matlab-emacs-discuss — Discuss Emacs/MATLAB integration

[Matlab-emacs-discuss] matlab-shell-region->script behavior with local functions

[Karthik C. <kar...@gm...>]

Hi,

I'm having trouble using matlab-shell-region->script when my script is structured as follows, and I call matlab-shell-run-region-or-line on the function call:

%---- 8< ----
%% Function call
func2(x)

%% Local functions
function y = func1(x)
  y = x * 2;
end  
  
function z = func2(x)
  y = func1(x);
  z = y + 1;
end
%---- 8< ----

When I call func2(x), matlab-shell-region->script copies over only func2 to the temporary file, even though func2 requires func1 to be defined.

The problem is that matlab-shell-region->script copies over only the functions that it finds in the region being evaluated. Since func1 is not part of this region, it's not copied over. From matlab-shell-region->script:

      (dolist (F functions)
	(save-excursion
	  (when (re-search-forward (semantic-tag-name F) nil t)
	    ;; Found, copy it in.
	    (let ((ft (matlab-semantic-tag-text F orig)))
	      (goto-char (point-max))
	      (insert "% Copy of " (semantic-tag-name F) "\n\n")
	      (insert ft)
	      (insert "\n%%\n")))))

The re-search-forward, which is run on the text being evaluated, is doing the filtering to func2 only. I got around this by simply copying over all local functions to the temp file:

      (dolist (F functions)
	(save-excursion
	  (let ((ft (matlab-semantic-tag-text F orig)))
	      ;; Copy over ALL local functions to the temp file.
              (goto-char (point-max))
	      (insert "% Copy of " (semantic-tag-name F) "\n\n")
	      (insert ft)
	      (insert "\n%%\n"))))

This probably makes the call to matlab-shell more expensive, but I think this should be the default behavior to avoid subtle failures like this.

Karthik

Re: [Matlab-emacs-discuss] [a problem in the branch: the return key]

[Uwe B. <ou...@ma...>]

>>> "EL" == Eric Ludlam <eri...@gm...> writes:

> Hi Uwe,
> Yes, that makes sense.  But also reminds me that way more than that should
> be in the news file.  Matlab mode used to do a lot of stuff that now has
> native support in emacs which changes how to configure things.  Sadly I
> didn't think to build a news like file for them other than all the check-in
> logs.  I could work on that but it will be a day or two as I'm doing family
> stuff this week.

Sure no hurry. I am occupied with family business as well. ;-)

Re: [Matlab-emacs-discuss] [a problem in the branch: the return key] (was: add-change-log-entry-other-window for matlab files)

[Eric L. <eri...@gm...>]

Hi Uwe,

Yes, that makes sense.  But also reminds me that way more than that should
be in the news file.  Matlab mode used to do a lot of stuff that now has
native support in emacs which changes how to configure things.  Sadly I
didn't think to build a news like file for them other than all the check-in
logs.  I could work on that but it will be a day or two as I'm doing family
stuff this week.

Some of the summary of changes is in an email I put together a month or so
ago, but not at the right detail level.

Eric

On Sun, Jun 20, 2021, 9:06 AM Uwe Brauer <ou...@ma...> wrote:

> >>> "EL" == Eric Ludlam <eri...@gm...> writes:
>
>
> > Sounds like a fine idea to me.  I posted a change to the hairyblocks
> > branch that will use the function name, or text after the cell start
> > as the 'defun' for add log.
>
> > The hairyblocks branch has been stable for a while with a lot of
> > developers at MW using it.  It should be ready to merge into the main
> > branch whenever you are ready for it.
>
>
>
> I am just running into a problem with the hairyblocks branch.
>
> I want to insert an empty line, in a matlab.m file hitting
> the return key, but I receive the following error
>
> ,----
> | Debugger entered--Lisp error: (void-function matlab-return)
> |   signal(void-function (matlab-return))
> |   help-fns--analyze-function(matlab-return)
> |   help-fns-function-description-header(matlab-return)
> |   describe-function-1(matlab-return)
> |   describe-key((([return] . [return])))
> |   funcall-interactively(describe-key (([return] . [return])))
> |   call-interactively(describe-key nil nil)
> |   command-execute(describe-key)
> `----
>
> Indeed my personal binding had (I don't know for how long)
>   (local-set-key [return]  'matlab-return)
>
>
> Eyeballing the changelog, I find
>   Remove matlab-comment-indent as a setting (use nil)
> │  (matlab-return-function, matlab-return, matlab-plain-ret)
> │  (matlab-indent-after-ret, matlab-indent-end-before-ret)
> │  (matlab-semicolon-on-return, matlab-indent-before-ret)
> │  (matlab-linefeed):
>
> Which I interpret that you removed matlab-return.
>
> So I suggest that before I merge that branch into master, we had into
> the NEWS file and entry were we tell which functions have been removed
> in order to avoid users complains.
>
> What do you think?
>
> Uwe
>
>

