.gitignore

@@ -68,3 +68,8 @@ Thumbs.db
 # Project outputs
 # =========================
 test/test_result/
+
+
+# Added by cargo
+
+/target

CMakeLists.txt

@@ -2,6 +2,8 @@ cmake_minimum_required(VERSION 3.20)
 
 project(compiler LANGUAGES C CXX)
 
+find_package(Java REQUIRED COMPONENTS Runtime)
+
 # 统一 C++ 标准（按实验环境可调整）
 set(CMAKE_CXX_STANDARD 17)
 set(CMAKE_CXX_STANDARD_REQUIRED ON)
@@ -39,6 +41,11 @@ endif()
 
 option(COMPILER_PARSE_ONLY "Build only the frontend parser pipeline" OFF)
 
+set(ANTLR4_JAR "${PROJECT_SOURCE_DIR}/third_party/antlr-4.13.2-complete.jar")
+if(NOT EXISTS "${ANTLR4_JAR}")
+  message(FATAL_ERROR "ANTLR jar not found: ${ANTLR4_JAR}")
+endif()
+
 # 使用仓库内 third_party 提供的 ANTLR4 C++ runtime（较新版本）
 # third_party 目录结构以仓库为准：runtime 源码位于 third_party/antlr4-runtime-4.13.2/runtime/src
 set(ANTLR4_RUNTIME_SRC_DIR "${PROJECT_SOURCE_DIR}/third_party/antlr4-runtime-4.13.2/runtime/src")

Cargo.lock

@@ -0,0 +1,7 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 4
+
+[[package]]
+name = "nudt-compiler-cpp"
+version = "0.1.0"

Cargo.toml

@@ -0,0 +1,6 @@
+[package]
+name = "nudt-compiler-cpp"
+version = "0.1.0"
+edition = "2024"
+
+[dependencies]

src/antlr4/SysY.g4

@@ -0,0 +1,228 @@
+grammar SysY;
+
+/*===-------------------------------------------===*/
+/* Lexer rules                                     */
+/*===-------------------------------------------===*/
+
+CONST: 'const';
+INT: 'int';
+FLOAT: 'float';
+VOID: 'void';
+IF: 'if';
+ELSE: 'else';
+WHILE: 'while';
+BREAK: 'break';
+CONTINUE: 'continue';
+RETURN: 'return';
+
+ASSIGN: '=';
+EQ: '==';
+NE: '!=';
+LT: '<';
+GT: '>';
+LE: '<=';
+GE: '>=';
+
+ADD: '+';
+SUB: '-';
+MUL: '*';
+DIV: '/';
+MOD: '%';
+NOT: '!';
+LAND: '&&';
+LOR: '||';
+
+LPAREN: '(';
+RPAREN: ')';
+LBRACE: '{';
+RBRACE: '}';
+LBRACK: '[';
+RBRACK: ']';
+COMMA: ',';
+SEMICOLON: ';';
+
+FLOAT_CONST
+    : DEC_FLOAT_CONST
+    | HEX_FLOAT_CONST
+    ;
+
+INT_CONST
+    : HEX_PREFIX HEX_DIGIT+
+    | '0' [0-7]+
+    | '0'
+    | [1-9] DIGIT*
+    ;
+
+ID: [a-zA-Z_][a-zA-Z_0-9]*;
+
+WS: [ \t\r\n] -> skip;
+LINECOMMENT: '//' ~[\r\n]* -> skip;
+BLOCKCOMMENT: '/*' .*? '*/' -> skip;
+
+fragment DEC_FLOAT_CONST
+    : DIGIT+ '.' DIGIT* EXP_PART?
+    | '.' DIGIT+ EXP_PART?
+    | DIGIT+ EXP_PART
+    ;
+
+fragment HEX_FLOAT_CONST
+    : HEX_PREFIX HEX_DIGIT+ '.' HEX_DIGIT* BIN_EXP_PART
+    | HEX_PREFIX '.' HEX_DIGIT+ BIN_EXP_PART
+    | HEX_PREFIX HEX_DIGIT+ BIN_EXP_PART
+    ;
+
+fragment EXP_PART: [eE] [+-]? DIGIT+;
+fragment BIN_EXP_PART: [pP] [+-]? DIGIT+;
+fragment HEX_PREFIX: '0' [xX];
+fragment HEX_DIGIT: [0-9a-fA-F];
+fragment DIGIT: [0-9];
+
+/*===-------------------------------------------===*/
+/* Syntax rules                                    */
+/*===-------------------------------------------===*/
+
+compUnit
+    : (decl | funcDef)+ EOF
+    ;
+
+decl
+    : constDecl
+    | varDecl
+    ;
+
+constDecl
+    : CONST bType constDef (COMMA constDef)* SEMICOLON
+    ;
+
+varDecl
+    : bType varDef (COMMA varDef)* SEMICOLON
+    ;
+
+bType
+    : INT
+    | FLOAT
+    ;
+
+constDef
+    : ID (LBRACK constExp RBRACK)* ASSIGN constInitVal
+    ;
+
+constInitVal
+    : constExp
+    | LBRACE (constInitVal (COMMA constInitVal)*)? RBRACE
+    ;
+
+varDef
+    : ID (LBRACK constExp RBRACK)* (ASSIGN initVal)?
+    ;
+
+initVal
+    : exp
+    | LBRACE (initVal (COMMA initVal)*)? RBRACE
+    ;
+
+funcDef
+    : funcType ID LPAREN funcFParams? RPAREN block
+    ;
+
+funcType
+    : VOID
+    | INT
+    | FLOAT
+    ;
+
+funcFParams
+    : funcFParam (COMMA funcFParam)*
+    ;
+
+funcFParam
+    : bType ID (LBRACK RBRACK (LBRACK exp RBRACK)*)?
+    ;
+
+block
+    : LBRACE blockItem* RBRACE
+    ;
+
+blockItem
+    : decl
+    | stmt
+    ;
+
+stmt
+    : lVal ASSIGN exp SEMICOLON
+    | exp? SEMICOLON
+    | block
+    | IF LPAREN cond RPAREN stmt (ELSE stmt)?
+    | WHILE LPAREN cond RPAREN stmt
+    | BREAK SEMICOLON
+    | CONTINUE SEMICOLON
+    | RETURN exp? SEMICOLON
+    ;
+
+exp
+    : addExp
+    ;
+
+cond
+    : lOrExp
+    ;
+
+lVal
+    : ID (LBRACK exp RBRACK)*
+    ;
+
+primaryExp
+    : LPAREN exp RPAREN
+    | lVal
+    | number
+    ;
+
+number
+    : INT_CONST
+    | FLOAT_CONST
+    ;
+
+unaryExp
+    : primaryExp
+    | ID LPAREN funcRParams? RPAREN
+    | unaryOp unaryExp
+    ;
+
+unaryOp
+    : ADD
+    | SUB
+    | NOT
+    ;
+
+funcRParams
+    : exp (COMMA exp)*
+    ;
+
+mulExp
+    : unaryExp ((MUL | DIV | MOD) unaryExp)*
+    ;
+
+addExp
+    : mulExp ((ADD | SUB) mulExp)*
+    ;
+
+relExp
+    : addExp ((LT | GT | LE | GE) addExp)*
+    ;
+
+eqExp
+    : relExp ((EQ | NE) relExp)*
+    ;
+
+lAndExp
+    : eqExp (LAND eqExp)*
+    ;
+
+lOrExp
+    : lAndExp (LOR lAndExp)*
+    ;
+
+constExp
+    : addExp
+    ;
+

