Re: [Axiom-developer] Unions in Spad

[Bill Page]

On 7/9/07, Ralf Hemmecke <address@hidden> wrote:
> ...
> Anyway, I would not say that Aldor does not allow
> "Union(Integer,String)". No hint in the AUG, but look at the error
> message given in
> http://lists.nongnu.org/archive/html/axiom-developer/2007-07/msg00181.html
>
> If you rewrite this program to
>
>      U == Union(machInt: MachineInteger, sStr: String);
>
> (just change this line) and then look at the error message.
>
> woodpecker:~/scratch>aldor -laldor -fx aaa.as
> "aaa.as", line 13:         if u case MachineInteger then {
>                     ..................^
> [L13 C19] #1 (Error) Argument 2 of `case' did not match any possible
> parameter type.
>      The rejected type is Join(
>                  PrimitiveType with
>                      ....
>      Expected one of:
>        -- Enumeration(machInt: Type)
>        -- Enumeration(sStr: Type)
>
> It looks as if one also could do
>
>    if u case MachineInteger then ...
>
> with the current Aldor compiler, I just don't know how to type
> "MachineInteger" since the compiler immediately takes the *type*
> MachineInteger and not the *tag* MachineInteger.

Use quotes like this?

 'MachineInteger'

That makes in a member of an Enumeration. Maybe that is what Aldor expects?

> Anyway, that usage of MachineInteger is not specified in the AUG so
> maybe one better does not use it that way.

I have become quite confident that Aldor takes the most general road
whenever possible inspite of what might or might not be written in the
user's guide, so I am inclined to think that you are right.
'Union(Integer,String)' might very well be acceptible to the Aldor
compiler, we just need to intuit the proper usage.

Regards,
Bill Page.