[antlr-interest] No viable alternative.

rkevinburton at charter.net rkevinburton at charter.net
Mon Jul 28 21:18:59 PDT 2008


Same problem. That is one of the first things that I tried. The grammer "auto-inserts" the semi. Whan I run it in the interpreter the presense or absence of the semi doesn't seem to affect the results.

Thank you.

Kevin
---- Chris Rebert <cvrebert at gmail.com> wrote: 
> Just a hunch, but what happens if you add a semicolon at the end of
> the first "var" line?
> 
> On Mon, Jul 28, 2008 at 12:12 PM,  <rkevinburton at charter.net> wrote:
> > I am relatively new to ANTLR and so I decided to look at "modifying" some existing grammars and see what that gives me. I decided to start with a ECMAScript grammar from Xebic Research. The only change I made to it was to add types to the arguments to functions and have it generate C# code. It didn't work :-( The following function declaration works just fine:
> >
> > function GetCatalogs(catalogContext:CatalogContext)
> > {
> >  var getCatalogs = catalogContext.GetCatalogs().Catalogs
> > }
> >
> >
> > BUt as soon as I add another variable declaration I get a NoViableAlt exception:
> >
> > function GetCatalogs(catalogContext:CatalogContext)
> > {
> >  var getCatalogs = catalogContext.GetCatalogs().Catalogs
> >  var catalogResponseType = new CatalogDataSet()
> > }
> >
> > Due to the size I will not try to reproduce the applicable rules here but I ran it under ANTLRWorks (Interpeter) and I received the same results. So if you need to test you could take the above function declaration (take out the argument type after the colon if you haven't modified the grammar to handle types) and you should get the same error (starting at functionDeclaration). Any ideas?
> >
> > Thank you.
> >
> > Kevin
> >



More information about the antlr-interest mailing list