Updated function works in Linux terminal in AWS SageMaker.

In R session, print command returns 19775.
But if I start a new R session, same print command returns 24904.  
And so on

Sent from my iPhone

> On Sep 10, 2025, at 11:01 AM, Tyler Smith <[email protected]> wrote:
> 
> On Wed, Sep 10, 2025, at 6:42 AM, Naresh Gurbuxani wrote:
>> It does not work in Linux terminal emacs run inside AWS SageMaker.  
>> Here it needs to open
>> /home/ec2-user/anaconda3/envs/R/lib/R/doc/html/index.html
> 
> This is interesting. If it needs to open a local file, that suggests the http 
> server is not started. Can you tell me what the output of 
> `print(tools::startDynamicHelp(NA))` (run from within R) is in that 
> situation? The new version of my function (below) may fix this.
> 
>> On Mac emacs GUI, this function does not work with default settings.  
>> This is the message displayed in in *eww* buffer:
>> Loading http://127.0.0.1:ting httpd help server ... done
>> [1] 23824/doc/html/index.html…
> 
> This is helpful! My original code didn't account for the extra text emitted 
> when the http server is started the first time. The following should fix that:
> 
> ```
> (defun ess-eww ()
>  (interactive)
>  (let ((BUF (generate-new-buffer "URL"))
>        (PORT))
>    (ess-command "print(tools::startDynamicHelp(NA))" BUF)
>    (with-current-buffer BUF
>      (setq PORT (buffer-substring (search-forward-regexp "\\[1\\] ") 
> (point-max)))
>      (message (buffer-substring (point-min) (point-max))))
>    (kill-buffer BUF)
>    (eww (concat "http://127.0.0.1:"; PORT "/doc/html/index.html"))))
> ```
> 
> This may also fix the problem in the Linux terminal. It works here using 
> terminal Emacs (on Debian).
> 
> Question for the ESS devs - is there a simpler way to get text from the R 
> process into Elisp? I've used `ess-command`, and then collected the relevant 
> info out of the buffer the result is printed to. This requires a bit of 
> fussing to get to the value I want (via `search-forward-regexp`). That might 
> be brittle. If I could directly transfer the first value of an R vector into 
> an Elisp variable that would be better.
> 
> - tyler
> 
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

Reply via email to