http://d.puremagic.com/issues/show_bug.cgi?id=1314

           Summary: Dupping an empty array creates a null array
           Product: D
           Version: 2.002
          Platform: PC
        OS/Version: Windows
            Status: NEW
          Keywords: wrong-code
          Severity: normal
          Priority: P2
         Component: DMD
        AssignedTo: bugzilla@digitalmars.com
        ReportedBy: brunodomedeiros+bugz@gmail.com


Dupping an empty array creates a null array:
----  ----
import std.stdio;

void main(char[][] args) {
        writefln("" is null);    // false
        writefln("".dup is null); // true
        writefln("".idup is null); // true
        writefln("".dup.ptr); // 0000
}


-- 

http://d.puremagic.com/issues/show_bug.cgi?id=1314





------- Comment #1 from bugzilla@digitalmars.com  2007-07-05 13:31 -------
Why is this a problem?


-- 

http://d.puremagic.com/issues/show_bug.cgi?id=1314





------- Comment #2 from brunodomedeiros+bugz@gmail.com  2007-07-05 17:56 -------
I tought you had changed D some releases ago, such that empty arrays became different than null arrays, wasn't that the case? (if not this is not a bug)


-- 

d-bugmail@puremagic.com wrote:
> http://d.puremagic.com/issues/show_bug.cgi?id=1314
> 
> ------- Comment #2 from brunodomedeiros+bugz@gmail.com  2007-07-05 17:56 -------
> I tought you had changed D some releases ago, such that empty arrays became
> different than null arrays, wasn't that the case? (if not this is not a bug)

He only changed it so that setting .length to 0 no longer sets .ptr to null. He didn't change anything else regarding empty arrays AFAIK.

http://d.puremagic.com/issues/show_bug.cgi?id=1314





------- Comment #4 from leikeze@gmail.com  2007-07-05 18:32 -------
(In reply to comment #0)
> Dupping an empty array creates a null array:
> ----  ----
> import std.stdio;
> 
> void main(char[][] args) {
>         writefln("" is null);    // false
>         writefln("".dup is null); // true
>         writefln("".idup is null); // true
>         writefln("".dup.ptr); // 0000
> }
> 

According to http://www.digitalmars.com/d/arrays.html shouldn't "".length be 1?
> String literals already have a 0 appended to them, so can be used directly
So shouldn't "".dup return a char array of length 1, containing the null character? Or is "" a special case that is interpreted as null?


-- 

http://d.puremagic.com/issues/show_bug.cgi?id=1314





------- Comment #5 from chris@dprogramming.com  2007-07-05 18:57 -------
Similar to why an empty struct should still allocate a byte.


-- 

http://d.puremagic.com/issues/show_bug.cgi?id=1314


brunodomedeiros+bugz@gmail.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




------- Comment #6 from brunodomedeiros+bugz@gmail.com  2007-07-06 04:03 -------
(In reply to comment #3)
> d-bugmail@puremagic.com wrote:
> > http://d.puremagic.com/issues/show_bug.cgi?id=1314
> > 
> > ------- Comment #2 from brunodomedeiros+bugz@gmail.com  2007-07-05 17:56 -------
> > I tought you had changed D some releases ago, such that empty arrays became different than null arrays, wasn't that the case? (if not this is not a bug)
> He only changed it so that setting .length to 0 no longer sets .ptr to null. He didn't change anything else regarding empty arrays AFAIK.

You're right, I misunderstood bug #429 .

(and "".length should be 0, there is no doubt about that)


--