Re: Re: Re: Re: bootstrap methods in the constant pool

On Tue, Jul 13, 2010 at 9:55 AM, Rémi Forax <[email protected]> wrote:

>>  Perhaps we could split it into two visit calls...
> yes, but in that case we loose backward compatibility.

  Backward compatibility to what?

>> or introduce a micro language, e.g. in desc field.
> you mean in owner field, invokedynamic has no owner.

  Either one.

>>  BTW, what would be the meaning of bootstrap method pointing to a method
>> handle of get/putField or get/putStatic type?
> This is not allowed.

  The data structures allow that.

>>  An alternative way to represent bootstrap method in the bytecode could be
>> to introduce a new attribute at the class, method and code level. This would
>> be also closer to semantic John is suggesting. Though that way you couldn't
>> enforce constrains on a pretense of the bootstrap methods, but that can be
>> done in the verifier.
>
> This is not the way chosen by the JSR 292 expert group,
> and creating of fake attribute will trouble ASM users.

  Is that a final decision? Or there is still change the expert group
could reconsider?

> In my opinion, the only valid path is to introduce a new method
> visitInvokeDynamicMethodInsn(int refType, String boostrapOwner, String
> bootstrapName, String bootstrapDesc, String invokedynamicName, String
> invokeDynamicDesc)

  All bytecode structures are attributes. I wonder if we can expose
just attributes and allow creating wrapper visitors that would
translate a high-level calls like you suggested into an attribute
construction.

  regards,
  Eugene

-- 
You receive this message as a subscriber of the [email protected] mailing list.
To unsubscribe: mailto:[email protected]
For general help: mailto:[email protected]?subject=help
OW2 mailing lists service home page: http://www.ow2.org/wws