Question

我对此有点困惑......我对野牛的语法看起来像这样..

%%
program:
    | program func_defn
    ;

datatype: INTEGER          { $$ = currentType = Type::IntegerTy; }
    | FLOAT                { $$ = currentType = Type::FloatTy; }
    | VOID                 { $$ = currentType = Type::VoidTy; }
    | STRING_LITERAL       { $$ = currentType = Type::StringTy; }
    ;

retlist : datatype
    {
        builder->pushDataType(&getType($1));
    }
    ;

arglist: datatype IDENTIFIER 
    {
        builder->pushDataType(&getType($1));
        Symbol *sym = builder->addSymbol($2, getType($1));
        builder->pushArgName(sym);
    }
    | arglist ',' datatype IDENTIFIER 
    {
        builder->pushDataType(&getType($3));
        Symbol *sym = builder->addSymbol($4, getType($3));
        builder->pushArgName(sym);
    }
    |
    ;
//as and when we find the identifiers , we need to add them to a list and use them while constructing the prototype/func

func_defn: FUNCTION IDENTIFIER LSQUARE retlist RSQUARE '(' arglist ')' '{' 
    {
        const std::string& string = $5;
        FunctionProtoType* fp = builder->getProtoType(string);//use current dataTypeList

        /*
        if(fp == NULL) //find the prototype in the module. if not found, add a new one
            builder->addProtoType(string, getType($1), &fp);
        */

        IcErr err = builder->addFunction(*fp);
        if(err != eNoErr)
            yyerror(errMsg[err]);
    }
    ;

%%

我正在尝试解析这一行......

function first_method [int] (int a, int b) {

然而，这在某种程度上似乎根本没有解析。我没有任何shift-reduce或reduce-reduce冲突。

Compiling file: ica_test\func_defn.icaStarting parse
Entering state 0
Reducing stack by rule 1 (line 80):
-> $$ = nterm program (1.1: )
Stack now 0
Entering state 1
Reading a token: Next token is token FUNCTION (1.1: )
Shifting token FUNCTION (1.1: )
Entering state 3
Reading a token:  Next token is token IDENTIFIER (1.1: )
Shifting token IDENTIFIER (1.1: )
Entering state 5
Reading a token: _Next token is token IDENTIFIER (1.1: )

function first_method [int] (int a, int b) {
          ^
Error on Line 1, Column 9: syntax error, unexpected IDENTIFIER, expecting LSQUAR
EError: popping token IDENTIFIER (1.1: )
Stack now 0 1 3
Error: popping token FUNCTION (1.1: )
Stack now 0 1
Error: popping nterm program (1.1: )
Stack now 0
Cleanup: discarding lookahead token IDENTIFIER (1.1: )
Stack now 0

Stopping compilation as we found some syntax errors in ica_test\func_defn.ica
!Press any key to continue . . .

Answer 1

我猜你的IDENTIFIER词法分析器规则不能包含_（下划线）字符，因此将first_method标记为标识符first，然后{ {1}}与任何规则都不匹配，因此会回显到stdout（输出中_之前的杂散下划线）。然后它会获得另一个标识符（Next），这会导致语法错误，因为它需要method。

我说“猜”，因为你没有显示你的词法分析器代码，但似乎很可能。

您应该通过修复[规则来解决此问题，但您还应添加最终的词法规则，例如：

IDENTIFIER

匹配任何单个字符并将其返回给解析器，而不是将其回显给输出。这样你就不会得到随机字符回显到输出并被忽略，你会得到更明智的语法错误 - 它们会在YYDEBUG文本输出中显示为标记，所以你会知道什么是继续。

用语法解析时的野牛错误

1 个答案: