www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.bugs - [Issue 11042] New: Inconsistent "static condition" behaviors

reply d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=11042

           Summary: Inconsistent "static condition" behaviors
           Product: D
           Version: D2
          Platform: All
        OS/Version: All
            Status: NEW
          Keywords: accepts-invalid, rejects-valid
          Severity: normal
          Priority: P2
         Component: DMD
        AssignedTo: nobody puremagic.com
        ReportedBy: k.hara.pg gmail.com



Inside static if condition, the logical operators || and && will be evaluated
lazily.

  static if (true || error) { pragma(msg, "OK"); }
  else                      { pragma(msg, "NG"); }
  // --> prints "OK"

This is useful feature, but it doesn't work on static assert condition.

  static assert(true || error);
  // --> Error: undefined identifier error

This is inconsistent. I think static assert should also evaluate the condition
lazily.

----

There's one associated bug.

  static if (is(typeof(true || error))) { pragma(msg, "OK"); }
  else                                  { pragma(msg, "NG"); }
  // --> prints "OK"

The expression "true || error" should not have valid type, but inside static if
condition, the lazy evaluation is incorrectly invoked. It's definitely a bug.

----

Template constraint is the one another place for "static condition".
Currently it behaves like "static if" condition. Its behavior should also be
fixed.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Sep 15 2013
next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=11042


Kenji Hara <k.hara.pg gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |pull



https://github.com/D-Programming-Language/dmd/pull/2558

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Sep 15 2013
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=11042


Andrej Mitrovic <andrej.mitrovich gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |andrej.mitrovich gmail.com



04:45:33 PDT ---
I would find it even more useful if we extended this to templates, for example
here's the workaround code I have to use in one of my libraries:

alias storages = ParameterStorageClassTuple;  // from Phobos

/** 
    Check whether $(D T) is a handler function which 
    can be called with the $(D Types). 
*/
template isEventHandler(T, Types...)
    if (isSomeFunction!T)
{
    alias stores = storages!T;

    // eager evaluation workaround
    static if (!stores.length)
        enum bool isEventHandler = false;
    else
        enum bool isEventHandler = is(typeof(T.init(Types.init))) &&
                                   stores[0] == stcType.scope_ &&
                                   is(ReturnType!T == void);
}

If I'm not using this "static if", the "stores[0]" code will issue a compiler
error (out of bounds) if there are no parameters. But,
"is(typeof(T.init(Types.init)))" would already fail before it, however it's not
lazy so "stores[0]" is evaluated eagerly and the compiler emits an error.

Can we do anything about this? Perhaps enabling lazy evaluation for manifest
constants in an eponymous template would work.. I'm not sure.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Sep 15 2013
prev sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=11042




04:46:13 PDT ---
Oh wait I just realized this is exactly what
http://d.puremagic.com/issues/show_bug.cgi?id=9073 was for, so it got rejected?
Shame..

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Sep 15 2013