add .codex

scripts/run_all_tests.sh

@@ -0,0 +1,50 @@
+#!/bin/bash
+
+# 批量测试所有.sy文件的语法解析
+
+test_dir="/home/lingli/nudt-compiler-cpp/test/test_case"
+compiler="/home/lingli/nudt-compiler-cpp/build/bin/compiler"
+
+if [ ! -f "$compiler" ]; then
+    echo "错误：编译器不存在，请先构建项目"
+    exit 1
+fi
+
+success_count=0
+failed_count=0
+failed_tests=()
+
+echo "开始测试所有.sy文件的语法解析..."
+echo "="
+
+# 获取所有.sy文件并排序
+for test_file in $(find "$test_dir" -name "*.sy" | sort); do
+    echo "测试: $(basename "$test_file")"
+    
+    # 运行解析测试，捕获输出
+    output=$("$compiler" --emit-parse-tree "$test_file" 2>&1)
+    exit_code=$?
+    
+    if [ $exit_code -eq 0 ]; then
+        echo "  ✓ 成功"
+        ((success_count++))
+    else
+        echo "  ✗ 失败"
+        echo "    错误信息: $output"
+        ((failed_count++))
+        failed_tests+=($(basename "$test_file"))
+    fi
+done
+
+echo "="
+echo "测试完成！"
+echo "总测试数: $((success_count + failed_count))"
+echo "成功: $success_count"
+echo "失败: $failed_count"
+
+if [ $failed_count -gt 0 ]; then
+    echo "失败的测试用例:"
+    for test in "${failed_tests[@]}"; do
+        echo "  - $test"
+    done
+fi

src/antlr4/SysY.g4

@@ -1,8 +1,4 @@
-// SysY 子集语法：支持形如
+// SysY 语法：扩展支持更多SysY特性
-// int main() { int a = 1; int b = 2; return a + b; }
-// 的最小返回表达式编译。
-
-// 后续需要自行添加
 
 grammar SysY;
 
@@ -10,21 +6,101 @@ grammar SysY;
 /* Lexer rules                                     */
 /*===-------------------------------------------===*/
 
+// 关键字
 INT: 'int';
+FLOAT: 'float';
+VOID: 'void';
+CONST: 'const';
 RETURN: 'return';
+IF: 'if';
+ELSE: 'else';
+WHILE: 'while';
+BREAK: 'break';
+CONTINUE: 'continue';
 
+// 操作符
 ASSIGN: '=';
 ADD: '+';
+SUB: '-';
+MUL: '*';
+DIV: '/';
+MOD: '%';
+LT: '<';
+LE: '<=';
+GT: '>';
+GE: '>=';
+EQ: '==';
+NE: '!=';
 
+// 逻辑操作符
+NOT: '!';
+AND: '&&';
+OR: '||';
+
+// 括号
 LPAREN: '(';
 RPAREN: ')';
 LBRACE: '{';
 RBRACE: '}';
+LBRACK: '[';
+RBRACK: ']';
+
+// 标点
 SEMICOLON: ';';
+COMMA: ',';
 
+// 标识符和字面量
 ID: [a-zA-Z_][a-zA-Z_0-9]*;
-ILITERAL: [0-9]+;
+ILITERAL
+    : DECIMAL_LITERAL
+    | OCTAL_LITERAL
+    | HEX_LITERAL
+    ;
 
+fragment DECIMAL_LITERAL
+    : [0-9]+
+    ;
+
+fragment OCTAL_LITERAL
+    : '0' [0-7]+
+    ;
+
+fragment HEX_LITERAL
+    : '0' ('x' | 'X') [0-9a-fA-F]+
+    ;
+
+// 浮点字面量
+FLITERAL
+    : (DECIMAL_FLOAT | HEX_FLOAT)
+    ;
+
+fragment DECIMAL_FLOAT
+    : ((DIGIT+ '.' DIGIT* | '.' DIGIT+)
+        (('E' | 'e') ('+' | '-')? DIGIT+)?)  
+    | ((DIGIT+ '.' DIGIT* | '.' DIGIT+ | DIGIT+)
+        (('E' | 'e') ('+' | '-')? DIGIT+))  
+    | ('0' [0-7]+ '.' [0-7]*
+        (('E' | 'e') ('+' | '-')? DIGIT+)?)  
+    ;
+
+fragment HEX_FLOAT
+    : '0' ('x' | 'X')
+        (HEXDIGIT* '.' HEXDIGIT+ | HEXDIGIT+ '.')
+        (('P' | 'p') ('+' | '-')? DIGIT+)
+    | '0' ('x' | 'X')
+        HEXDIGIT+
+        (('P' | 'p') ('+' | '-')? DIGIT+)
+    ;
+
+fragment DIGIT
+    : [0-9]
+    ;
+
+fragment HEXDIGIT
+    : [0-9a-fA-F]
+    ;
+
+// 空白和注释
 WS: [ \t\r\n] -> skip;
 LINECOMMENT: '//' ~[\r\n]* -> skip;
 BLOCKCOMMENT: '/*' .*? '*/' -> skip;
@@ -34,33 +110,62 @@ BLOCKCOMMENT: '/*' .*? '*/' -> skip;
 /*===-------------------------------------------===*/
 
