[Maxima-discuss] Displaying an "a\nbcd\nef" using very low-level tools

[Eduardo O. <edu...@gm...>]

Hi list!

First, some context. In this message from a few days ago

  https://sourceforge.net/p/maxima/mailman/message/59245670/

Robert Dodier showed a way to draw matrixes without border characters.
I have lots of objects - like trees and certain diagrams for teaching
set comprehensions - that can be represented nicely in ascii art, and
using RD's trick I can almost do what I want/need.

The most obvious way of handling that problem is saying that what I
want is wrong, but let me follow another path, and start by a minimal
example. Here is what I can do with RD's trick. If we run this in a
REPL in a recent Maxima,

  to_lisp();
    (setf (get '$barematrix 'dimension) 'dim-$barematrix)
    (defun dim-$barematrix (form result)
      (let (($display_matrix_brackets nil))
        (dim-$matrix form result)))
    (to-maxima)
  linel : 60;
  display2d_unicode : false;
  BM : barematrix(["a  "], ["bcd"], ["ef "]);
  matrix([BM, 42]);

We get this:

  (%i4) BM : barematrix(["a  "], ["bcd"], ["ef "]);
                              a

  (%o4)                       bcd

                              ef
  (%i5) matrix([BM, 42]);
                         [  a        ]
                         [           ]
  (%o5)                  [  bcd   42 ]
                         [           ]
                         [  ef       ]
  (%i6)

BM looks like "a\n\nbcd\n\nef", but I would like to get something that
looks like "a\nbcd\nef", without the interline spacings.

Here is an attempt to get a "a\nbcd\nef" using very low-level tools.
If we run this in a Maxima REPL,

  to_lisp();
    (setf (get '$abcdef 'dimension) 'dim-abcdef)
    (defun dim-abcdef (form result)
      (declare (ignorable form result))
      (setq depth 2)
      (setq maxdp 2)
      (setq width 3)
      '((-3 -2 #\f #\e) (-1 -1 #\d #\c #\b) #\a))
    (to-maxima)
  linel : 40;
  display2d_unicode : false;
  matrix([abcdef(), 42]);
  42 + abcdef();
  abcdef();

it works, or sort of, in some contexts, but it is very buggy. Look:

  (%i4) matrix([abcdef(), 42]);
                [ a   42 ]
  (%o4)         [ bcd    ]
                [ ef     ]

  (%i5) 42 + abcdef();
  a  + 42
  bcd
  ef
  (%i6) abcdef();
  24 +

    ef

I'm looking for suggestions on what to fix in my dim-abcdef. If anyone
wants to take a look at the tools that I am using to explore
displa.lisp and displm.lisp, they are here -
<http://anggtwu.net/MAXIMA/2025-testdispla.lisp.html> - but they are
not worth looking at now, and they are only "documented" using test
blocks: <http://anggtwu.net/eepitch.html#test-blocks>.

Suggestions like "trace the variables such and such" would be great,
and suggestions like "use the functions blah and blah instead of your
setqs", or "make your function return bleh instead of '((-3 -2 #\f
#\e) (-1 -1 #\d #\c #\b) #\a)" would be even better...

  Thanks in advance!
    Eduardo Ochs =)
    http://anggtwu.net/eev-maxima.html