[Matlab-emacs-discuss] [a problem in the branch: the return key] (was: add-change-log-entry-other-window for matlab files)

[Uwe B. <ou...@ma...>]

>>> "EL" == Eric Ludlam <eri...@gm...> writes:


> Sounds like a fine idea to me.  I posted a change to the hairyblocks
> branch that will use the function name, or text after the cell start
> as the 'defun' for add log.

> The hairyblocks branch has been stable for a while with a lot of
> developers at MW using it.  It should be ready to merge into the main
> branch whenever you are ready for it.



I am just running into a problem with the hairyblocks branch.

I want to insert an empty line, in a matlab.m file hitting 
the return key, but I receive the following error 

,----
| Debugger entered--Lisp error: (void-function matlab-return)
|   signal(void-function (matlab-return))
|   help-fns--analyze-function(matlab-return)
|   help-fns-function-description-header(matlab-return)
|   describe-function-1(matlab-return)
|   describe-key((([return] . [return])))
|   funcall-interactively(describe-key (([return] . [return])))
|   call-interactively(describe-key nil nil)
|   command-execute(describe-key)
`----

Indeed my personal binding had (I don't know for how long)
  (local-set-key [return]  'matlab-return) 


Eyeballing the changelog, I find 
  Remove matlab-comment-indent as a setting (use nil)
│  (matlab-return-function, matlab-return, matlab-plain-ret)
│  (matlab-indent-after-ret, matlab-indent-end-before-ret)
│  (matlab-semicolon-on-return, matlab-indent-before-ret)
│  (matlab-linefeed):

Which I interpret that you removed matlab-return.

So I suggest that before I merge that branch into master, we had into
the NEWS file and entry were we tell which functions have been removed
in order to avoid users complains.

What do you think?

Uwe 

Re: [Matlab-emacs-discuss] add-change-log-entry-other-window for matlab files

[Uwe B. <ou...@ma...>]

>>> "EL" == Eric Ludlam <eri...@gm...> writes:

> Sounds like a fine idea to me.  I posted a change to the hairyblocks
> branch that will use the function name, or text after the cell start
> as the 'defun' for add log.


Great I just saw it, thanks!


> The hairyblocks branch has been stable for a while with a lot of
> developers at MW using it.  It should be ready to merge into the main
> branch whenever you are ready for it.

Right, let me just pull, compile, and test it this afternoon. Then I
will merge and push.

Re: [Matlab-emacs-discuss] add-change-log-entry-other-window for matlab files

[Eric L. <eri...@gm...>]

Sounds like a fine idea to me.  I posted a change to the hairyblocks 
branch that will use the function name, or text after the cell start as 
the 'defun' for add log.

The hairyblocks branch has been stable for a while with a lot of 
developers at MW using it.  It should be ready to merge into the main 
branch whenever you are ready for it.

Eric

On 6/20/21 2:46 AM, Uwe Brauer wrote:
>>>> "EL" == Eric Ludlam <eri...@gm...> writes:
> Hi Eric,
>
> Thanks for he quick reply
>> Hiya,
>> matlab-mode has defined `add-log-current-defun-function' for a long time.
> You are right, I did not test this carefully enough.
>
> However in lisp files, for examples defcustom is also recognised, so the
> question is whether it would make sense to enhance the relevant  matlab functionality.
>
> What I am thinking of is a good old script.
>
> I see two candidates here
>
>      1. A cell with a header
>
> Like
> %% Test the implementation of Euler's method
>
>      2. A direct execution of a function like
> [t,y] = ode45(f, intv, y0);
>
>
> What do you think about it?
>
> Regards
>
> Uwe

Re: [Matlab-emacs-discuss] add-change-log-entry-other-window for matlab files

[Uwe B. <ou...@ma...>]

>>> "EL" == Eric Ludlam <eri...@gm...> writes:

Hi Eric,

Thanks for he quick reply
> Hiya,
> matlab-mode has defined `add-log-current-defun-function' for a long time.

You are right, I did not test this carefully enough.

