You are not logged in.

#1 2010-08-04 15:25:39

ewaller
Administrator
From: Pasadena, CA
Registered: 2009-07-13
Posts: 20,308

Emacs 23, semantic, wxWidgets code completion problem

Hello All.

I am making yet another venture into Emacs.  I think this is the time I may stick with it.  This version has Semantic and Project functions previously provided by Cedet  included.  I am impressed by the ability to handle Makefiles, AutoMake, etc...  It integrates well with Git, Gdb and Make.  In general, it is a nice development environment that I am really starting to embrace. 

Except...

I cannot figure out how to get code completion to descend into the base classes I use for my subclasses.  If I subclass, for example WxFrame, code completion can see all public class members that I have defined in subclass definition, but it will not show me any base class public members.

I have turned on tag decorations and can see that some of the includes were not found in the system path.  No problem,  add the system path.  Now I get the indication they are not parsed.  Okay, parse all headers.  Semantic reports no missing or unparsed headers -- Still no base class members.  So I manually descend the header tree,  ensuring everything is resolved right down to the function for which I seek, but still no joy.

Spent some time with Google and a glass of wine and found some references to some minor issues with wxWidgets having to do with a combination of compiled headers and #ifdef hell.  There was even helpful lisp code showing how to recursively descend the header tree explicitly loading the headers for the backend you are using (example Gtk).  The problem is that Emacs 23 does not seem to define several of the functions from Cedet on which this Lisp code depends, and I am a Lisp newbie (at best).  Regardless, several functions and constants resolved to nil.

For fun, I tried loading both the cedet  from community and cedet-cvs from AUR.  Mostly, they made a mess of things and seemed to conflict with some of the project management functionality.

I am wide open to suggestions.  Thanks.

Oh, on a related note.  I tried Python using Semantic and it showed signs of trying to work, but I put no effort in to this (yet).  I can imagine semantic might not really work on weakly typed languages such as Python and Ruby.  Does this work, or is it a bridge to far?


Nothing is too wonderful to be true, if it be consistent with the laws of nature -- Michael Faraday
Sometimes it is the people no one can imagine anything of who do the things no one can imagine. -- Alan Turing
---
How to Ask Questions the Smart Way

Offline

Board footer

Powered by FluxBB