typeType varName = identifier
+        | typeType varName = variableName
     )
     | DEFAULT
     ;
@@ -697,8 +703,9 @@ lambdaExpression
 
 lambdaParameters
     : identifier
+    | UNDERSCORE
     | '(' formalParameterList? ')'
-    | '(' identifier (',' identifier)* ')'
+    | '(' (identifier | UNDERSCORE) (',' (identifier | UNDERSCORE))* ')'
     | '(' lambdaLVTIList? ')'
     ;
 

chapi-ast-java/src/main/kotlin/chapi/ast/javaast/JavaAnalyser.kt

@@ -11,6 +11,11 @@ import org.antlr.v4.runtime.CommonTokenStream
 import org.antlr.v4.runtime.tree.ParseTreeWalker
 
 open class JavaAnalyser : TwoStepAnalyser() {
+    companion object {
+        // Reuse ParseTreeWalker instance to reduce object creation
+        private val walker = ParseTreeWalker()
+    }
+
     override fun analysis(code: String, filePath: String, parseMode: ParseMode): CodeContainer {
         return when (parseMode) {
             ParseMode.Basic -> identBasicInfo(code, filePath)
@@ -27,24 +32,29 @@ open class JavaAnalyser : TwoStepAnalyser() {
         val context = this.parse(str).compilationUnit()
         val listener = JavaFullIdentListener(fileName, classes)
 
-        ParseTreeWalker().walk(listener, context)
+        walker.walk(listener, context)
         return listener.getNodeInfo()
     }
 
     open fun identBasicInfo(str: String, fileName: String): CodeContainer {
         val context = this.parse(str).compilationUnit()
         val listener = JavaBasicIdentListener(fileName)
 
-        ParseTreeWalker().walk(listener, context)
+        walker.walk(listener, context)
 
         return listener.getNodeInfo()
     }
 
     open fun parse(str: String): JavaParser {
         val fromString = CharStreams.fromString(str)
         val lexer = JavaLexer(fromString)
+        // Remove default error listeners to reduce I/O overhead
+        lexer.removeErrorListeners()
+
         val tokenStream = CommonTokenStream(lexer)
         val parser = JavaParser(tokenStream)
+        // Remove default error listeners
+        parser.removeErrorListeners()
         return parser
     }
 }

chapi-ast-java/src/main/kotlin/chapi/ast/javaast/JavaBasicIdentListener.kt

@@ -12,8 +12,11 @@ open class JavaBasicIdentListener(fileName: String) : JavaAstListener() {
             Language = "java",
             Kind = ContainerKind.SOURCE_FILE
         )
-    private var classNodes: List<CodeDataStruct> = listOf()
-    private var imports: List<CodeImport> = listOf()
+    private var classNodes: MutableList<CodeDataStruct> = mutableListOf()
+    private var imports: MutableList<CodeImport> = mutableListOf()
+    // Index for fast import lookup by class name
+    // Using MutableList to handle import collisions
+    private var importsByClassName: MutableMap<String, MutableList<CodeImport>> = mutableMapOf()
 
     private var currentNode = CodeDataStruct()
     private var currentFunction = CodeFunction(IsConstructor = false)
@@ -58,8 +61,14 @@ open class JavaBasicIdentListener(fileName: String) : JavaAstListener() {
 
         codeImport.PathSegments = fullSource.split(".")
 
-        imports += codeImport
+        imports.add(codeImport)
         codeContainer.Imports += codeImport
+
+        // Build import index for non-wildcard imports only
+        if (!isWildcard) {
+            val className = fullSource.substringAfterLast('.')
+            importsByClassName.getOrPut(className) { mutableListOf() }.add(codeImport)
+        }
     }
 
     override fun enterPackageDeclaration(ctx: JavaParser.PackageDeclarationContext?) {
@@ -83,27 +92,39 @@ open class JavaBasicIdentListener(fileName: String) : JavaAstListener() {
     }
 
     override fun exitClassBodyDeclaration(ctx: JavaParser.ClassBodyDeclarationContext?) {
+        // Class members are handled in their own enter/exit callbacks.
+        // Finalizing the class here would duplicate nodes (one per member) and miss empty classes.
+    }
+
+    override fun exitClassDeclaration(ctx: JavaParser.ClassDeclarationContext?) {
         hasEnterClass = false
         if (currentNode.NodeName != "") {
-            classNodes += currentNode
+            classNodes.add(currentNode)
         }
         currentNode = CodeDataStruct()
     }
 
     open fun buildImplements(ctx: JavaParser.ClassDeclarationContext): List<String> {
         return ctx.typeList()
             .map { it.text }
-            .filter { imports.containsType(it) }
+            .filter { containsType(it) }
     }
 
-    private fun <T> List<T>.containsType(it: String?): Boolean {
-        return imports.filter { imp -> imp.Source.endsWith(".$it") }.toTypedArray().isNotEmpty()
+    private fun containsType(typeName: String?): Boolean {
+        if (typeName == null) return false
+        // Fast lookup using index
+        val importsByName = importsByClassName[typeName]
+        if (importsByName != null && importsByName.isNotEmpty()) {
+            return true
+        }
+        // Fallback to linear search for wildcard imports
+        return imports.any { imp -> imp.Source.endsWith(".$typeName") }
     }
 
     open fun buildImplements(ctx: JavaParser.EnumDeclarationContext): List<String> {
         val typeText = ctx.typeList().text
         return when {
-            imports.containsType(typeText) -> listOf(typeText)
+            containsType(typeText) -> listOf(typeText)
             else -> listOf()
         }
     }
@@ -120,7 +141,7 @@ open class JavaBasicIdentListener(fileName: String) : JavaAstListener() {
     override fun exitInterfaceDeclaration(ctx: JavaParser.InterfaceDeclarationContext?) {
         hasEnterClass = false
         if (currentNode.NodeName != "") {
-            classNodes += currentNode
+            classNodes.add(currentNode)
         }
     }
 
@@ -236,14 +257,13 @@ open class JavaBasicIdentListener(fileName: String) : JavaAstListener() {
     override fun exitEnumBodyDeclarations(ctx: JavaParser.EnumBodyDeclarationsContext?) {
         hasEnterClass = false
         if (currentNode.NodeName != "") {
-            classNodes += currentNode
+            classNodes.add(currentNode)
         }
         currentNode = CodeDataStruct()
     }
 
     fun getNodeInfo(): CodeContainer {
-        codeContainer.DataStructures = classNodes
+        codeContainer.DataStructures = classNodes.toList()
         return codeContainer
     }
-}
-
+}