 compUnit
-    : funcDef EOF
+    : (decl | funcDef)* EOF
     ;
 
+// 声明
 decl
-    : btype varDef SEMICOLON
+    : constDecl
+    | varDecl
+    ;
+
+constDecl
+    : CONST btype constDef (COMMA constDef)* SEMICOLON
+    ;
+
+varDecl
+    : btype varDef (COMMA varDef)* SEMICOLON
     ;
 
 btype
     : INT
+    | FLOAT
+    | VOID
+    ;
+
+constDef
+    : ID (LBRACK exp RBRACK)* ASSIGN initValue
     ;
 
 varDef
-    : lValue (ASSIGN initValue)?
+    : ID (LBRACK exp RBRACK)* (ASSIGN initValue)?
     ;
 
 initValue
     : exp
+    | LBRACE (initValue (COMMA initValue)*)? RBRACE
     ;
 
+// 函数定义
 funcDef
-    : funcType ID LPAREN RPAREN blockStmt
+    : funcType ID LPAREN (funcFParams)? RPAREN blockStmt
     ;
 
 funcType
     : INT
+    | FLOAT
+    | VOID
     ;
 
+funcFParams
+    : funcFParam (COMMA funcFParam)*
+    ;
+
+funcFParam
+    : btype ID (LBRACK (exp)? RBRACK)*
+    ;
+
+// 语句
 blockStmt
     : LBRACE blockItem* RBRACE
     ;
@@ -71,28 +176,77 @@ blockItem
     ;
 
 stmt
-    : returnStmt
+    : assignStmt
+    | returnStmt
+    | blockStmt
+    | ifStmt
+    | whileStmt
+    | breakStmt
+    | continueStmt
+    | expStmt
+    ;
+
+expStmt
+    : exp SEMICOLON
+    ;
+
+assignStmt
+    : lValue ASSIGN exp SEMICOLON
     ;
 
 returnStmt
-    : RETURN exp SEMICOLON
+    : RETURN (exp)? SEMICOLON
+    ;
+
+ifStmt
+    : IF LPAREN exp RPAREN stmt (ELSE stmt)?
+    ;
+
+whileStmt
+    : WHILE LPAREN exp RPAREN stmt
+    ;
+
+breakStmt
+    : BREAK SEMICOLON
+    ;
+
+continueStmt
+    : CONTINUE SEMICOLON
+    ;
+
+// 表达式
+lValue
+    : ID (LBRACK exp RBRACK)*
     ;
 
 exp
     : LPAREN exp RPAREN          # parenExp
-    | var                        # varExp
+    | lValue                     # lValueExp
     | number                     # numberExp
-    | exp ADD exp                # additiveExp
+    | ID LPAREN (funcRParams)? RPAREN # funcCallExp
+    | NOT exp                    # notExp
+    | ADD exp                    # unaryAddExp
+    | SUB exp                    # unarySubExp
+    | exp MUL exp                # mulExp
+    | exp DIV exp                # divExp
+    | exp MOD exp                # modExp
+    | exp ADD exp                # addExp
+    | exp SUB exp                # subExp
+    | exp LT exp                 # ltExp
+    | exp LE exp                 # leExp
+    | exp GT exp                 # gtExp
+    | exp GE exp                 # geExp
+    | exp EQ exp                 # eqExp
+    | exp NE exp                 # neExp
+    | exp AND exp                # andExp
+    | exp OR exp                 # orExp
     ;
 
-var
+funcRParams
-    : ID
+    : exp (COMMA exp)*
-    ;
-
-lValue
-    : ID
     ;
 
 number
     : ILITERAL
+    | FLITERAL
     ;