[prog] lex/yacc problem

[Jimen Ching]

Hi all,

I've been stuck on this lex/yacc problem for a day now, so I thought I
start asking for help.  ;-)

I'm writing a grammar that needs to parse the following:

	result = #1 'b1001;

Here, the #1 is a token, and 'b1001 is another token.  The yacc BNF looks
like this:

statement : assignment ';' ;
assignment : lvalue '=' optional_delay expression ;
lvalue : identifier ;
identifier : YYCHARSTR ;
optional_delay : /* nothing */ | YYPOUND unsigned_number ;
expression : primary ;
primary : unsigned_number ;
unsigned_number : YYNUMBER ;

The lexer has the following patterns:

WS		[ \t\r\b]
Digit		[0-9]
DigitU		[0-9_]
Letter		[a-zA-Z]
LetterU		[a-zA-Z_]
WordNum		[0-9a-zA-Z]
WordNumU	[0-9a-zA-Z_]
Number		{Digit}{DigitU}*
Word		{LetterU}{WordNumU}*
Binary		([-+]?{Number}{WS}*)?\'[bB]{WS}*[01xXzZ?][01xXzZ?_]*

<INITIAL>"#"		{ save_word("#"); return YYPOUND; }
<INITIAL>{Binary}	{ copy_word(yytext); return YYNUMBER; }
<INITIAL>{Word}		{ copy_word(yytext); return YYCHARSTR; }

The problem I'm running into is that a parser error occurs for the input
above.  If I remove the pound sign (#), then it works fine.  It seems like
the parser is trying to match the "1 'b1001" to a YYNUMBER instead of
treating the "#1" as a token.  How do I write the specification so it does
what I want?

Is this enough details to determine the problem?  If not, I can send my
original source code.  It is rather large, so that's why I didn't include
it here.

Thanks in advance.

--jc
-- 
Jimen Ching (WH6BRR)      jching at flex.com     wh6brr at uhm.ampr.org