On 8/4/08, Dieter Kaiser <drdieterkaiser at web.de> wrote:
> I had a look in the documentation. We find:
>
> The single quote does not prevent simplification.
> So the simplification of '(sin(1.0)) or of the Gamma function
> '(gamm(0.5)) is expected.
Yes.
> I think this seems to be only true for functions which have an alias and a
> reversealias on the property list. These functions are allready translated by
> the parser to a noun form.
I'm pretty sure any operator which has a simplification function
(as indicated by the OPERATORS property) is simplified (be it in
an evaluating or non-evaluating context).
There is probably an interaction with ALIAS or REVERSEALIAS.
If the simplification function is a property of the noun then it would
be called only after an expression is nounified.
> Functions which don't have an alias and reversealias are not automatically
> transformed to a noun form by the parser.
Most or all of the existing math functions are automatically
nounified by the parser. For the sake of consistency, I think we
should make new math functions work the same way.
Take a look at the stuff for math functions in src/mlisp.lisp and
src/suprv1.lisp --- maybe new math functions should be added
to the lists of functions there.
> That's all a bit confusing.
I think a rationale could be invented. It certainly isn't obvious;
sorry about that.
best
Robert Dodier