• David Nadlinger (9/9) Apr 29 2011 [repost from dmd-internals]
• Walter Bright (2/6) Apr 30 2011 int
• Andrei Alexandrescu (3/12) Apr 30 2011 I think he was surprised that the current implementation returns size_t.

David Nadlinger <see klickverbot.at> writes:

[repost from dmd-internals]

What should the _aApply... runtime functions, or in general, opApply() 
return? According to the current runtime implementation and 
http://www.digitalmars.com/d/2.0/statement.html, the return type should 
be int, but DMD expects Type::tindex (i.e. size_t) in 
ForeachStatement::semantic().

This is relevant to LDC, where the type is reflected in the backend 
(well, LLVM IR).

David

Walter Bright <newshound2 digitalmars.com> writes:

On 4/29/2011 5:02 PM, David Nadlinger wrote:
 What should the _aApply... runtime functions, or in general, opApply() return?
 According to the current runtime implementation and
 http://www.digitalmars.com/d/2.0/statement.html, the return type should be int,
 but DMD expects Type::tindex (i.e. size_t) in ForeachStatement::semantic().


int

Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:

On 4/30/11 12:51 PM, Walter Bright wrote:
 On 4/29/2011 5:02 PM, David Nadlinger wrote:
 What should the _aApply... runtime functions, or in general, opApply()
 return?
 According to the current runtime implementation and
 http://www.digitalmars.com/d/2.0/statement.html, the return type
 should be int,
 but DMD expects Type::tindex (i.e. size_t) in
 ForeachStatement::semantic().


 int

I think he was surprised that the current implementation returns size_t.

Andrei