org-babel: Bugs with inline src_* blocks

org-babel: Bugs with inline src_* blocks

[Paul Sexton]

I am experiencing a couple of significant bugs with inline src blocks in 
org-babel -- ie blocks of the form src_LANG{EXPRESSION}. I am using the
development version of org, checked out a few days ago. 

Pressing C-c C-c with the cursor on such a block is supposed to evaluate it and 
echo the result to the minibuffer. However in recent versions of org (the last 
3 months or so) this behaviour has become broken, at least for me.

The following is an example file.

------start-------
#+BABEL: :session s1 :exports value latex :results raw

#+BEGIN_SRC R :results none :exports none 
1+2+3
#+END_SRC

src_R{1+1}
------end--------

Pressing C-c C-c with the cursor on the inline block produces the error:

  'R' is not recognized as an internal or external command,
  operable program or batch file.

This happens even if the session named s1 is already running. However, if I
first evaluate the BEGIN_SRC ... END_SRC block, using
org-babel-execute-src-block, and then reattempt to evaluate the inline block, it
will work. If I then press C-c C-c on the '#+BABEL:' line at the start of the
file, the inline block goes back to producing the error.

The second, and more aggravating, error is do with the consequences of
evaluating inline blocks. Formerly the result would be echoed in the 
minibuffer, and the document itself would not be altered. Now org has taken to 
inserting the result after the block, the same behaviour as a non-inline block. 
The header arguments used for this insertion seem to carry over either from the 
previous non-inline block, or possibly the global settings (BABEL: line). 

So for example, if I press C-c C-c on the src_R{1+1} above, I get:

-------
src_R{1+1} #+BEGIN_LaTeX
2#+END_LaTeX

-------

I want inline code blocks to replace themselves with their result when 
exporting the document to latex etc. I *never* want them to paste their results
into the document while editing - that is what non-inline blocks are for.

Is this change in behaviour intentional? If so is there a setting that will
revert to the old behaviour?

Paul

Re: org-babel: Bugs with inline src_* blocks

[Suvayu Ali]

Hi Paul,

On Tue, 15 Feb 2011 02:31:27 +0000 (UTC)
Paul Sexton <psexton@xnet.co.nz> wrote:

> Is this change in behaviour intentional? If so is there a setting
> that will revert to the old behaviour?
> 

I don't know the answer to your question however my question would be
how up to date is your org-mode install? I remember there was a bugfix
for a problem similar to this very recently (sometime in the last 2
weeks). Maybe all you are missing is a =git pull=? 

> Paul

GL
-- 
Suvayu

Open source is the future. It sets us free.

Re: org-babel: Bugs with inline src_* blocks

[Eric Schulte]

Paul Sexton <psexton@xnet.co.nz> writes:

> I am experiencing a couple of significant bugs with inline src blocks in 
> org-babel -- ie blocks of the form src_LANG{EXPRESSION}. I am using the
> development version of org, checked out a few days ago. 
>
> Pressing C-c C-c with the cursor on such a block is supposed to evaluate it and 
> echo the result to the minibuffer. However in recent versions of org (the last 
> 3 months or so) this behaviour has become broken, at least for me.
>
> The following is an example file.
>
> ------start-------
> #+BABEL: :session s1 :exports value latex :results raw
>
> #+BEGIN_SRC R :results none :exports none 
> 1+2+3
> #+END_SRC
>
>
> src_R{1+1}
> ------end--------
>
> Pressing C-c C-c with the cursor on the inline block produces the error:
>
>   'R' is not recognized as an internal or external command,
>   operable program or batch file.
>
> This happens even if the session named s1 is already running. However, if I
> first evaluate the BEGIN_SRC ... END_SRC block, using
> org-babel-execute-src-block, and then reattempt to evaluate the inline block, it
> will work. If I then press C-c C-c on the '#+BABEL:' line at the start of the
> file, the inline block goes back to producing the error.
>

I just tried your example from above, and I can not reproduce this
error.  Maybe as Suvayu suggested you should try grabbing the latest
version of Org-mode and see if that fixes this issue.

>
> The second, and more aggravating, error is do with the consequences of
> evaluating inline blocks. Formerly the result would be echoed in the 
> minibuffer, and the document itself would not be altered. Now org has taken to 
> inserting the result after the block, the same behaviour as a non-inline block. 
> The header arguments used for this insertion seem to carry over either from the 
> previous non-inline block, or possibly the global settings (BABEL: line). 
>
> So for example, if I press C-c C-c on the src_R{1+1} above, I get:
>
> -------
> src_R{1+1} #+BEGIN_LaTeX
> 2#+END_LaTeX
>
> -------
>
> I want inline code blocks to replace themselves with their result when 
> exporting the document to latex etc. I *never* want them to paste their results
> into the document while editing - that is what non-inline blocks are for.
>

Hi, I agree with your intuition here, but I changed the default inline
header argument so that others would be able to use inline code blocks
and have the results inserted.  To regain the behavior you describe,
simply adjust the value of `org-babel-default-inline-header-args' as
follows in your .emacs file.

