Different Newline Recognition in JavaScript ANTLR4 Target

Question

This is a portion of my island grammar:

listing
    : LISTING (newlines listingPath)? (newlines label)? newlines caption
    ;

newlines
    : NL+
    ;

And for example some input text:

~~~~~

Listing 1 content...

~~~~~
{#lst:listing-1#}
##### Explanation for listing 1. #####

This gets recognized very well. Both in the IntelliJ prototype project (I think Java runs in the background) as well as in my application with a JavaScript target.

When I replace the "newlines" rule within the "listing" rule with "NL+" or "NL" I receive a parsing error that the first \\r\\n after the closing listing tag "~~~~~" is a bad token. The newline doesn't get recognized.

Both versions working well with IntelliJ in the prototype project. I tested the grammars with the ANTLR preview feature.

Am I missing something or is this a possible JavaScript target bug in ANTLR4?

Answer 1

It was just a small mistake within my parser grammar file.

Of course no bug within the ANTLR4 code generation. :)