I have gotten to the point, with my libclang emodule, that I can
attempt to prototype code completion.
What I am wondering about, is what kind of code completion backend is
the most "appropriate"? I've already tried with pcomplete, but for
anything but shell command completions, and similar syntaxen, it's a
bit hard to bend to the task.
The emodule itself needs a lot more work to be complete, so I'm not
putting the code out there just yet. Even a partial binding will need
more work. In particular, I have not made any effort to interface
with the garbage collector yet. Other than that, I have solved all
technical problems I've come across, so most of this is now "just
work" and not hard.
I hope I can get some replies on completion backends. There seem to
be quite a lot of completion mechanisms that have accumulated over the
years. And I hope something out there is more appropriate than
Have all the fun,
XEmacs-Beta mailing list