Re: [babel] feature request: debug messages

[Andreas Leha <andreas.leha@med.uni-goettingen.de>]

Hi Eric,

Eric Schulte <schulte.eric@gmail.com> writes:

> Andreas Leha <andreas.leha@med.uni-goettingen.de> writes:
>
>> Hi Eric,
>>
>>
>> Eric Schulte <schulte.eric@gmail.com> writes:
>>
>>> Hi Andreas,
>>>
>>> This should be easy to turn on or off using the newly introduced
>>> :prologue and :epilogue header arguments.  See the manual and the
>>> following example.
>>>
>>> #+Title: debug messages
>>> #+Property: session *R*
>>> #+Property: prologue (format "print(\"entering %s\")" (get-current-name))
>>>
>>> An elisp block to simplify the =:prologue= definition.
>>> #+begin_src emacs-lisp
>>>   (defun get-current-name ()
>>>     (save-excursion
>>>       (goto-char org-babel-current-src-block-location)
>>>       (while (and (forward-line -1)
>>>                   (looking-at org-babel-multi-line-header-regexp)))
>>>       (when (looking-at org-babel-src-name-w-name-regexp)
>>>         (org-no-properties (match-string 3)))))
>>> #+end_src
>>>
>>> Two blocks with simple assignments.
>>>
>>> #+name: block-1
>>> #+begin_src R
>>>   x <- 2 + 2
>>> #+end_src
>>>
>>> #+name: block-2
>>> #+begin_src R
>>>   y <- x + x
>>> #+end_src
>>>
>>> Execute the whole buffer =C-c C-v b= to see the prologue in action.
>>>
>>> Andreas Leha <andreas.leha@med.uni-goettingen.de> writes:
>>>
>>>> Hi all,
>>>>
>>>> I would love to see messages like 'entering block foo...' and
>>>> '...leaving block foo' printed to my R console.  This would be very
>>>> handy when I evaluate a subtree (C-c C-v s) with a lot of #+call lines
>>>> and some lengthy ones.
>>>>
>>>> I know that
>>>> (1) I could implement that myself at in the source blocks.  But I would
>>>>     love if orgmode did that for me
>>>> (2) Such messages are already printed to the emacs *Messages* buffer.
>>>>     But that buffer might not be visible and I can not switch to it,
>>>>     without interrupting the evaluation.  Anyway it would be much nicer
>>>>     to see that output together with the other output, that my code
>>>>     generates.
>>>>
>>>>
>>>> In essence it would be very helpful, if there was a variable
>>>> org-babel-print-debug-messages (or org-babel-debug-level...) which if
>>>> non-nil would cause that messages to be printed.  Or is there somewhere
>>>> already?
>>>>
>>>> Regards,
>>>> Andreas
>>>>
>>>>
>>>>
>>>>
>>
>>
>> thanks for the quick answer!  The :prologue and :epilogue header
>> arguments have indeed slipped my attention and they look really
>> interesting!  I see, that they are documented, but somehow, they seem to
>> not get their headline and TOC entry?
>>
>> I have three problems with your example, though:
>> 1) It does not run
>> 2) It does not work
>> 3) It won't be usable for 'my' epilogue, correct?
>> ;-)
>>
>
> Ah! My fault.  I had to add prologue and epilogue support to ob-R.el
> when working through the example I sent, but then I forgot to commit
> that support to Org-mode.  I've just pushed up that commit, and

Indeed, that fixed it, now it works.

> re-worked my example file to avoid the issue of prologue being applied
> to the emacs-lisp code block (using the very nice and also new
> language-specific PROPERTY header arguments).

Thanks.  So, I did the right thing here.

>
> Finally, I don't use epilogues in the example because (as the last thing
> evaluated) they would override the code block results.

Ok.  As I thought.  No "leaving code block" message, then.


>
> Hopefully the following:
> 1. will run
> 2. will work
> 3. will be usable

Yes.
Yes.
Yes.

I still believe, that there is a bug in the manual, where :epilogue
and :prologue are described.

Tanks a lot for your help and that really nice feature!

Cheers,
Andreas