Re: Generate new babel code blocks and/or initialized code/data?

["Thomas S. Dye" <tsd@tsdye.com>]

Aloha Lawrence,

Here

#+begin_src clojure :var i=(myfun1)
(inc i)
#+end_src

you are asking emacs-lisp to evaluate myfun1, which is not an emacs-lisp
function.

See http://orgmode.org/manual/var.html#var, "Emacs Lisp evaluation of
variables"

I'm not sure why you are able to bypass the :var functionality in
clojure.  What you are seeing is the behavior I expect in a :session.
I'm not a programmer--you'll need some input from a programmer on the
list to sort this one out.

hth,
Tom

Lawrence Bottorff writes:

> If I evaluate this:
>
> #+name: myfun1
> #+begin_src clojure
> (defn myfun1
>   [ ]
>   9)
> #+end_src
>
> #+RESULTS: myfun1
> : #'clojure-noob.core/myfun1
>
> then this
>
> #+begin_src clojure
> (inc (myfun1))
> #+end_src
>
> #+RESULTS:
> : 10
>
> I've got the right answer, but I've totally bypassed the :var
> functionality. In my elisp example
>
> #+begin_src emacs-lisp :var myx=(mylist1)
> (mapcar '1+ myx)
> #+end_src
>
> #+RESULTS:
> | 2 | 3 | 4 | 5 |
>
> it seems to be as you say, i.e., the function needs to be evaluated, and,
> yes, the mylist1 function in parens does the trick. But this
>
> #+begin_src clojure :var i=(myfun1)
> (inc i)
> #+end_src
>
> produces this in *Messages*:
>
> eval: Symbol’s function definition is void: myfun1
>
> However, this
>
> #+name: myfun1-test2
> #+begin_src clojure
> (myfun1)
> #+end_src
>
> #+RESULTS: myfun1-test2
> : 9
>
> does finally get seen and evaluated:
>
> #+begin_src clojure :var i=myfun1-test2
> (inc i)
> #+end_src
>
> #+RESULTS:
> : 10
>
> But this adds an extra step just to use :var. I'm guessing regular Lisp and
> maybe Scheme (geiser) know about var i=(myfun1), . Will test, but I've got
> to swap out my .emacs.d first. As far as noweb is concerned, doing
> <<myfun1>> doesn't win me anything, does it? It's the same as (myfun1)
> above.
>
>
>
>
>
>
>
>
>
>
> On Sat, Sep 24, 2016 at 8:03 PM, Thomas S. Dye <tsd@tsdye.com> wrote:
>
>> Aloha Lawrence,
>>
>> I don't know the Clojure dialect, but I think the problem is that the
>> myfun1 source code block returns a function.  It doesn't evaluate
>> the function and return a result, which is I think what you are
>> expecting.
>>
>> You can use noweb expansion of myfun1 to define the function inside
>> another source code block, then use the function in the normal way in
>> the source code block (or in the session, if you have that set).
>>
>> Or, you might change the myfun1 source code block to return a list,
>> rather than a function (if I'm reading Clojure correctly):
>>
>> #+name: myfun1
>> #+begin_source clojure
>> [8 9]
>> #+end_source
>>
>> hth,
>> Tom
>>
>> Lawrence Bottorff writes:
>>
>> > Not sure if you know Clojure, but here's what I've been toying with:
>> >
>> > #+name: my-test
>> > #+begin_src clojure :var i=[1 2]
>> > (map inc i)
>> > #+end_src
>> >
>> > #+RESULTS: my-test
>> > | 2 | 3 |
>> >
>> > looks good, but then
>> >
>> > #+name: myfun1
>> > #+begin_src clojure
>> > (defn myfun1
>> >   [ ]
>> >   [8 9])
>> > #+end_src
>> >
>> > #+begin_src clojure :var i=myfunc1
>> > (map inc i)
>> > #+end_src
>> >
>> > doesn't do anything, i.e., it doesn't process the myfunc1 and provide the
>> > vector [8 9]
>> >
>> > This elisp code works, though:
>> >
>> > #+name: mylist1
>> > #+begin_src emacs-lisp
>> > (defun mylist1 ()
>> >   (list 1 2 3 4))
>> > #+end_src
>> >
>> > then
>> >
>> > #+begin_src emacs-lisp :var myx=(mylist1)
>> > (mapcar '1+ myx)
>> > #+end_src
>> >
>> > #+RESULTS:
>> > | 2 | 3 | 4 | 5 |
>> >
>> > Note how I put mylist1 in parens. Without produced odd output
>> >
>> > #+RESULTS:
>> > | 110 | 122 | 109 | 106 | 116 | 117 | 50 |
>> >
>> > . . . which is literally taking the ascii letters of the word "mylist1"
>> and
>> > incrementing them. (Too much fun. . . ). What might be wrong with my
>> > Clojure attempt? I've tried (myfun1), myfun1, and myfun1() gives an
>> error.
>> >
>> >
>> >
>> > On Tue, Sep 20, 2016 at 3:33 PM, Thomas S. Dye <tsd@tsdye.com> wrote:
>> >
>> >> Aloha Lawrence,
>> >>
>> >> Lawrence Bottorff writes:
>> >>
>> >> > So I can run code for a REPL-type language like Clojure in a babel
>> code
>> >> > block and get "results," e.g., a Clojure code block takes in a vector
>> of
>> >> > mappings and produces new "results":
>> >> >
>> >> > #+RESULTS[abc5c51bb569a82c19c4eea1c385c74e839922c7]:
>> >> > symmetrize-body-parts-test
>> >> > | :name | head            | :size |  3 |
>> >> > | :name | left-eye        | :size |  1 |
>> >> > | :name | right-eye       | :size |  1 |
>> >> > | :name | left-ear        | :size |  1 |
>> >> > . . .
>> >> >
>> >> > but could I generate results that aren't just static output listed
>> after
>> >> a
>> >> > #+RESULTS tag, rather, embedded in a newly created babel code block?
>> I'd
>> >> > like such output "initialized" as far as the running REPL is concerned
>> >> too.
>> >> > Is it possible to generate new code/data that is immediately known to
>> the
>> >> > REPL session? Any examples don't have to be Clojure.
>> >>
>> >> You can use the :session header argument which will give you access to
>> >> any variables created during the session:
>> >>
>> >> http://orgmode.org/worg/org-contrib/babel/languages/ob-
>> >> doc-clojure.html#orgheadline13
>> >>
>> >> You can pass the function results to a variable argument, which makes
>> >> possible chaining (see http://www.jstatsoft.org/v46/i03):
>> >>
>> >> #+header: :var x=myfunc(2)
>> >>
>> >> You can also embed and call a function in a source code block using
>> noweb
>> >> syntax:
>> >>
>> >> http://orgmode.org/worg/org-contrib/babel/intro.html#
>> literate-programming
>> >>
>> >> hth,
>> >> Tom
>> >>
>> >> --
>> >> Thomas S. Dye
>> >> http://www.tsdye.com
>> >>
>>
>>
>> --
>> Thomas S. Dye
>> http://www.tsdye.com
>>


-- 
Thomas S. Dye
http://www.tsdye.com