From 3d0361e64827893ffaa9c49e54b289508b28e356 Mon Sep 17 00:00:00 2001
From: olivame <dyz@olivame.xyz>
Date: Mon, 23 Mar 2026 15:26:32 +0800
Subject: [PATCH] feat(frontend): complete lab1 SysY grammar support

---
 scripts/test_lab1.sh |  32 +++++++
 src/antlr4/SysY.g4   | 219 +++++++++++++++++++++++++++++++++++++------
 2 files changed, 223 insertions(+), 28 deletions(-)
 create mode 100755 scripts/test_lab1.sh

diff --git a/scripts/test_lab1.sh b/scripts/test_lab1.sh
new file mode 100755
index 0000000..619ebb1
--- /dev/null
+++ b/scripts/test_lab1.sh
@@ -0,0 +1,32 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+case_dir="${1:-test/test_case}"
+
+if [[ ! -d "$case_dir" ]]; then
+  echo "测试目录不存在: $case_dir" >&2
+  exit 1
+fi
+
+compiler="./build/bin/compiler"
+if [[ ! -x "$compiler" ]]; then
+  echo "未找到编译器: $compiler ，请先构建 parse-only 版本。" >&2
+  exit 1
+fi
+
+mapfile -t cases < <(find "$case_dir" -name '*.sy' | sort)
+if [[ ${#cases[@]} -eq 0 ]]; then
+  echo "未找到任何 .sy 测试文件: $case_dir" >&2
+  exit 1
+fi
+
+for f in "${cases[@]}"; do
+  echo "TEST $f"
+  "$compiler" --emit-parse-tree "$f" >/dev/null || {
+    echo "FAIL $f" >&2
+    exit 1
+  }
+done
+
+echo "ALL_PARSE_OK (${#cases[@]} cases)"
diff --git a/src/antlr4/SysY.g4 b/src/antlr4/SysY.g4
index 263aeef..5f6b0c8 100644
--- a/src/antlr4/SysY.g4
+++ b/src/antlr4/SysY.g4
@@ -1,67 +1,155 @@
-// SysY 子集语法：支持形如
-// 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;
 
 /*===-------------------------------------------===*/
 /* Lexer rules                                     */
 /*===-------------------------------------------===*/
 
+CONST: 'const';
 INT: 'int';
+FLOAT: 'float';
+VOID: 'void';
+IF: 'if';
+ELSE: 'else';
+WHILE: 'while';
+BREAK: 'break';
+CONTINUE: 'continue';
 RETURN: 'return';
 
+LE: '<=';
+GE: '>=';
+EQ: '==';
+NE: '!=';
+AND: '&&';
+OR: '||';
+
 ASSIGN: '=';
+LT: '<';
+GT: '>';
 ADD: '+';
+SUB: '-';
+MUL: '*';
+DIV: '/';
+MOD: '%';
+NOT: '!';
 
 LPAREN: '(';
 RPAREN: ')';
+LBRACK: '[';
+RBRACK: ']';
 LBRACE: '{';
 RBRACE: '}';
+COMMA: ',';
 SEMICOLON: ';';
 
 ID: [a-zA-Z_][a-zA-Z_0-9]*;
-ILITERAL: [0-9]+;
+
+HEX_FLOAT_LITERAL
+    : ('0x' | '0X') HEX_DIGIT* '.' HEX_DIGIT+ BINARY_EXPONENT
+    | ('0x' | '0X') HEX_DIGIT+ '.' HEX_DIGIT* BINARY_EXPONENT
+    | ('0x' | '0X') HEX_DIGIT+ BINARY_EXPONENT
+    ;
+
+DEC_FLOAT_LITERAL
+    : DEC_DIGIT+ '.' DEC_DIGIT* DEC_EXPONENT?
+    | '.' DEC_DIGIT+ DEC_EXPONENT?
+    | DEC_DIGIT+ DEC_EXPONENT
+    ;
+
+HEX_INT_LITERAL
+    : ('0x' | '0X') HEX_DIGIT+
+    ;
+
+OCT_INT_LITERAL
+    : '0' OCT_DIGIT+
+    ;
+
+DEC_INT_LITERAL
+    : '0'
+    | [1-9] DEC_DIGIT*
+    ;
 
 WS: [ \t\r\n] -> skip;
 LINECOMMENT: '//' ~[\r\n]* -> skip;
 BLOCKCOMMENT: '/*' .*? '*/' -> skip;
 
