我已经完成了c-minus的antlr格式语法文件

[Gao wei]

/*
 *       Parser for C-Minus
 * Copyright (C) 2005 Gao Wei <rock...@gmail.com>
 *
 *
 */

class CMinusParser extends Parser;
options{
k=4;
buildAST=true;
}

program : declaration_list ;
declaration_list :declaration(COMMA declaration)*;
declaration:var_declaration | fun_declaration;
var_declaration:type_specifier ID ";"
                |type_specifier ID "[" NUM "];";
type_specifier:INT|VOID;
fun_declaration:type_specifier ID "(" params ")" compound_stmt;
params:param_list|VOID;
param_list:param (COMMA param)*;
param:type_specifier ID|type_specifier ID "[]";
compound_stmt:"{" local_declarations statement_list "}";
local_declarations:(var_declaration )*;
statement_list:(statement)*;
statement:expression_stmt
          |compound_stmt
          |selection_stmt
          |interation_stmt
          |return_stmt;
expression_stmt:expression ";"|";";
selection_stmt: "if" "(" expression ")" statement ("else" statement) ;
interation_stmt: "while" "(" expression ")" statement;
return_stmt: RETURN ";"|RETURN expression ";";

var : ID|ID"[" expression "]";
expression : var "=" simple_expression;
simple_expression:additive_expression (relop additive_expression);
additive_expression:term (addop term)*;
addop:"+"|"-";
term:factor (mulop factor)*;
factor:"(" expression ")"|var|call|NUM;
mulop:"*"|"/";
relop:"<"|">"|"="|"!=";
call:ID "(" args ")";
args:(arg_list)*;
arg_list:expression ( COMMA expression )*;



class CMinusLexer extends Lexer;
options {
    exportVocab=Java;      // call the vocabulary "Java"
    testLiterals=false;    // don't automatically test for literals
    k=4;                   // four characters of lookahead
    charVocabulary='\u0003'..'\uFFFF';
    // without inlining some bitset tests, couldn't do unicode;
    // I need to make ANTLR generate smaller bitsets; see
    // bottom of JavaLexer.java
    codeGenBitsetTestThreshold=20;
}
{
    void deferredNewline(){
    return;
    }
}


LPARENT  : '(';
RPARENT  : ')';
COMMA    : ',';
ID       : ('a'..'z'|'A'..'Z')+;
NUM      : ('0'..'9')+ ;
WS       : ( ' '
          | '\r' '\n'
          | '\n'
          | '\t'
          )
          {$setType(Token.SKIP);}
          ;    
// multiple-line comments
ML_COMMENT
    :    "/*"
        (    /*    '\r' '\n' can be matched in one alternative or by matching
                '\r' in one iteration and '\n' in another.  I am trying to
                handle any flavor of newline that comes in, but the language
                that allows both "\r\n" and "\r" and "\n" to all be valid
                newline is ambiguous.  Consequently, the resulting grammar
                must be ambiguous.  I'm shutting this warning off.
             */
            options {
                generateAmbigWarnings=false;
            }
        :
            { LA(2)!='/' }? '*'
        |    '\r' '\n'        {newline();}
        |    '\r'            {newline();}
        |    '\n'            {newline();}
        |    ~('*'|'\n'|'\r')
        )*
        "*/"
        {$setType(Token.SKIP);}
    ;

CPPComment
        :
                "//" ( ~('\n') )*
                        {$setType(Token.SKIP);}
        ;