www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - two semantic change proposals

reply davidl <davidl nospam.org> writes:
I believe the following allow the runtime reflection wrapper. Though it  
changes the old semantics, the old one can be implemented on top of the  
new semantics

http://d.puremagic.com/issues/show_bug.cgi?id=2868


-- 
使用 Opera 革命性的电子邮件客户程序: http://www.opera.com/mail/
Apr 21 2009
next sibling parent reply "Steven Schveighoffer" <schveiguy yahoo.com> writes:
On Tue, 21 Apr 2009 07:58:32 -0400, davidl <davidl nospam.org> wrote:


 I believe the following allow the runtime reflection wrapper. Though it  
 changes the old semantics, the old one can be implemented on top of the  
 new semantics

 http://d.puremagic.com/issues/show_bug.cgi?id=2868




That's cool.

However, you have misunderstood how the opDot function name should be  
passed via templates.

The opDot signature should not take a type as the first argument, but a  
string.  Otherwise, you have no compile-time abilities with the function  
name:

opDot(string methodname, T...)(T args)

instead of

opDot(U:immutable(char)[], T...)(U methodname, T args)

-Steve
Apr 21 2009
parent Don <nospam nospam.com> writes:
davidl wrote:

 在 Tue, 21 Apr 2009 20:21:05 +0800，Steven Schveighoffer 
 <schveiguy yahoo.com> 写道:
 

 On Tue, 21 Apr 2009 07:58:32 -0400, davidl <davidl nospam.org> wrote:


 I believe the following allow the runtime reflection wrapper. Though 
 it changes the old semantics, the old one can be implemented on top 
 of the new semantics

 http://d.puremagic.com/issues/show_bug.cgi?id=2868




 That's cool.

 However, you have misunderstood how the opDot function name should be 
 passed via templates.

 The opDot signature should not take a type as the first argument, but 
 a string.  Otherwise, you have no compile-time abilities with the 
 function name:

 opDot(string methodname, T...)(T args)

 instead of

 opDot(U:immutable(char)[], T...)(U methodname, T args)

 -Steve



 It's said there will be soon "static string" param.



I first proposed that about 3 years ago. I seriously don't believe it's 
going to happen any time soon. In fact, IIRC, Walter said that he'd 
found a problem with it. Maybe in D3.


 I think the prototype might best be refined that time.



We need something that works in D2.
Apr 21 2009
prev sibling parent davidl <davidl nospam.org> writes:
在 Tue, 21 Apr 2009 20:21:05 +0800，Steven Schveighoffer  
<schveiguy yahoo.com> 写道:


 On Tue, 21 Apr 2009 07:58:32 -0400, davidl <davidl nospam.org> wrote:


 I believe the following allow the runtime reflection wrapper. Though it  
 changes the old semantics, the old one can be implemented on top of the  
 new semantics

 http://d.puremagic.com/issues/show_bug.cgi?id=2868




 That's cool.

 However, you have misunderstood how the opDot function name should be  
 passed via templates.

 The opDot signature should not take a type as the first argument, but a  
 string.  Otherwise, you have no compile-time abilities with the function  
 name:

 opDot(string methodname, T...)(T args)

 instead of

 opDot(U:immutable(char)[], T...)(U methodname, T args)

 -Steve



It's said there will be soon "static string" param.

I think the prototype might best be refined that time.


-- 
使用 Opera 革命性的电子邮件客户程序: http://www.opera.com/mail/
Apr 21 2009