[antlr-interest] language independant grammars
jaluber at gmx.de
Fri Jul 27 14:14:06 PDT 2007
Loring Craymer wrote:
> I've been incredibly busy the past several months, but
> am now back in a position where I have time and
> research funding. I should have a current version
> posted to sourceforge soon; I am in the process of
> finishing a v2 to Yggdrasil translator (includes
> template conversion of v2 actions).
This looks like the answer to one of my suggestions at the beginning of
my ANTLR usage. The main thing what I would change besides the lack of
any examples is to adapt the syntax more to ANTLR 3. The reason is
simple: V3 will be around for a long time. Many users will be only
familiar with this version, so it will only a source of confusion.
People expect automatically that software environments like Yggdrasil
have the newest software versions under the hood. When they discover the
truth it will put them off. Even people like me, who would like to have
language independent versions of their grammars, won't be that willing
to switch if they have to change more than the actions.
Another thing I noticed is that you praise the use of diff3. Well, I'm
not happy to have to use such a tool. I prefer inheritance, partial
classes or marking functions etc. as programmer supplied, because these
methods (in combination or alone) allow it to generate the source code
freely without any intermediate steps. At least I don't know any
language, which doesn't have a feature to circumvent this step.
Hopefully my thoughts are useful to you.
More information about the antlr-interest