However in lisp files, for examples defcustom is also recognised, so the
question is whether it would make sense to enhance the relevant  matlab functionality.

What I am thinking of is a good old script.

I see two candidates here

    1. A cell with a header

Like 
%% Test the implementation of Euler's method

    2. A direct execution of a function like 
[t,y] = ode45(f, intv, y0);


What do you think about it?

Regards

Uwe 

Re: [Matlab-emacs-discuss] add-change-log-entry-other-window for matlab files

[Eric L. <eri...@gm...>]

Hiya,

matlab-mode has defined `add-log-current-defun-function' for a long time.

I'm using the hairyblocks branch, and tried the below command and it 
added the function name of a nested function to the record similar to below.

While the implementation of add-log-current-defun-function has changed 
in my branch, how add-log it is integrated with in matlab-mode hasn't 
really changed.

Eric

On 6/18/21 4:22 PM, Uwe Brauer wrote:
> Hi
>
> I use  add-change-log-entry-other-window a lot.
>
> The nice thing about elisp files is that you get entries like
>
> 	* matlab_init.el (my-matlab-shell-mode-hook): 296
>
> There is nothing similar for matlab.
>
> Does anybody have a hack for this feature?
>
> Regards
>
> Uwe
>
>
> _______________________________________________
> Matlab-emacs-discuss mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matlab-emacs-discuss

[Matlab-emacs-discuss] add-change-log-entry-other-window for matlab files

[Uwe B. <ou...@ma...>]

Hi

I use  add-change-log-entry-other-window a lot. 

The nice thing about elisp files is that you get entries like

	* matlab_init.el (my-matlab-shell-mode-hook): 296 

There is nothing similar for matlab. 

Does anybody have a hack for this feature?

Regards

Uwe 

[Matlab-emacs-discuss] when starts the gud menu in all buffers

[Uwe B. <ou...@ma...>]

Hi

I just realized when I start the matlab-shell, the gud menu pops up in
all buffers. The problem with this is that overwrites also certain 
keybindings of  other modes.

Any idea how to deal with this. I eye balled all my matlab relevant
files but could not find anything.

Andy comments?

Uwe 

Re: [Matlab-emacs-discuss] matlab-shell-save-and-go from another pwd

[Julien C. <jul...@mo...>]

Hi Uwe,

Sorry, this is my fault : this patch does not correct the fact that
matlab-shell-save-and-go can work with scripts from anywhere. It only
corrects that we can now run functions with 1 or more parameters.

Here is an example :

1 - Open matlab-shell
2 - mkdir test-matlab-shell-save-and-go/
3 - Create a new file test.m under this new directory :
  function test(p1)
  disp(['hello ' num2str(p1) '!'])
4 - Do a matlab-shell-save-and-go from this file.
5 - Enter (5) in the minibuffer prompt and press enter
6 - Without the patch you get one error:
  >> run('/home/jc/test-matlab-shell-save-and-go/test')
  Not enough input arguments.

  Error in test (line 2)
  disp(['hello ' num2str(p1) '!'])

  Error in run (line 91)
  evalin('caller', strcat(script, ';'));

7 - Apply the patch, and do the same. It works :
  >> addpath('/home/jc/test-matlab-shell-save-and-go/')
  >> test(5)
  hello 5!

Next time, I will provide an example immediately. I wrote this patch some
time ago (it was residing into my init.el file) and I was remembering wrong.
I put again the patch in attachment with a correct name.

Let me know if you note the difference.

Regards,

Julien


Le sam. 8 mai 2021 à 11:56, Uwe Brauer <ou...@ma...> a écrit :

> >>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <
> mat...@li...> writes:
>
> > Hi,
> > I'm posting this patch for improving the matlab-shell-save-and-go
> command.
> > I'm not sure to understand all the different parts/customization of this
> > function but, in my case (by default) I end up at the end of it : where
> it
> > tries to run the full path file using the matlab "run" command.
> > This works only if the matlab shell is in the current file directory.
> > Moreover, it works only with scripts (no matlab function).
> > Here is my proposition:
> > @@ -2166,2 +2166,4 @@ Similar to  `comint-send-input'."
> > -     ;; If not changing dir, maybe we need to use 'run' command instead?
> > -     (let ((cmd (concat "run('" dir fn-name "')")))
> > +     ;; If not changing dir, maybe we need to use addpath command before
> > +     (let ((cmd (concat "addpath('" dir "')")))
> > +       (matlab-shell-send-command cmd))
> > +     (let ((cmd (concat fn-name param)))
>
> > Instead of using the run matlab function, first, we addpath the file
> > directory and then, we run the file with parameters if any. This way, we
> > can run a script or a function from within any location.
> > I check that addpath at each time do not increase the path uselessly:
> > matlab does nothing if it is already into the path.
>
> I am currently testing your patch, can you tell me where I can note a
> difference to the unpatched matlab-shell.el?
>
> I run some tests and cannot spot a difference. An example would
> definitely help
>
> Uwe
>


