[armedbear-devel] [PATCH] jcall(there are multiple matchingmethods)

[Erik Huelsmann]

On Tue, Dec 22, 2009 at 9:54 PM, Alan Ruttenberg
<alanruttenberg at gmail.com> wrote:
> This would be incompatible with other common lisps, and make porting
> to other java enabled lisps in the future a pain. My gut says don't
> mess with the define reading behavior.

We could look at the way CCL (Clozure) integrates with Cocoa: it uses
a special syntax to identify foreign calls too. That would be a nice
way of integrating; besides: that kind of macro character would
defacto be unuseable anyway: CCL has a defined purpose for it already.

Regards,

Erik

>
> -Alan
>
> On Tuesday, December 22, 2009, Tobias C. Rittweiler <tcr at freebits.de> wrote:
>> Alessio Stalla writes:
>>
>>> On Mon, Nov 23, 2009 at 2:47 PM, Alessio Stalla wrote:
>>>
>>> Now jcall can be used with less verbose syntax, for example:
>>> (jcall "toString" 4) ==> "4"
>>> (jcall "compareTo" 4 5) ==> -1
>>>
>>> it would be nice to add (#"methodName" args) a la JSS for even shorter syntax.
>>
>> FWIW, I do not like the special read syntax. I think, the real solution
>> that ABCL should strive for is
>>
>>   - to implement per-package symbol case sensitivity a la Clisp
>>
>>     See  http://clisp.cons.org/impnotes/package-case.html
>>
>>   - introduce a package JAVA (with nickname J) where symbols are
>>     interned in while preserving case.
>>
>> Ultimatively, ABCL should strive for making
>>
>>   (funcall j:compareTo 4 5)  [read in as (FUNCALL J:|compareTo| 4 5)]
>>
>> work.
>>
>> Sounds like a nice project for the free time between the years.
>>
>> Any takers?
>>
>>   -T.
>>
>>
>> _______________________________________________
>> armedbear-devel mailing list
>> armedbear-devel at common-lisp.net
>> http://common-lisp.net/cgi-bin/mailman/listinfo/armedbear-devel
>>
>
> _______________________________________________
> armedbear-devel mailing list
> armedbear-devel at common-lisp.net
> http://common-lisp.net/cgi-bin/mailman/listinfo/armedbear-devel
>