Question

我尝试使用LEX和YACC创建一个语法分析器，它与C ++非常相似...... 这部分是LEX部分：

%{
#include <stdio.h>
#include "y.tab.h"
%}
%x comment
%x comment_line
%%

"//" {BEGIN comment_line;}
<comment_line>\n {BEGIN 0;}
<comment_line>. ;
"/*" {BEGIN comment;} 
<comment>"*/" {BEGIN 0;}
<comment>. ;

"bool"|"const"|"int"|"float"|"char"|"string" {return TIP;}
"void" {return TIP_VOID;}
"class" {return TIP_CLASS;}
\'.\' {return VAL_CHAR;}
\".*\" {return VAL_STRING;}
"start_program" {return BGIN;}
"end_program" {return END;}
"start_func" {return START_FUNC;}
"end_func" {return END_FUNC;}
"start_class" {return START_CLASS;}
"end_class" {return END_CLASS;}
"if" {return IF;}
"then" {return THEN;}
"end_then" {return END_THEN;}
"else" {return ELSE;}
"end_else" {return END_ELSE;}
"while" {return WHILE;}
"do" {return DO;};
"end_while" {return END_WHILE;}
"for" {return FOR;}
"end_for" {return END_FOR;}
"return" {return RETURN;}
"true"|"false" {return VAL_BOOL;}
"and"|"or" {return OPERATOR_BOOL;}
"<"|">"|"<="|"=>" {return OPERATOR_COMPARISON;}
"=="|"!=" {return OPERATOR_INEQUALITY;}
"<-" {return ASSIGN;}
[_a-zA-Z][_a-zA-Z0-9]* {return ID;}
0|[1-9][0-9]* {return NR_INT_U;}
-?0|([1-9][0-9]*) {return NR_INT;}
-?(0|[1-9][0-9]*)","[0-9]* {return NR_FLOAT;}
[ \t] ;
\n {yylineno++;}
. { return yytext[0];}

虽然这部分是YACC部分：

%{
#include <stdio.h>
extern FILE* yyin;
extern char* yytext;
extern int yylineno;
%}
%token ID TIP BGIN END ASSIGN NR NR_FLOAT NR_INT NR_INT_U TIP_VOID START_FUNC END_FUNC RETURN TIP_CLASS START_CLASS END_CLASS OPERATOR_COMPARISON OPERATOR_INEQUALITY VAL_BOOL IF THEN END_THEN ELSE END_ELSE DO WHILE END_WHILE FOR END_FOR VAL_CHAR VAL_STRING OPERATOR_BOOL
%start progr
%%
progr: declarations bloc {printf("Syntacticaly correct program\n");}
     ;
declarations : declaration ';'
       | declarations declaration ';'
       ;
declaration : variable_declaration
       | function body_function
       | class_declaration
       ;

class_declaration : TIP_CLASS ID START_CLASS declarations END_CLASS
            ;

variable_declaration : TIP ID 
            | TIP ID ASSIGN value_asign
            | TIP ID vector_declaration
            | TIP_CLASS ID ID
            ;
value_asign: NR_INT_U 
        | VAL_BOOL
        | NR_INT 
        | NR_FLOAT 
        | VAL_CHAR 
        | VAL_STRING
        | ID '(' list_call ')'
        | ID '.' ID '(' list_call ')'
        ;
/*functions declaration */
function : TIP ID '(' parameter_list ')'
       | TIP ID '(' ')'
       | TIP_VOID ID '(' parameter_list ')'
       | TIP_VOID ID '(' ')'
       ;

body_function: START_FUNC list_function END_FUNC
        ;
list_function : RETURN expression ';'
         | list_function RETURN expression ';'
         | statement ';' 
         | list_function statement ';'
         | bloc_logic_function
         | list bloc_logic_function
         ;
bloc_logic_function: IF expression_logic 
            THEN list_function END_THEN
            ELSE list_function END_ELSE
          | WHILE expression_logic
            list_function 
            END_WHILE
          | FOR statement WHILE expression_logic DO statement
            list_function
             END_FOR
          ;


/*declaration of multidimensional arrays*/
vector_declaration : '[' NR_INT_U ']'
                 | vector_declaration '[' NR_INT_U ']'
                 ;
vector : '[' expression ']'
       | vector '[' expression ']'
       ;

parameter_list : param
            | parameter_list ','  param 
            ;

param : TIP ID      
      ; 

/* bloc */
bloc : BGIN list END  
     ;

/* list of instructions */
list : bloc_logic
     | list bloc_logic
     | statement ';' 
     | list statement ';'   
     ;
bloc_logic: IF expression_logic 
        THEN list END_THEN
        ELSE list END_ELSE
      | WHILE expression_logic
        list 
        END_WHILE
      | FOR statement WHILE expression_logic DO statement
        list
        END_FOR
      ;

