Very Computer

Hi,
        I'm not sure if this is the group to post a question about
flex/bison(lex/yacc) problems. I'm having trouble parsing multiple
line in a file with a grammar I wrote.
Note: flex is 2.5.4 and bison 1.25 this has been tried on linux and
solaris 2.5.2 and also using lex/yacc.
The lex/grammar must parse multiple lines of the sort that follows:

bx*123*ab*a 1b*;
bx**3*cd2;
bx*;

Now the ";" isn't important and will eventually be removed but the
lex file is:

%{
#include        "y.tab.h"     /* or the appropriate file bison generates */
#define token(x)        x
%}
%%

^bx     { return token(ID); }
\*      { return token(SEPARATOR); }
[0-9]+  { return token(NUMBER); }
[A-Z ]+ { return token(ALPHA); }
[A-Za-z0-9 ]+ { return token{ALPHNUM); }
;               { return token(TERMINATOR); }
\n      /* ignore */
.       { return token(ERROR); }

The bison/yacc file is:

%token ID
%token SEPARATOR
%token NUMBER
%token ALPHA
%token ALPHNUM
%token TERMINATOR
%token ERROR

%%

strt
        : line TERMINAL
        ;

line
        : ID SEPARATOR fieldexpression
        ;

fieldexpression
        : fields
        | fieldexpression SEPARATOR fields
        ;

fields
        : /* null */
        | NUMERIC
        | ALPHA
        | ALPHNUM
        ;

Of course the test.c containing the main() just defines a few
externs such as yyparse=1, yylen, yytext[], and yylineno and call yyparse
with a function for errors defined as yyerror().

main()
{
        yyparse();

Quote:}

yyerror(s)
        char *s;
{
        fprintf(stdout, " parse error at lineno=%d for token=%s\n",
                yylineno, yytext);

Quote:}

Also not the lex code has been tested and will return or not return the
proper tokens. It is only when the lex and parser are trying to parse
more than two lines.

--
        Mike,

Hi,
        I'm not sure if this is the group to post a question about
flex/bison(lex/yacc) problems. I'm having trouble parsing multiple
line in a file with a grammar I wrote.
Note: flex is 2.5.4 and bison 1.25 this has been tried on linux and
solaris 2.5.2 and also using lex/yacc.
The lex/grammar must parse multiple lines of the sort that follows:

bx*123*ab*a 1b*;
bx**3*cd2;
bx*;

Now the ";" isn't important and will eventually be removed but the
lex file is:

%{
#include        "y.tab.h"     /* or the appropriate file bison generates */
#define token(x)        x
%}
%%

^bx     { return token(ID); }
\*      { return token(SEPARATOR); }
[0-9]+  { return token(NUMBER); }
[A-Z ]+ { return token(ALPHA); }
[A-Za-z0-9 ]+ { return token{ALPHNUM); }
;               { return token(TERMINATOR); }
\n      /* ignore */
.       { return token(ERROR); }

The bison/yacc file is:

%token ID
%token SEPARATOR
%token NUMBER
%token ALPHA
%token ALPHNUM
%token TERMINATOR
%token ERROR

%%

strt
        : line TERMINAL
        ;

line
        : ID SEPARATOR fieldexpression
        ;

fieldexpression
        : fields
        | fieldexpression SEPARATOR fields
        ;

fields
        : /* null */
        | NUMERIC
        | ALPHA
        | ALPHNUM
        ;

Of course the test.c containing the main() just defines a few
externs such as yyparse=1, yylen, yytext[], and yylineno and call yyparse
with a function for errors defined as yyerror().

main()
{
        yyparse();

Quote:}

yyerror(s)
        char *s;
{
        fprintf(stdout, " parse error at lineno=%d for token=%s\n",
                yylineno, yytext);

Quote:}

Also not the lex code has been tested and will return or not return the
proper tokens. It is only when the lex and parser are trying to parse
more than two lines.

--
        Mike,

Quote:>Hi,
>    I'm not sure if this is the group to post a question about
>flex/bison(lex/yacc) problems.

I think comp.compilers would be better, but well....

 I'm having trouble parsing multiple

Quote:>line in a file with a grammar I wrote.
>Note: flex is 2.5.4 and bison 1.25 this has been tried on linux and
>solaris 2.5.2 and also using lex/yacc.

[del del]
yacc file:

Quote:>%%
>strt
>    : line TERMINAL
>    ;

You should use a recursive rule here to allow multiple line to be parsed.
Like :

strt    : oneline
        | oneline strt

oneline : line TERMINAL

Sorry for the delay, but I was offline for two days (on business travel
to and from Berlin).

Uli
--

Stormweg 24                | listening to: Burning Desire (Funeral Oration)
24539 Neumuenster, Germany | Back With You/What Am I Supposed To Do (Buzzcocks)