Re: org-sbe: error when passing strings as parameters to/from Python blocks

[Daniel Herzig <daniel.herzig@outlook.at>]

Karl Voit <devnull@Karl-Voit.at> writes:

> Hi!
Hi!
>
> I want to test/use Python with org-sbe:
>
> #+NAME: classificationfm
> #+BEGIN_SRC python :exports none :var prob :var impact
>
> result = ""
> if prob == 'high' and impact == 'high':
>     return 'A'
> if prob == 'low' and impact == 'high':
>     return 'B'
> if prob == 'high' and impact == 'low':
>     return 'C'
> if prob == 'low' and impact == 'low':
>     return 'D'
> return 'undefined'
> #+END_SRC
>
After some trying I found that the variables as set in the source-code
header need standard values set:

#+NAME: classificationfm
#+BEGIN_SRC python :var prob="high" :var impact="high"
  if prob == "high" and impact == "high":
      return "A"
  if prob == "low" and impact == "high":
      return "B"
  if prob == "high" and impact == "low":
      return "C"
  if prob == "low" and impact == "low":
      return "D"
#+END_SRC

If I don't set them I get exactly the same errors as you. Like this I
get the following:

| prob | impact | class |
|------+--------+-------|
| high | high   | A     |
| low  | high   | B     |
| high | low    | C     |
| low  | low    | D     |
#+TBLFM: @2$3..@>$3='(org-sbe classificationfm (prob $$1) (impact $$2))

Evaluation is being asked for each line then.


> (Yes, I should move to elif and the Python code might be coded with
> less characters in general: this should not be the point here ;-) )
>
> | prob | impact | class  |
> |------+--------+--------|
> | high | high   | #ERROR |
> | low  | high   | #ERROR |
> | high | low    | #ERROR |
> | low  | low    | #ERROR |
>
> #+TBLFM: @2$3..@>$3='(org-sbe "classificationfm" (prob $1) (impact $2))
>
> I'd expect to get the values of the "class" table column: A, B, C, D
> instead of #ERROR.
>
>
> Reading the manual, I found:
> | NOTE: By default, string variable names are interpreted as
> | references to source-code blocks, to force interpretation of a
> | cell’s value as a string, prefix the identifier a "$" (e.g.,"$$2"
> | instead of "$2" or "$@2$2" instead of "@2$2").
>
> Therefore, I tried the following lines ...
>
> #+TBLFM: @2$3..@>$3='(org-sbe "classificationfm" (prob "$1") (impact "$2"))
> #+TBLFM: @2$3..@>$3='(org-sbe "classificationfm" (prob $$1) (impact $$2)) 
>
> #+TBLFM: @2$3..@>$3='(org-sbe "classificationfm" (prob "$$1") (impact "$$2")) 
> ... all with same error in the result.
>
>
> My setup: Org mode version 9.1.6 on GNU Emacs 26.0.90 and GNU Emacs 25.1.1
>
>
> I started a reddit thread[1]. In this thread, somebody was posting
> this table showing the error:
>
> #+NAME: myfunc
> #+BEGIN_SRC python :var n="1"
>
> return "ok"
> #+END_SRC
>
> | 10.3 |  9 | -18 | a      | 14 | 11 |  1 |
> |   ok | ok |  ok | #ERROR | ok | ok | ok |
>
> #+TBLFM: @2='(org-sbe myfunc (n @1))
>
>
> However, using the comment from the manual about strings, I
> prepended the reference with "$" ...
>
> | 10.3 |  9 | -18 | a  | 14 | 11 |  1 |
> |   ok | ok |  ok | ok | ok | ok | ok |
>
> #+TBLFM: @2='(org-sbe mytestfunc (n $@1))
>
> ... which now looks OK ;-)
>
>
> So, back to the initial situation: what is my error or do we have a
> bug in Org?
>
>
> [1] https://www.reddit.com/r/orgmode/comments/b0ll1v/embedding_python_code_in_table_formula/


I hope this helps, I'm on orgmode 8.2.1 here.

Cheers,
Daniel