-- 

Cordialement, Kind regards


*Julien Claisse*

Software Engineer

*Moduleus S.A.S,*
34 rue Pierre et Marie Curie
37100 Tours, France

+33 (0)2 47 49 08 16 <+33%202%2047%2049%2008%2016>

www.moduleus.com

Re: [Matlab-emacs-discuss] matlab-shell-save-and-go from another pwd

[Uwe B. <ou...@ma...>]

>>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <mat...@li...> writes:

> Hi,
> I'm posting this patch for improving the matlab-shell-save-and-go command.
> I'm not sure to understand all the different parts/customization of this
> function but, in my case (by default) I end up at the end of it : where it
> tries to run the full path file using the matlab "run" command.
> This works only if the matlab shell is in the current file directory.
> Moreover, it works only with scripts (no matlab function).
> Here is my proposition:
> @@ -2166,2 +2166,4 @@ Similar to  `comint-send-input'."
> -     ;; If not changing dir, maybe we need to use 'run' command instead?
> -     (let ((cmd (concat "run('" dir fn-name "')")))
> +     ;; If not changing dir, maybe we need to use addpath command before
> +     (let ((cmd (concat "addpath('" dir "')")))
> +       (matlab-shell-send-command cmd))
> +     (let ((cmd (concat fn-name param)))

> Instead of using the run matlab function, first, we addpath the file
> directory and then, we run the file with parameters if any. This way, we
> can run a script or a function from within any location.
> I check that addpath at each time do not increase the path uselessly:
> matlab does nothing if it is already into the path.

I am currently testing your patch, can you tell me where I can note a
difference to the unpatched matlab-shell.el?

I run some tests and cannot spot a difference. An example would
definitely help

Uwe  

Re: [Matlab-emacs-discuss] matlab-shell-save-and-go from another pwd

[Julien C. <jul...@mo...>]

Sure, no problem. Happy to help.

Le ven. 7 mai 2021 à 16:46, Uwe Brauer <ou...@ma...> a écrit :

> >>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <
> mat...@li...> writes:
>
> Hi Julien
>
>
> > Hi,
> > I'm posting this patch for improving the matlab-shell-save-and-go
> command.
> > I'm not sure to understand all the different parts/customization of this
> > function but, in my case (by default) I end up at the end of it : where
> it
> > tries to run the full path file using the matlab "run" command.
> > This works only if the matlab shell is in the current file directory.
> > Moreover, it works only with scripts (no matlab function).
>
> I am extremely interested in this patch, because lately I have to use
> this function quite a bit in different directories.
>
> May I ask you to send this patch as an attachment please?
>
> Uwe
>


-- 

Cordialement, Kind regards


*Julien Claisse*

Software Engineer

*Moduleus S.A.S,*
34 rue Pierre et Marie Curie
37100 Tours, France

+33 (0)2 47 49 08 16 <+33%202%2047%2049%2008%2016>

www.moduleus.com

Re: [Matlab-emacs-discuss] matlab-shell-save-and-go from another pwd

[Uwe B. <ou...@ma...>]

>>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <mat...@li...> writes:

Hi Julien


> Hi,
> I'm posting this patch for improving the matlab-shell-save-and-go command.
> I'm not sure to understand all the different parts/customization of this
> function but, in my case (by default) I end up at the end of it : where it
> tries to run the full path file using the matlab "run" command.
> This works only if the matlab shell is in the current file directory.
> Moreover, it works only with scripts (no matlab function).

I am extremely interested in this patch, because lately I have to use
this function quite a bit in different directories.

May I ask you to send this patch as an attachment please?

