Hi,
> Klaus Berndl has concocted some fixes for the C parser based on the
> 1.4.x sources. He will be checking them into the 1p4 branch at some
> point when he is satisfied.
Great! Thanks Klaus!
> One suggestion he had was to provide the language support in
> separate distributions so that languages can develop at a different
> speed from the core parser utilities.
>
> I suspect that the inversion stuff would help, but I haven't had
> many thoughts on a good way to deal with it. Perhaps ship everything
> in one tar file with language updates in a separate file=3F Perhaps it=
> doesn't occur often enough to worry about yet=3F Thoughts=3F
The cedet toolkit (and particularly semantic) is already sufficiently
complex to manage, and I wouldn't like to add more complexity to
handle such probably rare cases.
It would be a less time consuming strategy to improve our release
cycles and release the cedet toolkits more often.
Another difficulty is that the grammar framework has changed in
Semantic 2.0, and the C grammar in HEAD branch is in the new c.by
file. So if fixes are made in the c.bnf file in the v1p4 branch, we
probably need to apply them manually to the c.by file in the trunk :-(
Finally it is always possible for users to get particular fixed stuff
from the CVS repository (JDEE's users already do that when an
important fix is checked-in before the next release).
Just my opinion ;-)
David
|