src/frontend/CMakeLists.txt

@@ -3,15 +3,44 @@ add_library(frontend STATIC
   SyntaxTreePrinter.cpp
 )
 
+set(ANTLR4_GRAMMAR "${PROJECT_SOURCE_DIR}/src/antlr4/SysY.g4")
+set(ANTLR4_GENERATED_FILES
+  "${ANTLR4_GENERATED_DIR}/SysYLexer.cpp"
+  "${ANTLR4_GENERATED_DIR}/SysYLexer.h"
+  "${ANTLR4_GENERATED_DIR}/SysYLexer.interp"
+  "${ANTLR4_GENERATED_DIR}/SysYLexer.tokens"
+  "${ANTLR4_GENERATED_DIR}/SysYParser.cpp"
+  "${ANTLR4_GENERATED_DIR}/SysYParser.h"
+  "${ANTLR4_GENERATED_DIR}/SysY.interp"
+  "${ANTLR4_GENERATED_DIR}/SysY.tokens"
+  "${ANTLR4_GENERATED_DIR}/SysYBaseVisitor.h"
+  "${ANTLR4_GENERATED_DIR}/SysYVisitor.h"
+)
+
+add_custom_command(
+  OUTPUT ${ANTLR4_GENERATED_FILES}
+  COMMAND ${CMAKE_COMMAND} -E make_directory "${ANTLR4_GENERATED_DIR}"
+  COMMAND ${Java_JAVA_EXECUTABLE} -jar "${ANTLR4_JAR}"
+          -Dlanguage=Cpp
+          -visitor
+          -no-listener
+          -Xexact-output-dir
+          -o "${ANTLR4_GENERATED_DIR}"
+          "${ANTLR4_GRAMMAR}"
+  DEPENDS "${ANTLR4_GRAMMAR}" "${ANTLR4_JAR}"
+  COMMENT "Generating ANTLR4 parser sources from SysY.g4"
+  VERBATIM
+)
+
+add_custom_target(antlr4_generated DEPENDS ${ANTLR4_GENERATED_FILES})
+add_dependencies(frontend antlr4_generated)
+
+target_sources(frontend PRIVATE
+  "${ANTLR4_GENERATED_DIR}/SysYLexer.cpp"
+  "${ANTLR4_GENERATED_DIR}/SysYParser.cpp"
+)
+
 target_link_libraries(frontend PUBLIC
   build_options
   ${ANTLR4_RUNTIME_TARGET}
 )
-
-# 自动纳入构建目录中的 Lexer/Parser 生成源码（若存在）
-file(GLOB_RECURSE ANTLR4_GENERATED_SOURCES CONFIGURE_DEPENDS
-  "${ANTLR4_GENERATED_DIR}/*.cpp"
-)
-if(ANTLR4_GENERATED_SOURCES)
-  target_sources(frontend PRIVATE ${ANTLR4_GENERATED_SOURCES})
-endif()

src/main.rs

@@ -0,0 +1,3 @@
+fn main() {
+    println!("Hello, world!");
+}