I'm currently using the ANTLR version that
emits C# code and can't take it anymore. The main
reason I picked ANTLR to do this project was all
the advanced and cool features of ANTLR like
near-infinite lookahead given dissambiguating
rules, controlling lexer return types, etc.
Unfortunately it looks like whoever wrote the C#
port tried it on some "Hello World" level examples
and then considered his work done without checking
to see if any of the advanced features worked
correctly. Of course, my freaking employment
contract prevents me from simply fixing these bugs
and giving them back to the original author and
instead I have to work around them.
Click below for another popular reason Open Source
projects don't have as many bug fixes as you'd
expect.