« Return to Thread: scalax.reflect?
Re: scalax.reflect?
by Andrew.Foggin
::
Rate this Message:
I've been working (very part time) on a ScalaSig parser using the scalax
rule combinator as I proposed on the list a few weeks ago. I have quite
a bit of ScalaSig figured out now and it's producing simple Mirror objects.
I'm happy to share it. It's not in any state suitable for addition to
scalax yet but I could throw it up somewhere if you're interested.
Regards,
Andrew
David MacIver wrote:
> I've been tinkering with the bare bones of a little reflection library
> for Scala in my spare time and it occurred to me to wonder if there
> was any interest in merging something like that into scalax and
> turning it into more of a community effort. I'm sure other people have
> done some stuff towards this as well (I know Aaron Harnly has been
> doing some interesting things with reflection. That's part of what
> prompted me to look into this), so pooling efforts and getting
> additional use cases for it seemed like a good idea.
>
> The basic goal at the moment is to be able to:
>
> - Dynamically invoke methods on Scala objects as something like
> call("foo", args). This should be aware of Scala specific features
> like symbolic names, call by name arguments, varargs, etc as well as
> resolve overloads at runtime.
> - Create new instances of Scala types in a similar way.
> - Create proxies in a way that understand the traits implementation,
> so you only need to proxy abstract methods in a trait and you'll get
> the rest for free.
>
> I've got the basics of the first two working (symbolic names work,
> although not for converting to primitive equivalents yet, varargs and
> call by name arguments don't except in the sense that if you pass
> instances of the wrapper type it will work, overloading is currently
> only kinda sorta resolved but I know how to fix it) and not done much
> on the third.
>
> Also I'm trying to design the library in a way that doesn't require
> explicit mention of anything from java.lang.reflect. I've no interest
> in working on a .NET backend for it, but it should probably be
> designed in such a way as to not preclude the existence of one.
>
>
>
rule combinator as I proposed on the list a few weeks ago. I have quite
a bit of ScalaSig figured out now and it's producing simple Mirror objects.
I'm happy to share it. It's not in any state suitable for addition to
scalax yet but I could throw it up somewhere if you're interested.
Regards,
Andrew
David MacIver wrote:
> I've been tinkering with the bare bones of a little reflection library
> for Scala in my spare time and it occurred to me to wonder if there
> was any interest in merging something like that into scalax and
> turning it into more of a community effort. I'm sure other people have
> done some stuff towards this as well (I know Aaron Harnly has been
> doing some interesting things with reflection. That's part of what
> prompted me to look into this), so pooling efforts and getting
> additional use cases for it seemed like a good idea.
>
> The basic goal at the moment is to be able to:
>
> - Dynamically invoke methods on Scala objects as something like
> call("foo", args). This should be aware of Scala specific features
> like symbolic names, call by name arguments, varargs, etc as well as
> resolve overloads at runtime.
> - Create new instances of Scala types in a similar way.
> - Create proxies in a way that understand the traits implementation,
> so you only need to proxy abstract methods in a trait and you'll get
> the rest for free.
>
> I've got the basics of the first two working (symbolic names work,
> although not for converting to primitive equivalents yet, varargs and
> call by name arguments don't except in the sense that if you pass
> instances of the wrapper type it will work, overloading is currently
> only kinda sorta resolved but I know how to fix it) and not done much
> on the third.
>
> Also I'm trying to design the library in a way that doesn't require
> explicit mention of anything from java.lang.reflect. I've no interest
> in working on a .NET backend for it, but it should probably be
> designed in such a way as to not preclude the existence of one.
>
>
>
« Return to Thread: scalax.reflect?
| Free embeddable forum powered by Nabble | Forum Help |