#+begin_src emacs-lisp
  (setf org-babel-default-inline-header-args
        '((:session . "none")
          (:results . "silent")
          (:exports . "results")))
#+end_src

Best -- Eric

>
> Is this change in behaviour intentional? If so is there a setting that will
> revert to the old behaviour?
>
> Paul
>
>
>
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Please use `Reply All' to send replies to the list.
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode

Re: org-babel: Bugs with inline src_* blocks

[Paul Sexton]

Eric Schulte <schulte.eric <at> gmail.com> writes:
> Hi, I agree with your intuition here, but I changed the default inline
> header argument so that others would be able to use inline code blocks
> and have the results inserted.  To regain the behavior you describe,
> simply adjust the value of `org-babel-default-inline-header-args' as
> follows in your .emacs file.
> 
> #+begin_src emacs-lisp
>   (setf org-babel-default-inline-header-args
>         '((:session . "none")
>           (:results . "silent")
>           (:exports . "results")))
> #+end_src

Thanks Eric. The 'default' behaviour I experienced where inline blocks
were inheriting behaviour from elsewhere seems odd however. Is it intended
to work this way?

Paul

Re: org-babel: Bugs with inline src_* blocks

[Dan Davison]

Paul Sexton <psexton@xnet.co.nz> writes:

> Eric Schulte <schulte.eric <at> gmail.com> writes:
>> Hi, I agree with your intuition here, but I changed the default inline
>> header argument so that others would be able to use inline code blocks
>> and have the results inserted.  To regain the behavior you describe,
>> simply adjust the value of `org-babel-default-inline-header-args' as
>> follows in your .emacs file.
>> 
>> #+begin_src emacs-lisp
>>   (setf org-babel-default-inline-header-args
>>         '((:session . "none")
>>           (:results . "silent")
>>           (:exports . "results")))
>> #+end_src
>
> Thanks Eric. The 'default' behaviour I experienced where inline blocks
> were inheriting behaviour from elsewhere seems odd however.

Hi Paul, Eric,

Paul -- I do agree with you. I've had to manually delete unwanted inline
src output a few times in the last few days.

I think we did have the behaviour we want before -- replacement during
export but no modification of the buffer in interactive use. Could we
make the default behave so that it does "replace" during export and
"silent" during normal interactive evaluation? (Or some similar change
to ob-exp.el?)

Dan


> Is it intended
> to work this way?
>
> Paul
>
>
>
> _______________________________________________
> Emacs-orgmode mailing list
> Please use `Reply All' to send replies to the list.
> Emacs-orgmode@gnu.org
> http://lists.gnu.org/mailman/listinfo/emacs-orgmode

Re: Re: org-babel: Bugs with inline src_* blocks

[Eric Schulte]

Dan Davison <dandavison7@gmail.com> writes:

> Paul Sexton <psexton@xnet.co.nz> writes:
>
>> Eric Schulte <schulte.eric <at> gmail.com> writes:
>>> Hi, I agree with your intuition here, but I changed the default inline
>>> header argument so that others would be able to use inline code blocks
>>> and have the results inserted.  To regain the behavior you describe,
>>> simply adjust the value of `org-babel-default-inline-header-args' as
>>> follows in your .emacs file.
>>> 
>>> #+begin_src emacs-lisp
>>>   (setf org-babel-default-inline-header-args
>>>         '((:session . "none")
>>>           (:results . "silent")
>>>           (:exports . "results")))
>>> #+end_src
>>
>> Thanks Eric. The 'default' behaviour I experienced where inline blocks
>> were inheriting behaviour from elsewhere seems odd however.
>
> Hi Paul, Eric,
>
> Paul -- I do agree with you. I've had to manually delete unwanted inline
> src output a few times in the last few days.
>
> I think we did have the behaviour we want before -- replacement during
> export but no modification of the buffer in interactive use. Could we
> make the default behave so that it does "replace" during export and
> "silent" during normal interactive evaluation? (Or some similar change
> to ob-exp.el?)
>

The easiest way to have header argument values vary between interactive
and export evaluation is to place raw elisp in the header argument
value.  The following (after pulling the latest version of Org-mode)
results in the behavior you describe.

#+begin_src emacs-lisp :results silent
  (setf org-babel-default-inline-header-args
        '((:session . "none")
          (:results . (if (boundp 'org-current-export-file) "replace" "silent"))
          (:exports . "results")))
#+end_src

I think this is generally the most desirable behavior, and I'd vote that
this become the default header argument value for inline code blocks.

Best -- Eric

>
> Dan
>
>
>> Is it intended
>> to work this way?
>>
>> Paul
>>