Uwe 

[Matlab-emacs-discuss] matlab-shell-save-and-go from another pwd

[Julien C. <jul...@mo...>]

Hi,

I'm posting this patch for improving the matlab-shell-save-and-go command.
I'm not sure to understand all the different parts/customization of this
function but, in my case (by default) I end up at the end of it : where it
tries to run the full path file using the matlab "run" command.
This works only if the matlab shell is in the current file directory.
Moreover, it works only with scripts (no matlab function).
Here is my proposition:
@@ -2166,2 +2166,4 @@ Similar to  `comint-send-input'."
-     ;; If not changing dir, maybe we need to use 'run' command instead?
-     (let ((cmd (concat "run('" dir fn-name "')")))
+     ;; If not changing dir, maybe we need to use addpath command before
+     (let ((cmd (concat "addpath('" dir "')")))
+       (matlab-shell-send-command cmd))
+     (let ((cmd (concat fn-name param)))

Instead of using the run matlab function, first, we addpath the file
directory and then, we run the file with parameters if any. This way, we
can run a script or a function from within any location.
I check that addpath at each time do not increase the path uselessly:
matlab does nothing if it is already into the path.

Regards,

*Julien Claisse*

[Matlab-emacs-discuss] Revised indentation engine available for matlab-mode

[Eric L. <eri...@gm...>]

Hi all,

John C and I had been working on a revised indentation engine for
matlab-mode back in March.  After a month of testing (which John has been
working on) we think is ready for more general testing and integration into
the mainline.

The new engine fixes a wide range of bugs and includes massive performance
improvements enabling large M files that used to take multiple hours to
indent to now fully indent in less than a minute.

You can access the new version on the 'hairyblocks' branch on sourceforge;
named for some rather challenging block indentation problems we solved.

Here are a few highlights of what is in the new branch:
* Comments and strings are now handled via syntax tables.  This means:
    * Faster font lock & indentation
    * Standard navigation & edit commands now works w/ strings & comments
    * Standard tools that operate on strings & comments now work w/ matlab
mode
    * More font lock color options for various types of strings and comments
    * List navigation (via (), [], or {}) via std commands now works.
* Argument, Properties, Methods, Events blocks now detect and indent
correctly.  This means:
    * matlab-mode detects 'arguments' and other keywords as a variable or
keyword depending on context
    * faster forward/backward block navigation.
    * electric indent for block keywords.
* Tested the indentation engine against a huge corpus of M code.  This
means:
    * Lots of indentation bug fixes for random edge cases.
    * Focused performance improvements for large files and classes.
    * Improved local test suite to keep things working.
* Mode overrides for sexp & defun detection.  This means:
    * special keybindings for forward/back sexp no longer needed
    * special keybindings for begin/end defun no longer needed
    * tools that use sexp navigation (kill, mark, etc) now work.
    * tools that use defun navigation (mark, narrow, etc) now work.
* Improved file end-type detection.  This means:
    * shows end-type of the file in mode line
    * auto-detects on save w/ fewer questions.
    * mlint will add 'end' in the correct place if an end is missing.
* Removed keybindings for 'return'.  This means:
    * global settings for behavior of RET now also apply to matlab-mode.
    * blank lines will have no trailing spaces after RET or during indent
region.
* Modernized font-lock by using newer font lock faces
* Update the version #

There are some downsides, however.  Several chunks of compatibility code
for Emacs version 19, 20, and probably XEmacs were removed along the way,
mostly because I don't have any versions of Emacs that old so had no way to
verify it continued to work as I updated the code.

For anyone who has the interest, give it a try and let us know how it
goes.  If all goes well, I propose this be merged into the master branch.

Thanks
Eric & John

Re: [Matlab-emacs-discuss] [your patch]

[Julien C. <jul...@mo...>]

Awesome, thanks!

I'm using doom emacs and it seems that doom fetches it from sourceforge :
jc@localhost:matlab-mode$ git remote -v
origin  https://git.code.sf.net/p/matlab-emacs/src (fetch)
So yes, I think you can.

Julien


Le mar. 4 mai 2021 à 16:45, Uwe Brauer <ou...@ma...> a écrit :

> >>> "JC" == Julien Claisse <jul...@mo...> writes:
>
> > Sure, here it is.
> > Regards
> > Julien
>
> Thanks, applied, committed and pushed.
>
>
> I realized that I have not pushed to the github mirror for a while:
> however little has been changed on master, most of the commits where on
> the new branches, which BTW, I could/should merge to master?
>


-- 

Cordialement, Kind regards


*Julien Claisse*

Software Engineer

*Moduleus S.A.S,*
34 rue Pierre et Marie Curie
37100 Tours, France

+33 (0)2 47 49 08 16 <+33%202%2047%2049%2008%2016>

www.moduleus.com

Re: [Matlab-emacs-discuss] [your patch]

[Uwe B. <ou...@ma...>]

>>> "JC" == Julien Claisse <jul...@mo...> writes:

> Sure, here it is.
> Regards
> Julien

Thanks, applied, committed and pushed. 