/* instruction */
statement: variable_declaration
     | ID ASSIGN ID
     | ID ASSIGN NR_INT_U
     | ID ASSIGN NR_INT
     | ID ASSIGN NR_FLOAT 
     | ID ASSIGN VAL_BOOL
     | ID ASSIGN expression
     | ID vector ASSIGN expression
     | ID '(' list_call ')'
     | ID '.' ID '(' list_call ')'
     ;

expression: simbol
    | expression '+' expression
    | '('expression '+' expression ')'
    | expression '-' expression
    | '('expression '-' expression ')'
    | expression '*' expression
    | '('expression '*' expression ')'
    | expression ':' expression
    | '('expression ':' expression ')'
    ;
simbol: ID
      | '-' ID
      | '-' ID vector
      | ID vector
      | NR_INT_U
      | NR_INT
      | NR_FLOAT
      | ID '(' list_call ')'
      | ID '.' ID '(' list_call ')'
      ;
expression_logic : logical_condition
        | '(' logical_condition ')'
        | expression_logic OPERATOR_BOOL expression_logic
        | '(' expression_logic OPERATOR_BOOL expression_logic ')'
        ;
logical_condition : VAL_BOOL OPERATOR_INEQUALITY VAL_BOOL
        | VAL_BOOL OPERATOR_INEQUALITY expression
        | expression OPERATOR_COMPARISON expression  
        | expression OPERATOR_INEQUALITY expression 
        | expression OPERATOR_INEQUALITY VAL_BOOL
        ;
list_call : expression
       | list_call ',' expression
       | VAL_STRING
       | list_call ',' VAL_STRING
       | VAL_CHAR
       | list_call ',' VAL_CHAR
       | VAL_BOOL
       | list_call ',' VAL_BOOL
       |  /*empty*/
       ;
%%
int yyerror(char * s){
printf("Error: %s at line:%d\n",s,yylineno);
}

int main(int argc, char** argv){
yyin=fopen(argv[1],"r");
yyparse();
}

这是我尝试运行它的例子：

 int factorial(int n)
  start_func
    if n==0
      then
        return 1;
      end_then
    else
      return n*factorial(n-1);
    end_else
  end_func;

class book
  start_class
    int code;
    string name;
    float price;
  void setPrice(float x)
        start_func
         price <- x;
       end_func;

    string getNume()
       start_func
         return name;
       end_func;
  end_class;

start_program

   class book python;
   float y<-1,37;
   string auxiliar;

   python.setPrice(y);
   auxiliar<-getNume();

  int v[50];
  int j<-0;
  for i<-0 while i<5 do i<-i+1
  v[i]<-j+1;
  end_for 

end_program

我正在尝试使用这些说明在Putty上运行它：

lex tema.l
yacc -d tema.y 2>/dev/null
gcc y.tab.c lex.yy.c -ll
./a.out <prog2.txt

我似乎无法找到问题...我已经搜索了每一行代码，但是我在屏幕上显示的只是我在第一行显示错误，这是没有意义的，因为我可以'看到任何错误......也许我在这里失明了......你怎么看？

Answer 1

显然你在第一行有回车或换文或其他一些奇怪的字符。您的空白和行尾规则过于严格。尝试

[ \t\f]+ ;
"\r\n"|"\n" { yylineno++; }

第一行将换页作为空格处理，添加+使得效率更高。第二行将\r\n和\n作为行终止符处理。

此外，你需要解决所有这些问题：

"bool"|"const"|"int"|"float"|"char"|"string" {return TIP;}
"true"|"false" {return VAL_BOOL;}
"and"|"or" {return OPERATOR_BOOL;}
"<"|">"|"<="|"=>" {return OPERATOR_COMPARISON;}
"=="|"!=" {return OPERATOR_INEQUALITY;}

不要＆＃39;优化＆＃39;像这样。您在这里丢失了信息，您将无法生成正确的代码。此外，通过合并and和or，您已经拒绝了自己有机会获得正确的优先权。让解析器完成解析器的工作。这样做，适当调整你的语法和%token声明：

"bool" {return TIP_BOOL;}
"const" {return TIP_CONST;}
"int" {return TIP_INT;}
"float" {return TIP_FLOAT;}
"char" {return TIP_CHAR;}
"string" {return TIP_STRING;}
"true" {return VAL_TRUE;}
"false" {return VAL_FALSE;}
"and" { return OPERATOR_AND; }
"or" {return OPERATOR_OR;}
"<=" { return OPERATOR_LE; }
"=>" {return OPERATOR_GE;}
"==" {return OPERATOR_EQUALITY;}
"!=" {return OPERATOR_INEQUALITY;}

并注意，您根本不必处理<，>，=，+等等：它们由

. { return yytext[0];}

因此您可以直接在语法中使用'<'，'>'，'='等。

虽然我无法在YACC中检测到它，但仍然会出现语法错误

1 个答案: