[antlr-interest] non-determinism warnings again

[Tarun Khanna]

Hey thanks for changing the code....I'll try to see why antlr does not 
generate warnings now..

I ran the code and it does not parse correctly. I get a "invalid token null" 
error.

try giving this as an input to the parser "a.b + b.g". Theoratically this 
should work.
You can save the following code in a file called Main.java and use it for 
testing.

Thanks and regards,
Tarun


import java.io.*;
public class Main {
public static void main(String[] args) throws Exception {
try {
//testLexer lexer = new testLexer(new FileInputStream(args[0]));
testLexer lexer = new testLexer(new StringReader("a.b + b + x"));

testParser parser = new testParser(lexer);

parser.exp();

}
catch(Exception e) 
{
System.err.println("exception: "+e);
e.printStackTrace(System.err); // so we can get stack trace
}
}

}
On 7/8/05, John B. Brodie <jbb at acm.org> wrote:
> 
> Hello :-
> 
> Attached please find an updated grammar which removes your ambiguity (no
> longer need to set any k=x option at all).
> 
> I can not explain why, but I changed your factor rule to utilize an 
> optional
> `qualification` rule (e.g. the DOT (ID | TAB) stuff) and this did the 
> trick.
> Note I have not actually tried to execute the generated parser, I have 
> only
> run it through the antlr.Tool and got no complaints.
> 
> Again do not know why this fixed the problem...
> 
> I also made some gratuitous changes to your lexer:
> 1) put the "mod" keyword into a tokens {} section; and
> 2) removed the need for k=2 in the lexer by updating the NL rule.
> 
> Hope this helps...
> -jbb
> 
> /*-------------------------begin test.g-------------------------*/
> 
> class testParser extends Parser;
> 
> //options { k = 2; }
> 
> exp : term ( (PLUS | MINUS) term )* ;
> 
> term : factor ( (TIMES | DIV | MOD) factor )* ;
> 
> factor :
> NUM
> | ( ( LPAREN exp RPAREN ) | IDENT ) ( qualification )?
> | TAB
> ;
> 
> qualification :
> DOT
> ( ( ID ( qualification )? )
> | TAB )
> ;
> 
> class testLexer extends Lexer;
> 
> options {
> charVocabulary = '\0'..'\377';
> testLiterals = true;
> //k=2;
> }
> 
> tokens {
> MOD = "mod";
> }
> 
> LPAREN : '(' ;
> RPAREN : ')' ;
> PLUS : '+' ;
> MINUS : '-' ;
> TIMES : '*' ;
> DIV : '/' ;
> DOT : '.' ;
> COMMA : ',' ;
> LBRACK : '{' ;
> RBRACK : '}' ;
> SEMI : ';' ;
> 
> NUM : ('0'..'9')+ ;
> 
> IDENT : ('a'..'z'|'A'..'Z'|'_') ('a'..'z'|'A'..'Z'|'0'..'9'|'_')* ;
> 
> TAB : '\t' ;
> 
> protected NL :
> ( '\r' // '\r' used on Macintosh
> ( '\n' )? // "\r\n" used on DOS/Windows
> | '\n' ) // '\n' used on Unix
> { newline(); }
> ;
> 
> WS : ( ' ' | '\f' | NL ) { $setType(Token.SKIP); } ;
> 
> /*------------------------- end test.g -------------------------*/
> 



-- 
Tarun Khanna
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.antlr.org/pipermail/antlr-interest/attachments/20050708/ddd4d245/attachment.html