[ESS] C-M-e and C-M-a: move to next end/previous beginning?

[Martin Maechler]

> I have opened an issue for this (https://github.com/emacs-ess/ESS/issues/251)
> 
> Looks like this behavior has been there since ESS inception, but I still think
> we need to comply with the standard behavior across emacs modes.

yes (2 x): It was always like that in ESS, but after switching
from C code editing, I had also found it "a pity" that ESS would
not allow to more easily move from function to function..

> 
> I personally never noticed this because I am mostly relying on paragraph
> navigation functions.
> 
>   Vitalie


I'm using C-M-a very often and C-M-e still often....
and I'll be very happy if the (older) issue about their
occasional failure --- notably when there's an internal
function() definition or even a *commented* internal function
--- can be solved soonish. 

Martin

> 
> >> On Mon, Oct 05 2015 12:54, Ramon Diaz-Uriarte wrote:
> 
> > Dear Vitalie,
> 
> > On Mon, 05-10-2015, at 11:24, Vitalie Spinu <spinuvit using gmail.com> wrote:
> >> There is an explicit check in the code for the point to be inside the
> >> function. So I guess this is a bug which is there by design.
> >>
> >> Fixing it would be very easy, but it would break the paragraph navigation
> >> functionality.
> >>
> >> The reason for the -or-paragraph is that as compared to elisp or C++ R code
> >> commonly contain paragraphs of code which are not defuns, like a series of
> >> one-liners which you want to evaluate as one block. I personally would like that
> >> C-M-a/e should perceive such paragraphs as defuns.
> >>
> >> These commands are meant to be consistent with C-c C-c
> >> (ess-eval-region-or-function-or-paragraph-and-step) and C-M-x
> >> (ess-eval-region-or-function-or-paragraph).
> >>
> 
> > Oh, I see. Of course, I actually use C-c C-c and C-M-x a lot and expect
> > that "series of one-liners" to work; I wasn't aware of the
> > relationship. And I also intuitively expect that paragraph behavior to work
> > with C-M-a/e.
> 
> > Thanks for the explanation.
> 
> >> One fix wold be to detect if two defuns follow one after another and then jump
> >> to previous/next function, otherwise jump to paragraphs in between.
> >>
> 
> > I think that is what I'd intuitively expect. But maybe everyone else is
> > happy with the current behavior.
> 
> > Best,
> 
> > R.
> 
> >>  Vitalie
> >>
> >>>> On Sun, Oct 04 2015 21:57, Ramon Diaz-Uriarte wrote:
> >>
> >>> Dear All,
> >>
> >>> With C-M-e ( ess-goto-end-of-function-or-para) and C-M-a
> >>> (ess-goto-beginning-of-function-or-para), once you reach the end/beginning
> >>> of a function, further C-M-e/a will not move you to the end/beginning of
> >>> the next/previous function.
> >>
> >>> Is it possible to somehow achieve this? Or the "otherwise go to the end of
> >>> paragraph" that C-M-e does precludes it?
> >>
> >>> Why do I want to do this? Yes, the ESS manual explicitly says "Don’t forget
> >>> the usual Emacs commands for moving over balanced expressions and
> >>> parentheses", so combining C-M-e with, say, C-M-n will do what I want, but
> >>> I'd rather just keep hitting "e" than alternating C-M-e and C-M-n. And I
> >>> also keep tripping on this as, for instance, C-M-e and C-M-a in C++ code
> >>> (c-end/beginning-of-defun) or elisp code (end/beginning-of-defun) keep
> >>> moving you to subsequent/previous ends/beginnings. 
> >>
> >>> Thanks,
> >>
> >>> R.
> 
> ______________________________________________
> ESS-help using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/ess-help