Im pretty new to compiler design. im trying to take the first step in lexical analyzer trying to run the following code.
%{
#include <stdlib.h>
#include <stdio.h>
#include "symboltable.h"
#include "tokens.h"
entry_t** symbol_table;
entry_t** constant_table;
int cmnt_strt = 0;
%}
letter [a-zA-Z]
digit [0-9]
ws [ \t\r\f\v]
identifier (_|{letter})({letter}|{digit}|_){0,31}
hex [0-9a-f]
/* Exclusive states */
%x CMNT
%x PREPROC
%%
/* Keywords*/
"int" {printf("\t%-30s : =\n",yytext,INT);}
"long" {printf("\t%-30s : =\n",yytext,LONG);}
"long long" {printf("\t%-30s : =\n",yytext,LONG_LONG);}
"short" {printf("\t%-30s : =\n",yytext,SHORT);}
"signed" {printf("\t%-30s : =\n",yytext,SIGNED);}
"unsigned" {printf("\t%-30s : =\n",yytext,UNSIGNED);}
"for" {printf("\t%-30s : =\n",yytext,FOR);}
"break" {printf("\t%-30s : =\n",yytext,BREAK);}
"continue" {printf("\t%-30s : =\n",yytext,CONTINUE);}
"if" {printf("\t%-30s : =\n",yytext,IF);}
"else" {printf("\t%-30s : =\n",yytext,ELSE);}
"return" {printf("\t%-30s : =\n",yytext,RETURN);}
{identifier} {printf("\t%-30s : =\n", yytext,IDENTIFIER);
insert( symbol_table,yytext,IDENTIFIER );}
{ws} ;
[ \-]?[0][x|X]{hex} [lLuU]? {printf("\t%-30s : =\n", yytext,HEX_CONSTANT);
insert( constant_table,yytext,HEX_CONSTANT);}
[ \-]?{digit} [lLuU]? {printf("\t%-30s : =\n", yytext,DEC_CONSTANT);
insert( constant_table,yytext,DEC_CONSTANT);}
"/*" {cmnt_strt = yylineno; BEGIN CMNT;}
<CMNT>.|{ws} ;
<CMNT>\n {yylineno ;}
<CMNT>"*/" {BEGIN INITIAL;}
<CMNT>"/*" {printf("Line =: Nested comments are not valid!\n",yylineno);}
<CMNT><<EOF>> {printf("Line =: Unterminated comment\n", cmnt_strt); yyterminate();}
^"#include" {BEGIN PREPROC;}
<PREPROC>"<"[^<>\n] ">" {printf("\t%-30s : =\n",yytext,HEADER_FILE);}
<PREPROC>{ws} ;
<PREPROC>\"[^"\n] \" {printf("\t%-30s : %3d\n",yytext,HEADER_FILE);}
<PREPROC>\n {yylineno ; BEGIN INITIAL;}
<PREPROC>. {printf("Line %3d: Illegal header file format \n",yylineno);}
"//".* ;
\"[^\"\n]*\" {
if(yytext[yyleng-2]=='\\') /* check if it was an escaped quote */
{
yyless(yyleng-1); /* push the quote back if it was escaped */
yymore();
}
else
insert( constant_table,yytext,STRING);
}
\"[^\"\n]*$ {printf("Line %3d: Unterminated string %s\n",yylineno,yytext);}
{digit} ({letter}|_) {printf("Line %3d: Illegal identifier name %s\n",yylineno,yytext);}
\n {yylineno ;}
"--" {printf("\t%-30s : %3d\n",yytext,DECREMENT);}
" " {printf("\t%-30s : %3d\n",yytext,INCREMENT);}
"->" {printf("\t%-30s : %3d\n",yytext,PTR_SELECT);}
"&&" {printf("\t%-30s : %3d\n",yytext,LOGICAL_AND);}
"||" {printf("\t%-30s : %3d\n",yytext,LOGICAL_OR);}
"<=" {printf("\t%-30s : %3d\n",yytext,LS_THAN_EQ);}
">=" {printf("\t%-30s : %3d\n",yytext,GR_THAN_EQ);}
"==" {printf("\t%-30s : %3d\n",yytext,EQ);}
"!=" {printf("\t%-30s : %3d\n",yytext,NOT_EQ);}
";" {printf("\t%-30s : %3d\n",yytext,DELIMITER);}
"{" {printf("\t%-30s : %3d\n",yytext,OPEN_BRACES);}
"}" {printf("\t%-30s : %3d\n",yytext,CLOSE_BRACES);}
"," {printf("\t%-30s : %3d\n",yytext,COMMA);}
"=" {printf("\t%-30s : %3d\n",yytext,ASSIGN);}
"(" {printf("\t%-30s : %3d\n",yytext,OPEN_PAR);}
")" {printf("\t%-30s : %3d\n",yytext,CLOSE_PAR);}
"[" {printf("\t%-30s : %3d\n",yytext,OPEN_SQ_BRKT);}
"]" {printf("\t%-30s : %3d\n",yytext,CLOSE_SQ_BRKT);}
"-" {printf("\t%-30s : %3d\n",yytext,MINUS);}
" " {printf("\t%-30s : %3d\n",yytext,PLUS);}
"*" {printf("\t%-30s : %3d\n",yytext,STAR);}
"/" {printf("\t%-30s : %3d\n",yytext,FW_SLASH);}
"%" {printf("\t%-30s : %3d\n",yytext,MODULO);}
"<" {printf("\t%-30s : %3d\n",yytext,LS_THAN);}
">" {printf("\t%-30s : %3d\n",yytext,GR_THAN);}
. {printf("Line %3d: Illegal character %s\n",yylineno,yytext);}
%%
int yywrap(){ return 1;}
int main()
{
yyin=fopen("testcases/test-case-1.c","r");
symbol_table=create_table();
constant_table=create_table();
yylex();
printf("\n\tSymbol table");
display(symbol_table);
printf("\n\tConstants Table");
display(constant_table);
printf("NOTE: Please refer tokens.h for token meanings\n");
}
I tried to run it using:
flex lexer.l
gcc lex.yy.c -o lexrun
This gives me an error as
lexer.l: In function 'yylex':
lexer.l:46:14: error: 'yylineno' undeclared (first use in this function)
"/*" {cmnt_strt = yylineno; BEGIN CMNT;}
^~~~~~~~
lexer.l:46:14: note: each undeclared identifier is reported only once for each function it appears in
I tried many things on the net such as adding
extern int yylineno;
This started giving another error as
undefined reference to `yylineno'
I have installed Flex version 2.5.4 and runs this in Vscode. Any clarification to overcome this would be much appreciated. Thanks in advance.
CodePudding user response：
If you want flex to track line numbers you need to add
%option yylineno
to your flex prologue.
I also suggest the following:
-
%option noinput nounputwhich will allow you to compile the generated scanner without compiler warnings (Always compile with
-Wall, even generated code, and fix whatever warnings are reported.) -
%option noyywrapwhich avoids the need to define
yywrap -
%option nodefualtwhich will warn you if you don't have a rule for every possible input.
As a final note, extern int yylineno; can't work, since extern means "this variable is defined in a different translation unit", and there is no other translation unit in your code. I presume that you found that in the discussion of a different file intended to be linked together with the scanner. (If the place you found that suggests putting the extern declaration in a .l file, you need to discard it as a source of information.)
If this is the first time you've attempted to write a C application with more than one source file, you should probably take a few minutes to review how linking multiple files works in C. That will save you a lot of frustration later.