I realized that I have not pushed to the github mirror for a while:
however little has been changed on master, most of the commits where on
the new branches, which BTW, I could/should merge to master?

Re: [Matlab-emacs-discuss] [your patch] (was: emacsrunregion from a tramp opened file)

[Julien C. <jul...@mo...>]

Sure, here it is.

Regards
Julien

Le mar. 4 mai 2021 à 15:12, Uwe Brauer <ou...@ma...> a écrit :

> >>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <
> mat...@li...> writes:
>
> > Hi,
> > First, thanks to the developer team for maintaining this emacs package.
> I'm
> > using it almost every day.
>
> > I'm sending this mail because I want to suggest one improvement in the
> > emacsrunregion.m matlab function for being able to use it into a tramp
> > opened file (I'm also doing a lot of remote development on several
> targets).
>
> > This improvement is relatively straightforward, it adds the following
> lines
> > at the beginning of the function :
>
> > modified   toolbox/emacsrunregion.m
> > @@ -2,6 +2,12 @@ function emacsrunregion(file, startchar, endchar)
> >  % Run code from FILE between STARTCHAR and ENDCHAR.
> >  % Command sent by Emacs for run-cell & run-region functionality.
>
> > +    % Filter out emacs tramp file path prefix
> > +    trampMatch = regexp(file, {'/*:',':/'});
> > +    if (~isempty(trampMatch{1}))
> > +        file = file((trampMatch{2}+1):end);
> > +    end
> > +
> >      if ~exist(file,'file')
> >          error('You must save your region into a file accessible by
> MATLAB
> > process.');
> >      end
>
> The patch is damaged by the MTA, could you please send it again, as an
> attachment?
>


-- 

Cordialement, Kind regards


*Julien Claisse*

Software Engineer

*Moduleus S.A.S,*
34 rue Pierre et Marie Curie
37100 Tours, France

+33 (0)2 47 49 08 16 <+33%202%2047%2049%2008%2016>

www.moduleus.com

[Matlab-emacs-discuss] [your patch] (was: emacsrunregion from a tramp opened file)

[Uwe B. <ou...@ma...>]

>>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <mat...@li...> writes:

> Hi,
> First, thanks to the developer team for maintaining this emacs package. I'm
> using it almost every day.

> I'm sending this mail because I want to suggest one improvement in the
> emacsrunregion.m matlab function for being able to use it into a tramp
> opened file (I'm also doing a lot of remote development on several targets).

> This improvement is relatively straightforward, it adds the following lines
> at the beginning of the function :

> modified   toolbox/emacsrunregion.m
> @@ -2,6 +2,12 @@ function emacsrunregion(file, startchar, endchar)
>  % Run code from FILE between STARTCHAR and ENDCHAR.
>  % Command sent by Emacs for run-cell & run-region functionality.

> +    % Filter out emacs tramp file path prefix
> +    trampMatch = regexp(file, {'/*:',':/'});
> +    if (~isempty(trampMatch{1}))
> +        file = file((trampMatch{2}+1):end);
> +    end
> +
>      if ~exist(file,'file')
>          error('You must save your region into a file accessible by MATLAB
> process.');
>      end

The patch is damaged by the MTA, could you please send it again, as an
attachment?

Re: [Matlab-emacs-discuss] emacsrunregion from a tramp opened file

[Julien C. <jul...@mo...>]

Hi Eric,

Thanks for the feedback and your nice emacs package.

Yes, even if there are still missing features, TRAMP works really well and
I think there is not a lot of work for having almost all features of your
emacs package which work under it.
I will let you know about other patches as I come across issues.

