[antlr-interest] Re: C# ?

micheal_jor open.zone at virgin.net
Fri Apr 5 12:44:02 PST 2002


--- In antlr-interest at y..., Terence Parr <parrt at j...> wrote:
> I'm happy to host the C# code generator....we should probably 
decide on 
> whose to use though...sounds like a bunch of your folks are doing 
the 
> same thing.

AFAICT, there is currently one "open" versions - Eric's C# port for 
2.7.1. I now have two new ports working with 2.7.1 and 2.7.2a2 ANTLR 
versions. The plan is to release them asap so others can continue 
testing. The largest grammar that I have tested is the 'java' example 
(supports the -showtree option too).

Someone (was it Scott?) mentioned that they had developed an 
optimized C# port for 2.7.2a2 that had also been .NET-tified with 
regards to naming and uses of Properties for accessors/mutators etc. 
I expressed an interest in seeing the code when it was offered to the 
list but so far I've had no reply.

My intent in case anyone wondered was to backport the optimizations 
to my version which retains the Java-esque naming conventions. I 
actually like the idea of a .NET-tified version but I'm willing to go 
with Eric's point regarding debugging.

> Also a friend of mine at Microsoft has a J# version for 2.7.0...he 
> probably doesn't have time to update it for latest stuff.   He says 
> (reprinted with permission):

J# is good(-ish) for maintianing a kinda no-effort-
required "native" .NET version of ANTLR itself. I don't consider it a 
solution since it can't generate C# parsers without the C# codegen.

> Does this help anybody needing C# right away?  Might be able to use 
the 
> "bridge".

OK. I'll send what I have for 2.7.2a2 to you soon as it is done. I 
haven't had much joy with the debugging "bits". No reply to my 
previous post on the subject  ;-).

A few questions for you Ter if I may:

1. Is the debugging stuff intended for use while ANLTR itself is 
generating the Lexer/Parsers/Whatnots or is it for use with generated 
parsers?

2. In a similar vein, is ParseView for use with ANLTR or the stuff 
that ANTLR generates?

3. How does the debugging stuff work?. What is the thinking behind 
it. If anyone can explain it in enough detail, I'll have a go at 
fixing ParseView for 2.7.2a2.

Thanks in advance

Micheal



 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 



More information about the antlr-interest mailing list