On 04/04/14 00:48, Ralf Hemmecke wrote:
That still has some parts that need "pretend". Suggestions are welcome
how to remove them.
In fact, I would be happy to separate the "algebraic" transformation
from the one that is connected to type and language. I somehow believe
that the language stuff doesn't need translation boot->spad. Am I wrong?
Ralf
I think it would be really good if the whole interpreter were converted
to SPAD. The reason that I say that is that I would like changes to it
such as the ability to specify operator precedence per domain. The lack
of this is a serious problem which makes Clifford algebra unusable for
my purposes. Also both compiler and interpreter need improvements to
type system.
I get the impression that the interpreter, as it stands now, is hard to
change because small changes are likely to break things in unexpected
ways. I'm not just talking about the unpredictable nature of the Pratt
compiler but more about the loosely typed and poorly documented code.
I get the impression this might be improved if the interpreter used
higher level, better typed structures.
I don't think a line-by-line translation from boot to spad would be good
enough, on its own, because the loose typing still remains, its just
hidden in SExpression. I think these SExpressions need to be replaced by
higher level structures which are properly typed.
For example, in something relatively simple like this:
(1) -> parseFromString("2*3.4")$Lisp
(1) (* 2 (($elt (Float) float) 34 - 1 10))
Type: SExpression
we might just get away without high level structures, but when we get to
something more complicated like the following, then high level
structures would provide proper compile level type checking and avoid a
lot of runtime errors. It would also make the whole thing easier to
understand.
(2) -> a := mkAtreeWithSrcPos(parseFromString("2*3.4")$Lisp,nil)$Lisp
(2)
([*,NIL,NIL,NIL,NIL]
[--immediateData--,NIL,PositiveInteger()(),NIL,NIL]
([Dollar,NIL,NIL,NIL,NIL] (Float)
([float,NIL,NIL,NIL,NIL]
[--immediateData--,NIL,PositiveInteger()(),NIL,NIL]
[--immediateData--,NIL,Integer()(),NIL,NIL]
[--immediateData--,NIL,PositiveInteger()(),NIL,NIL])
)
)
Type: SExpression
Martin
--
You received this message because you are subscribed to the Google Groups "FriCAS -
computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/fricas-devel.
For more options, visit https://groups.google.com/d/optout.
