Thanks Raul,
First, for clarifying my understanding of the way that J uses UTF-8 as its
representation of the 'literal' type and second, for the clear explanation of
the challenges that unicode presents.
I don't actually think that the two examples that you gave create 'literal'
results, at least as defined by datatype (defined in system/main/stdlib.ijs).
datatype
3 : 0
n=. 1 2 4 8 16 32 64 128 1024 2048 4096 8192 16384 32768 65536 131072
t=. '/boolean/literal/integer/floating/complex/boxed/extended/rational'
t=. t,'/sparse boolean/sparse literal/sparse integer/sparse floating'
t=. t,'/sparse complex/sparse boxed/symbol/unicode'
(n i. 3!:0 y) pick <;._1 t
)
datatype 'a',(u:600+i.5),u:30000+/i.5
unicode
datatype ": 'a',(u:600+i.5),u:30000+/i.5
literal
There also seems to be a 'hole' in the type representation for 2^256 and 2^512
which may be available for type use, but that is probably just me
misunderstanding the scope of how the language uses these numbers to represent
types. My guess is that your reference to available type numbers refers to a
different context from the one presented in datatype.
The most important thing to me is the question that you posed at the end. What
do you want the user to understand?
My goal would be that by looking at the representation of a literal array, the
user would have the option to use spatial cues (width reflecting UTF-8) to
understand the underlying encoding of the UTF-8 represented characters. This
would also make the width of the boxed nouns consistent, as the current ragged
boxed edge is a result of the difference between the actual character width and
number of integers required to encode those characters in UTF-8. This would not
work for a purely textual interface because it would overload our
interpretation of character spacing, but with the right background containers
it may allow a user to intuitively understand what you already know about the
variability of UTF-8 encoding.
Thanks again for taking the time to contribute to my education. It really is
appreciated. :-)
Cheers, bob
> On Jul 5, 2016, at 9:17 AM, Raul Miller <rauldmil...@gmail.com> wrote:
>
> Which brings me back to questions like "What do you want the user to
> understand?"
>
> Thanks,
>
> --
> Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