Regards,

Julien

Le mar. 4 mai 2021 à 13:12, Eric Ludlam <eri...@gm...> a écrit :

> Hi,
>
> Thanks Julien for the patch.
>
> The nature of this patch seems good to me.  The workflow of Emacs and
> MATLAB on different machines is an interesting one, and it is nice that it
> can be made to work in such a simple way!
>
> Eric
>
> On Mon, May 3, 2021 at 12:06 PM Uwe Brauer <ou...@ma...> wrote:
>
>> >>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <
>> mat...@li...> writes:
>>
>> Hi Julien,
>>
>>
>> > Hi,
>> > First, thanks to the developer team for maintaining this emacs package.
>> I'm
>> > using it almost every day.
>>
>> Most of the work is done by Eric and John.
>>
>>
>> > I'm sending this mail because I want to suggest one improvement in the
>> > emacsrunregion.m matlab function for being able to use it into a tramp
>> > opened file (I'm also doing a lot of remote development on several
>> targets).
>>
>> > This improvement is relatively straightforward, it adds the following
>> lines
>> > at the beginning of the function :
>>
>> > modified   toolbox/emacsrunregion.m
>> > @@ -2,6 +2,12 @@ function emacsrunregion(file, startchar, endchar)
>> >  % Run code from FILE between STARTCHAR and ENDCHAR.
>> >  % Command sent by Emacs for run-cell & run-region functionality.
>>
>> > +    % Filter out emacs tramp file path prefix
>> > +    trampMatch = regexp(file, {'/*:',':/'});
>> > +    if (~isempty(trampMatch{1}))
>> > +        file = file((trampMatch{2}+1):end);
>> > +    end
>> > +
>> >      if ~exist(file,'file')
>> >          error('You must save your region into a file accessible by
>> MATLAB
>> > process.');
>> >      end
>>
>> > This way, file tramp path prefixes like "/scp:host@host_ip:" are
>> filtered
>> > out and the MATLAB remote shell can find the file without any trouble.
>>
>>
>> That looks ok to me, but since this is Eric's brainchild,
>> I think he should also agree.
>>
>>
>> > How can I submit this pull request ? Do I simply need to commit and
>> push ?
>>
>> Commit and push to where?
>>
>> The main development is done in sourceforge, github is just a mirror. I
>> am not sure whether you have write access to sourceforge I rather think
>> not.
>> I not a huge fan of pull request[1] although I know most people consider
>> it
>> today as a standard.
>>
>> Once we agreed on that patch I can apply it to Master, or Eric for that
>> matter.
>>
>> Regards
>>
>> Uwe
>>
>> Footnotes:
>> [1] I prefer to discuss patches on the mailing list and then apply them
>>      or to give a person write access to that he/she can push.
>>
>> _______________________________________________
>> Matlab-emacs-discuss mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matlab-emacs-discuss
>>
> _______________________________________________
> Matlab-emacs-discuss mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matlab-emacs-discuss
>


-- 

Cordialement, Kind regards


*Julien Claisse*

Software Engineer

*Moduleus S.A.S,*
34 rue Pierre et Marie Curie
37100 Tours, France

+33 (0)2 47 49 08 16 <+33%202%2047%2049%2008%2016>

www.moduleus.com

Re: [Matlab-emacs-discuss] emacsrunregion from a tramp opened file

[Uwe B. <ou...@ma...>]

>>> "EL" == Eric Ludlam <eri...@gm...> writes:

> Hi,
> Thanks Julien for the patch.

> The nature of this patch seems good to me.  The workflow of Emacs and
> MATLAB on different machines is an interesting one, and it is nice that it
> can be made to work in such a simple way!

Ok then. I apply his patch to master and push.

Re: [Matlab-emacs-discuss] emacsrunregion from a tramp opened file

[Eric L. <eri...@gm...>]

Hi,

Thanks Julien for the patch.

The nature of this patch seems good to me.  The workflow of Emacs and
MATLAB on different machines is an interesting one, and it is nice that it
can be made to work in such a simple way!

Eric

On Mon, May 3, 2021 at 12:06 PM Uwe Brauer <ou...@ma...> wrote:

