Re: [Matlab-emacs-discuss] emacsinit/ using emacs as an external Matlab editor

[emacs u. <use...@gm...>]

hi Eric, I tried using matlab-shell again.  Looks great.

A couple of questions: I wasn't able to get cedet to work as it gives the
following error message when I edit a Matlab file.  I tried both the cedet
in emacs 24.3 and downloading it from sourceforce, version 1.1.  I don't
know if this is a result of cedet not working, but when I get an error
message in Matlab-shell the line number pointing to the problem in the
Matlab code does not seem to appear as a clickable link.

I wasn't able to see how to report issues/ bugs in the Matlab-emacs
sourceforce page so am sending this here.

best, E

Debugger entered--Lisp error: (void-variable matlab-mode)
  byte-code("..." [matlab-mode
semantic-matlab-dependency-system-include-path
semantic-matlab-root-directory semantic-matlab-system-paths-include tmp0
mode defcustom-mode-local-semantic-dependency-system-include-path mapcar
#[(cur) "\302! P\207" [semantic-matlab-root-directory cur
file-name-as-directory] 2] "The system include paths from MATLAB."
semantic-idle-summary-format-matlab-mode mode-local-bind
semantic-idle-summary-function (mode-variable-flag t variable-documentation
"Search project files, then search this omniscience database.\nIt is not
necessary to to system or recursive searching because of\nthe omniscience
database.") mode-local-map-mode-buffers #[nil "..." [tmp0
semantic-idle-summary-function make-local-variable] 2] put symbol-name nil
intern-soft mode-local-symbol-table mode-local-parent derived-mode-parent
variable-documentation "Function to use when displaying tag information
during idle time." symbol bind name mode-local-active-mode major-mode] 6)
  semantic-default-matlab-setup()
  run-hooks(matlab-mode-hook)
  matlab-mode()
  set-auto-mode-0(matlab-mode nil)
  set-auto-mode()
  normal-mode(t)
  after-find-file(nil t)
  find-file-noselect-1(#<buffer analyze_flow_solve.m>
"~/Internal-waves-DIA/my_flow_solve/Postprocessing/analyze_flow_solve.m"
nil nil
"~/Internal-waves-DIA/my_flow_solve/Postprocessing/analyze_flow_solve.m"
(22407275 16777218))

find-file-noselect("~/Internal-waves-DIA/my_flow_solve/Postprocessing/analyze_flow_solve.m"
nil nil t)

find-file("~/Internal-waves-DIA/my_flow_solve/Postprocessing/analyze_flow_solve.m"
t)
  call-interactively(find-file)
  find-file-at-point()
  call-interactively(find-file-at-point nil nil)



On Mon, Jul 8, 2013 at 12:29 PM, Eric Ludlam <Eri...@ma...>wrote:

>  Hi,****
>
> ** **
>
> I think I understand a bit better.  Clicking on the “line 3” which is
> supposed to call ‘opentoline’ is not calling the shadowed version. ****
>
> ** **
>
> I gave this a try and discovered the same problem with newer MATLABs.  I
> don’t know what the problem is.  Mostly it looks like the command window
> isn’t actually running the MATLAB command in the link since the debugger
> won’t stop there either.  I’ll need to look into it to see what’s going on.
> ****
>
> ** **
>
> In the meantime, you may need to use the matlab-shell in Emacs to get
> those links to work.****
>
> ** **
>
> Eric****
>
> ** **
>
> *From:* emacs user [mailto:use...@gm...]
> *Sent:* Tuesday, July 02, 2013 5:17 PM
> *To:* Eric Ludlam
> *Cc:* mat...@li...
> *Subject:* Re: [Matlab-emacs-discuss] emacsinit/ using emacs as an
> external Matlab editor****
>
> ** **
>
> Hi, thanks for the prompt response, I think I didn't explain that very
> clearly, let me try again.  I have an init.m file:****
>
> ** **
>
> addpath('~/.emacs.d/matlab-emacs/toolbox','-begin');****
>
> rehash;****
>
> emacsinit('/Applications/Emacs.app/Contents/MacOS/bin/emacsclient');****
>
> ** **
>
> and then a opentoline_test.m file:****
>
> a=zeros(3,4);****
>
> b=zeros(3,3);****
>
> c=a+b;****
>
> ** **
>
> in Matlab I do:****
>
> ** **
>
> >> which opentoline****
>
> /Applications/MATLAB_R2013a.app/toolbox/matlab/codetools/opentoline.m****
>
> >> init****
>
> >> which opentoline****
>
> /Users/eli/.emacs.d/matlab-emacs/toolbox/opentoline.m****
>
> ** **
>
> so the correct opentoline seems to be shadowing.  Next,****
>
> ** **
>
> >> opentoline_test****
>
> Error using  + ****
>
> Matrix dimensions must agree.****
>
> ** **
>
> Error in opentoline_test (line 3)****
>
> c=a+b;****
>
>  ****
>
> now if I click on "line 3" I get the matlab editor, if instead, I say ****
>
> ** **
>
> >> opentoline('opentoline_test.m',3,0)****
>
> ** **
>
> then emacs is correctly invoked.****
>
> ** **
>
> does this help, or did I simply misunderstand your response?  thanks
> again, best, Eli****
>
> ** **
>
> ** **
>
> On Tue, Jul 2, 2013 at 4:46 PM, Eric Ludlam <Eri...@ma...>
> wrote:****
>
> Hi,****
>
>  ****
>
> Not all the ‘edit’ like functions obey the preference to use an alternate
> editor.  If you look in the toolbox directory that came with matlab.el, you
> will see “opentoline.m” which shadows the one that comes with matlab.
> Chances are that netcdf.open (which I am not familiar with at all) might
> need something similar.   If you look at that fcn, it calls some lib
> function that you might be able to shadow in order to get what you want.
> Just stick it in the toolbox directory for Emacs, and you’ll be all set.
> If it isn’t related to some MATLAB toolbox, perhaps we can add the shadow
> to the matlab-emacs repository.****
>
>  ****
>
> Good Luck****
>
> Eric****
>
>  ****
>
> *From:* emacs user [mailto:use...@gm...]
> *Sent:* Monday, July 01, 2013 3:01 PM
> *To:* mat...@li...
> *Subject:* [Matlab-emacs-discuss] emacsinit/ using emacs as an external
> Matlab editor****
>
>  ****
>
> hi, ****
>
>  ****
>
> I am trying the above using Emacs 24.3 on Mac OSX, using a fresh download
> of matlab-emacs and Matlab 2013a.  when I follow the install directions,
> trying to click on the line number in something like this:****
>
>  ****
>
> Error using netcdf.open (*line 45*)****
>
>  ****
>
> brings up the matlab editor instead of emacsclient.  if I try****
>
>  ****
>
> >> opentoline('myfile.m',1477,0)****
>
>  ****
>
> emacsclient is invoked and goes to the right line etc.****
>
>  ****
>
> Any idea what I may be doing wrong?  additional information I can povide?
> ****
>
> many thanks...****
>
>  ****
>
> ** **
>