[ESS] Ever consider changing indentation on #?

[Paul Johnson]

On Thu, May 4, 2017 at 3:51 AM, Stephen Eglen <sje30 using cam.ac.uk> wrote:
>
> hi Paul,
>
>> "what the Hell, the folding is borked!". This is more and more common
>> with the adoption of markdown-style commentary within R files.
>
> can you give an example of the markdown-style commentary?
>
>

Here is an R file that has "Rstudio sections" that begin with #.  I just tested
with their gui and the magic recipe for howto is written below.  It is
something like
Emacs org mode, I suppose. Code folding. This is an example of a
file that gets "destroyed" by Emacs/ESS indentation

## Paul Johnson
## 20170504

## Open this file in Rstudio, you'll see what I mean

# Here is a section from Code-> insert section ----------------------------

x <- rnorm(100)
y <- rnorm(100)


## The effect of "Code -> Insert section" is to create a left aligned single #.
## Apparently, it follows with dots out to right edge.

# Here is a second section from Code -> insert section --------------------

z <- function(a, b, c){
    y <-3
    y + 7
}

##  After this, in editor Rstudio, then the sections can be folded (hidden)


# Third section ----

## How to create a section without using Rstudio menu?
## TO create this section, I did not use Rstudio menu. I insert
## 1. one #
## 2. a space
## 3. some words
## 4. Four dashes

## the # at begin and dashes at end is signal to R studio this is a section.


I understand I can fix my init.el go avoid this, but it would be nicer
for my purpose
if this were in ESS itself, because I keep telling students to use it
and they are put off/discouraged
by this.

I'm not an Rstudio user, but try to cooperate with some of them :)

pj

>> Until now, I've said "too bad" when they complain about their code,
>> but now I'm thinking we ought to consider changing the behavior of
>> Emacs/ESS.  Does it really really need to do # indentation that way?
>> Why?  I've always thought it is odd. And counterintuitive. In what
>> backwards world would one suppose # gets the most indent, while ###
>> gets none?. I don't see any benefit in that # indentation that way. I
>> do see a big benefit in the ## indent at the code level. I'm only
>> quibbling about #.
>
> just for background context, ESS follows the Emacs convention for
> comments (### for column 0, ## for same level, # for col 32 or more), that is
> used widely in elisp.  It is probably fair to say though that these
> conventions are not widely known outside of Emacs.
>
>> Would you ever consider changing the ESS default for R files so # is
>> flush left?  Could I offer you some cash to consider it?
>>
>> I understand I *could* learn code for my emacs.init to do that, but it
>> seems like a change that would be broadly beneficial to new Emacs &
>> ESS users.  I preach the message about Emacs to the students here, but
>> little quirks like this seem like an unnecessary hassle.
>
> Its likely to be fixable, but first it might be worth hearing Martin's
> views on this, as he may deal with differing commenting styles a lot in
> R-core.
>
> Stephen



-- 
Paul E. Johnson   http://pj.freefaculty.org
Director, Center for Research Methods and Data Analysis http://crmda.ku.edu

To write to me directly, please address me at pauljohn at ku.edu.