> >>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <
> mat...@li...> writes:
>
> Hi Julien,
>
>
> > Hi,
> > First, thanks to the developer team for maintaining this emacs package.
> I'm
> > using it almost every day.
>
> Most of the work is done by Eric and John.
>
>
> > I'm sending this mail because I want to suggest one improvement in the
> > emacsrunregion.m matlab function for being able to use it into a tramp
> > opened file (I'm also doing a lot of remote development on several
> targets).
>
> > This improvement is relatively straightforward, it adds the following
> lines
> > at the beginning of the function :
>
> > modified   toolbox/emacsrunregion.m
> > @@ -2,6 +2,12 @@ function emacsrunregion(file, startchar, endchar)
> >  % Run code from FILE between STARTCHAR and ENDCHAR.
> >  % Command sent by Emacs for run-cell & run-region functionality.
>
> > +    % Filter out emacs tramp file path prefix
> > +    trampMatch = regexp(file, {'/*:',':/'});
> > +    if (~isempty(trampMatch{1}))
> > +        file = file((trampMatch{2}+1):end);
> > +    end
> > +
> >      if ~exist(file,'file')
> >          error('You must save your region into a file accessible by
> MATLAB
> > process.');
> >      end
>
> > This way, file tramp path prefixes like "/scp:host@host_ip:" are
> filtered
> > out and the MATLAB remote shell can find the file without any trouble.
>
>
> That looks ok to me, but since this is Eric's brainchild,
> I think he should also agree.
>
>
> > How can I submit this pull request ? Do I simply need to commit and push
> ?
>
> Commit and push to where?
>
> The main development is done in sourceforge, github is just a mirror. I
> am not sure whether you have write access to sourceforge I rather think
> not.
> I not a huge fan of pull request[1] although I know most people consider it
> today as a standard.
>
> Once we agreed on that patch I can apply it to Master, or Eric for that
> matter.
>
> Regards
>
> Uwe
>
> Footnotes:
> [1] I prefer to discuss patches on the mailing list and then apply them
>      or to give a person write access to that he/she can push.
>
> _______________________________________________
> Matlab-emacs-discuss mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matlab-emacs-discuss
>

Re: [Matlab-emacs-discuss] emacsrunregion from a tramp opened file

[Uwe B. <ou...@ma...>]

>>> "JCvM" == Julien Claisse via Matlab-emacs-discuss <mat...@li...> writes:

Hi Julien,


> Hi,
> First, thanks to the developer team for maintaining this emacs package. I'm
> using it almost every day.

Most of the work is done by Eric and John.


> I'm sending this mail because I want to suggest one improvement in the
> emacsrunregion.m matlab function for being able to use it into a tramp
> opened file (I'm also doing a lot of remote development on several targets).

> This improvement is relatively straightforward, it adds the following lines
> at the beginning of the function :

> modified   toolbox/emacsrunregion.m
> @@ -2,6 +2,12 @@ function emacsrunregion(file, startchar, endchar)
>  % Run code from FILE between STARTCHAR and ENDCHAR.
>  % Command sent by Emacs for run-cell & run-region functionality.

> +    % Filter out emacs tramp file path prefix
> +    trampMatch = regexp(file, {'/*:',':/'});
> +    if (~isempty(trampMatch{1}))
> +        file = file((trampMatch{2}+1):end);
> +    end
> +
>      if ~exist(file,'file')
>          error('You must save your region into a file accessible by MATLAB
> process.');
>      end

> This way, file tramp path prefixes like "/scp:host@host_ip:" are filtered
> out and the MATLAB remote shell can find the file without any trouble.


That looks ok to me, but since this is Eric's brainchild, 
I think he should also agree.


> How can I submit this pull request ? Do I simply need to commit and push ?

Commit and push to where? 

The main development is done in sourceforge, github is just a mirror. I
am not sure whether you have write access to sourceforge I rather think
not.
I not a huge fan of pull request[1] although I know most people consider it
today as a standard.

Once we agreed on that patch I can apply it to Master, or Eric for that
matter.

Regards

Uwe 

Footnotes:
[1] I prefer to discuss patches on the mailing list and then apply them
     or to give a person write access to that he/she can push.

[Matlab-emacs-discuss] emacsrunregion from a tramp opened file

[Julien C. <jul...@mo...>]

Hi,

First, thanks to the developer team for maintaining this emacs package. I'm
using it almost every day.

I'm sending this mail because I want to suggest one improvement in the
emacsrunregion.m matlab function for being able to use it into a tramp
opened file (I'm also doing a lot of remote development on several targets).

This improvement is relatively straightforward, it adds the following lines
at the beginning of the function :

modified   toolbox/emacsrunregion.m
@@ -2,6 +2,12 @@ function emacsrunregion(file, startchar, endchar)
 % Run code from FILE between STARTCHAR and ENDCHAR.
 % Command sent by Emacs for run-cell & run-region functionality.

+    % Filter out emacs tramp file path prefix
+    trampMatch = regexp(file, {'/*:',':/'});
+    if (~isempty(trampMatch{1}))
+        file = file((trampMatch{2}+1):end);
+    end
+
     if ~exist(file,'file')
         error('You must save your region into a file accessible by MATLAB
process.');
     end

This way, file tramp path prefixes like "/scp:host@host_ip:" are filtered
out and the MATLAB remote shell can find the file without any trouble.

How can I submit this pull request ? Do I simply need to commit and push ?

*Julien Claisse*