+fragment DEC_DIGIT: [0-9];
+fragment OCT_DIGIT: [0-7];
+fragment HEX_DIGIT: [0-9a-fA-F];
+fragment DEC_EXPONENT: [eE] [+-]? DEC_DIGIT+;
+fragment BINARY_EXPONENT: [pP] [+-]? DEC_DIGIT+;
+
 /*===-------------------------------------------===*/
 /* Syntax rules                                    */
 /*===-------------------------------------------===*/
 
 compUnit
-    : funcDef EOF
+    : topLevelItem* EOF
+    ;
+
+topLevelItem
+    : decl
+    | funcDef
     ;
 
 decl
-    : btype varDef SEMICOLON
+    : constDecl
+    | varDecl
+    ;
+
+constDecl
+    : CONST bType constDef (COMMA constDef)* SEMICOLON
     ;
 
-btype
+varDecl
+    : bType varDef (COMMA varDef)* SEMICOLON
+    ;
+
+bType
     : INT
+    | FLOAT
+    ;
+
+constDef
+    : ID constIndex* ASSIGN constInitVal
     ;
 
 varDef
-    : lValue (ASSIGN initValue)?
+    : ID constIndex* (ASSIGN initVal)?
+    ;
+
+constIndex
+    : LBRACK constExp RBRACK
     ;
 
-initValue
+constInitVal
+    : constExp
+    | LBRACE (constInitVal (COMMA constInitVal)*)? RBRACE
+    ;
+
+initVal
     : exp
+    | LBRACE (initVal (COMMA initVal)*)? RBRACE
     ;
 
 funcDef
-    : funcType ID LPAREN RPAREN blockStmt
+    : funcType ID LPAREN funcFParams? RPAREN block
     ;
 
 funcType
-    : INT
+    : VOID
+    | INT
+    | FLOAT
+    ;
+
+funcFParams
+    : funcFParam (COMMA funcFParam)*
+    ;
+
+funcFParam
+    : bType ID (LBRACK RBRACK (LBRACK exp RBRACK)*)?
     ;
 
-blockStmt
+block
     : LBRACE blockItem* RBRACE
     ;
 
@@ -71,28 +159,103 @@ blockItem
     ;
 
 stmt
-    : returnStmt
+    : 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
     ;
 
-returnStmt
-    : RETURN exp SEMICOLON
+exp
+    : addExp
     ;
 
-exp
-    : LPAREN exp RPAREN          # parenExp
-    | var                        # varExp
-    | number                     # numberExp
-    | exp ADD exp                # additiveExp
+cond
+    : lOrExp
     ;
 
-var
-    : ID
+lVal
+    : ID (LBRACK exp RBRACK)*
     ;
 
-lValue
-    : ID
+primaryExp
+    : LPAREN exp RPAREN
+    | lVal
+    | number
     ;
 
 number
-    : ILITERAL
+    : intConst
+    | floatConst
+    ;
+
+intConst
+    : DEC_INT_LITERAL
+    | OCT_INT_LITERAL
+    | HEX_INT_LITERAL
+    ;
+
+floatConst
+    : DEC_FLOAT_LITERAL
+    | HEX_FLOAT_LITERAL
+    ;
+
+unaryExp
+    : primaryExp
+    | ID LPAREN funcRParams? RPAREN
+    | unaryOp unaryExp
+    ;
+
+unaryOp
+    : ADD
+    | SUB
+    | NOT
+    ;
+
+funcRParams
+    : exp (COMMA exp)*
+    ;
+
+mulExp
+    : unaryExp
+    | mulExp MUL unaryExp
+    | mulExp DIV unaryExp
+    | mulExp MOD unaryExp
+    ;
+
+addExp
+    : mulExp
+    | addExp ADD mulExp
+    | addExp SUB mulExp
+    ;
+
+relExp
+    : addExp
+    | relExp LT addExp
+    | relExp GT addExp
+    | relExp LE addExp
+    | relExp GE addExp
+    ;
+
+eqExp
+    : relExp
+    | eqExp EQ relExp
+    | eqExp NE relExp
+    ;
+
+lAndExp
+    : eqExp
+    | lAndExp AND eqExp
+    ;
+
+lOrExp
+    : lAndExp
+    | lOrExp OR lAndExp
+    ;
+
+constExp
+    : addExp
     ;