• Pillsy <pillsbury gmail.com> Aug 12 2010
• dsimcha <dsimcha yahoo.com> Aug 12 2010

Pillsy <pillsbury gmail.com> writes:

Steven Schveighoffer Wrote:
[...]
 Hold on, can't we have enforce and lenforce (lazy enforce)?


Or couldn't we have an overload for enforce that takes a string, and another
overload that takes a void delegate returning a string? 

It makes the syntax a little grottier, but...

  From a simple grep, 99% of enforce instances are:


 enforce(condition)
   -or-
 enforce(condition, "compile-time-string")
   -or-
 enforce(condition, "compile-time-string" ~ type.stringof)


...if you have to write

    enforce(condition, 
               (){ return "compile-time-string" ~ type.stringof; });

one time in a hundred, is it really that bad? 
 
Cheers,
Pillsy

dsimcha <dsimcha yahoo.com> writes:

== Quote from Pillsy (pillsbury gmail.com)'s article
 ...if you have to write
     enforce(condition,
                (){ return "compile-time-string" ~ type.stringof; });
 one time in a hundred, is it really that bad?
 Cheers,
 Pillsy


I think the point was that even this should be automagically evaluated at
compile
time, since type.stringof is a compile time constant.