>>> "Berndl, Klaus" <klaus.berndl@...> seems to think that:
>>I'm happy to announce that the python parser has been fixed up
>>sufficiently well such that latest CVS version of ECB works quite well
>>with python files. For example, ECB shows member functions and
>>variables of python classes just it does for C++ classes.
>
>Sounds good..... my interest is mor ethe ECB-side than the python-side ;-),
>but i must say, the more parsers available for semantic the better for
>ECB... ;-)
>
>But what sounds really good: It seems that you can use ECB without changes
>with forthcoming semantic 2.0betaX, right? This would be fine, if there are
>no ECB changes necessary to work proper with the next semantic generation.
[ ... ]
Hi Klaus,
I think of semantic as having 3 interfaces. One interface is what
the user uses, and may put in their .emacs file. I don't usually
mind changing this API.
Another is the interface developers use to support different
languages. The language part (The .bnf, .by, or .wy files)
hopefully will not change drastically once written, but other parts
are changing drastically in semantic 2.0.
Lastly, there is the API applications use to get tokens out of
semantic. This API has not changed at all, except, perhaps, to be
extended.
Internally, of course, functions and structures will change on a
regular basis. (Internal parsing functions, lexical behavior, and
even the token structure.) As long as ECB doesn't use the tokens'
internal structure, it should be forward compatible for a long time.
Have fun
Eric
--
Eric Ludlam: zappo@..., eric@...
Home: http://www.ludlam.net Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net GNU: http://www.gnu.org
|