[antlr-interest] Token Names

[Adam Connelly]

Hi,

When antlr's doing error reporting it uses the tokenNames array to get the
name of tokens that it couldn't parse (from what I can see).  If I want to
change what's displayed in the error message I can easily alter the values
in this array to do so, but is there a way to do this in the grammar?

Here's an example:

rule: expression ';' ;

could give a tokenNames array something like this:

public static final String[] tokenNames = new String[] {
        "<invalid>", "<EOR>", "<DOWN>", "<UP>", "';'"
    };

whereas

rule: expression SEMI ;

SEMI: ';' ;

could give something like this:

public static final String[] tokenNames = new String[] {
        "<invalid>", "<EOR>", "<DOWN>", "<UP>", "SEMI"
    };

Is there a way in the grammar to tell antlr to put the actual ';' in the
tokenNames array rather than the name "SEMI" (which apart from the innuendo
might not be amazingly useful to users :))?

Cheers,

Adam
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.antlr.org/pipermail/antlr-interest/attachments/20080108/2edf2a70/attachment.html