Hi again,

I was able to solve my problem, it seems that everything was related to the #include statements in the source-code and now after executing C-u M-x bovinate almost all the dependencies are solved. However, every time I initialize a new emacs session I'm obligated to call bovinate, is there a possible way to do it only once? 

Regards,

Pamela



On Mon, May 13, 2013 at 9:41 AM, Pamela Carreño <carrenopamela@gmail.com> wrote:
Hi,

thank you for your replies. I tried using 

:file (expand-file-name "~/Documents/PhD/Code/Kd-Trees_NNS/Flann/CMakeLists.txt") as LLuis suggested, however it did not work and there are still several unsolved includes.

I tried to use semantic-add-system-include for every directory like this

(defun my-flann-semantic-hook ()
(setq FLANN-base-dir "~/Documents/PhD/Code/Kd-Trees_NNS/Flann/")
(semantic-add-system-include (concat FLANN-base-dir "test/") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "examples/") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "src/cpp/flann") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "src/cpp/flann/algorithms") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "src/cpp/flann/io") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "src/cpp/flann/mpi") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "src/cpp/flann/nn") 'c++-mode)
(semantic-add-system-include (concat FLANN-base-dir "src/cpp/flann/util") 'c++-mode))
(add-hook 'semantic-init-hooks 'my-flann-semantic-hook)

and although Semantic-c-describe-environment now list every path... It did not work at all.

What else can I try? 

Thanks,

Pamela


On Sun, May 12, 2013 at 11:36 PM, Eric M. Ludlam <eric@siege-engine.com> wrote:
On 05/07/2013 04:19 PM, Lluís wrote:
Pamela Carreño writes:

Hi,
I'm using ede-cpp-root-project to create an ede-project from existing files.

(ede-cpp-root-project "FLANN"
:name "Flann1.8"
:file "~/Documents/PhD/Code/Kd-Trees_NNS/Flann/CMakeLists.txt"

I think EDE does not expand filenames for you (maybe that should be implemented,
or at least clearly documented). Try this instead:

     :file (expand-file-name "~/Documents/PhD/Code/Kd-Trees_NNS/Flann/CMakeLists.txt")


I checked, and I'm pretty sure the :file is expanded.  For example, see this from ede/cpp-root.el:

  (let ((f (expand-file-name (oref this :file))))
...
    (oset this :file f)

Is there something else causing problems?


BTW, this reminds me that all my 'ede-cpp-root-project' are guarded with a
'file-exists-p', as EDE otherwise throws out an error complaining about
non-existing files.

Is there a reason for this behaviour? Wouldn't it be more comfortable to check
that inside EDE?

What do you propose EDE do if the incoming file doesn't exist?  Wouldn't it still throw an error, or do you think it needs a better error?

Thanks
Eric



--
Pamela Carreño Medrano
PhD. candidate in Computer Science
IRISIA Lab/ Seaside team
University of Southern Brittany - UBS
Vannes, France