FP6: Types
Autrijus Tang
FP6. I intend to hold off subtyping until the simple types can be
checked and matched against smoothly. The types are currently:
VBool Boolean
VInt Integer
VNum Number
VStr String
VRef Reference
VPair Pair
VList Lazily evaluated list
VRoutine Nameable executable objects
VBlock Unnameable executable objects
Implementation types are:
IScalar Perl scalar
IArray Perl array
IHash Perl hash
This list differs from S06v6's "Types" section, in that S06 did
not list which types are I-Types (implementation types) and which are
V-Types (value types). Also "Pair" is not listed among standard types.
Hence, the list above reflects my current intuition, based on
past experience with Perl5's tie() system. If my intuition are way off
base, please do let me know. :-)
Thanks,
/Autrijus/
Autrijus Tang
> Uh, as this is an ask-for-clarification for a Synopsis, I think p6l
> is the correct forum. Of course, as I'm new to this list, I may be
> totally wrong, in which case please tell me so. :-)
To be more precise, according to S06:
"A value type is given as a prefix or with the 'returns' or 'of' keywords."
"The implementation type is given as a trait for the variable via 'is'."
It seems clear to me that the distinction is visible in a language level;
as I'm implementing the parser, this is crucial information to me, and
I suspect perl6-internals won't be of much help here...
Another interpretation is that there is only one default I-Type, namely Scalar,
and Array and Hash are merely V-Types types. In that case, I'd like to
inquire whether Lists and Arrays are always intercastable.
Thanks,
/Autrijus/
(PS: Mark, your server is blocking my email server at 168.95.4.166,
so I'm replying your mail on-list. I'm sorry for the spamming. :-/)
Mark J. Reed
On 2005-02-03 at 11:13:30, Autrijus Tang wrote:
> Today I've started implementing typing relations for value types in
Shouldn't this have gone to perl6-internals, not perl6-language?
Autrijus Tang
Uh, as this is an ask-for-clarification for a Synopsis, I think p6l
is the correct forum. Of course, as I'm new to this list, I may be
totally wrong, in which case please tell me so. :-)
Thanks,
/Autrijus/
Autrijus Tang
> perl6-c...@perl.org seems a good forum to me for your endeavor
> because you write a compiler for a subset of Perl6.
Okay. If that is the desired forum, tomorrow I'll switch my questions
to p6c and unsubscribe from p6l. Sorry for bothering the people
involved. :-)
> You are speaking of types, do you intend to explore type inference?
Yes, eventually.
> The problem is more complex than with statically typed language, but
> somewhat simplified by the choice of a purely functional subset of Perl6.
Which is why I started with such a subset...
Thanks,
/Autrijus/
Stéphane Payrard
>
perl6-c...@perl.org seems a good forum to me for your endeavor
because you write a compiler for a subset of Perl6.
You are speaking of types, do you intend to explore type inference?
The problem is more complex than with statically typed language, but
somewhat simplified by the choice of a purely functional subset of Perl6.
--
stef
Matt Fowles
Actually, I think that p6l is the correct place for this discussion.
My logic is that you are asking about specific facets of the language,
not helping the perl 6 compiler or parrot.
Matt
--
"Computer Science is merely the post-Turing Decline of Formal Systems Theory."
-???
Juerd
> perl6-c...@perl.org seems a good forum to me for your endeavor
> because you write a compiler for a subset of Perl6.
What someone *does* is irrelevant. The topic of the mail message
matters, and if that is the language, this is the right list.
If all the lists are too strictly separated, we end up with two very
different Perl 6s: one that exists only in design and one, made by the
internals people, with awful hacks because the design didn't work that
well in practice.
Juerd
--
http://convolution.nl/maak_juerd_blij.html
http://convolution.nl/make_juerd_happy.html
http://convolution.nl/gajigu_juerd_n.html