[Matlab-emacs-discuss] [SOLVED] (was: make test fails because of commit 864ff8a925a28f)

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

>>> "UB" == Uwe Brauer <ou...@ma...> writes:

> Hi 

> currently make lisp works but 
> make test fails with the error 


Please ignore this message. I caused this problem by playing around with too many different ways to define 

 matlab-mode-version 




> ebugger entered--Lisp error: (void-variable matlab-mode-version)
>   (string= package-version matlab-mode-version)
>   (not (string= package-version matlab-mode-version))
>   (if (not (string= package-version matlab-mode-version)) (progn (user-error "Version from matlab-mode.el \";; Version: %s\" != ma..." package-version matlab-mode-version)))
>   (let ((package-version (let ((temp-buffer (generate-new-buffer " *temp*" t))) (save-current-buffer (set-buffer temp-buffer) (unwind-protect (progn (insert-file-contents "../matlab-mode.el") (if ... ...) (match-string 1)) (and (buffer-name temp-buffer) (kill-buffer temp-buffer))))))) (if (not (string= package-version matlab-mode-version)) (progn (user-error "Version from matlab-mode.el \";; Version: %s\" != ma..." package-version matlab-mode-version))))
>   metest-check-version()
>   metest-all-syntax-tests()
>   command-line-1(("--eval" "(setq debug-on-error t)" "-l" "metest.el" "-e" "metest-all-syntax-tests"))
>   command-line()
>   normal-top-level()

> make[1]: *** [Makefile:35: modetests] Error 255
> make[1]: Leaving directory '/home/oub/ALLES/src/src-emacs-pkg/Matlab/Github/SSH/emacs-matlab-hg/tests'
> make: *** [Makefile:86: .tests.tstamp] Error 2




> The error was caused by commit 864ff8a925a28f (from the fill-fix branch
> which I merged into default and did not notice the problem)

> The problem is this change in metest.el


> +(defun metest-check-version ()
> +  "Validate matlab-mode version numbers are consistent."
> +  (let ((package-version (with-temp-buffer
> +                           (insert-file-contents "../matlab-mode.el")
> +                           (when (not (re-search-forward
> +                                       "^;; Version: \\([0-9]+\\.[0-9]+\\)[ \t]*$" nil t))
> +                             (user-error "Failed to find version in ../matlab-mode.el"))
> +                           (match-string 1))))
> +    (when (not (string= package-version matlab-mode-version))
> +      (user-error "Version from matlab-mode.el \";; Version: %s\" != matlab-mode-version %s"
> +                  package-version matlab-mode-version))))
> +
> +


> John you introduced this.

> I think it is time to discuss several things

>     1. Should we keep the function matlab-show-version (for matlab-shell)

>     2. Shall we have the defconst 

> (defconst matlab-mode-version "6.2"
>   "Current version of MATLAB(R) mode.")
   
>     Either in matlab.el or matlab-mode.el?

> I had to revert to commit 

> B2baf6b106587480 to reintroduce the matlab-mode-version, otherwise make
> lisp fails which is a show-stopper for ELPA and MELPA.

> I hesitate to revert metest.el before we discuss its implication

> I think we should settle this as soon as possible. I will put this as an
> issue in github as well

> Uwe 

-- 
I strongly condemn Hamas heinous despicable pogroms/atrocities on Israel
I strongly condemn Putin's war of aggression against Ukraine.
I support to deliver weapons to Ukraine's military. 
I support the EU and NATO membership of Ukraine.