Re: [O] [Babel] Lisp error: (wrong-type-argument listp hline)

[Sebastien Vauban]

Hi Eric,

Eric Schulte wrote:
> I think you're confused by headers which are re-added by the colnames
> machinery.

Blush!  I mixed the two in my head, yes.

> See the following which returns scalar output avoiding any colnames
> post-processing.

Why are you talking of *post* processing machinery for colnames?

The documentation[1] says that ":colnames no" means "no column name *pre*
processing takes place."

I may miss something important here.

> #+Property: results scalar
>
> #+name: table
> | Key | Value |
> |-----+-------|
> | ABJ |     1 |
> | DEK |     2 |
>
> #+begin_src emacs-lisp :var data=table :hlines no
>   data
> #+end_src
>
> #+RESULTS:
> : (("Key" "Value") ("ABJ" 1) ("DEK" 2))
>
> #+begin_src emacs-lisp :var data=table :hlines yes
>   data
> #+end_src
>
> #+RESULTS:
> : (("Key" "Value") hline ("ABJ" 1) ("DEK" 2))
>
>> * Set :hlines to "yes"
>>
>> `:hlines yes' should leave the horizontal line, but generates an error.
>>
>>   #+begin_src emacs-lisp :var data=table :results output :hlines yes
>>   (mapc (lambda (item) (princ (format "UPDATE dim SET val=%s WHERE 
>> code='%s'\n"
>>                                       (nth 1 item) (nth 0 item))))
>>    data)
>>   #+end_src
>
> This is not a babel error, this is an error in your code block body,
> which assumes that every element of `data' will be a list.  The symbol
> `hline' is not a list.

OK, that now becomes clear to me!

>> Finally, how am I then supposed to *ignore the header line*?  By adding
>>
>>   #+begin_src emacs-lisp
>>   (setq table (cdr table))
>>   #+end_src
>>
>> ?
>>
>> Do I have to do that explicitly in my code block, or is there an option for
>> telling that to Babel?
>
> Use the :colnames processing to strip the headings.
>
>     #+begin_src emacs-lisp :var data=table :colnames yes
>       data
>     #+end_src
>
>     #+RESULTS:
>     : (("ABJ" 1) ("DEK" 2))

I always wondered why emacs-lisp is the _only_ language with ":colnames no" as
its default. Is there a reason therefore?  If no really good reason, could we
suppress that?

Best regards,
  Seb

--
Sebastien Vauban

[1] http://orgmode.org/org.html#colnames