Okay, I just don't get it.

This code below doesn't compile in V1 dmd.

version(D_Version2)
{
    alias const(wchar)[] wstring;
    alias const(dchar)[] dstring;
}
else
{
    alias char[] string;
    alias wchar[] wstring;
    alias dchar[] dstring;
}


I can understand why - because the first two alias declarations have invalid V1 syntax - but doesn't that make the whole D_Version2 thing pointless?

-- 
Derek Parnell
Melbourne, Australia
skype: derek.j.parnell

On Sun, 2 Sep 2007 22:50:32 +1000, Derek Parnell wrote:

P.S. I finally got Bud to compile using V2 after about a thousand tiny changes and now I want to tweak so I can also compile the same source in V1.

Needless to say, but there will be a slight delay in the next Bud release until I sort out this next bloody irritation.

-- 
Derek Parnell
Melbourne, Australia
skype: derek.j.parnell

Derek Parnell wrote:
> On Sun, 2 Sep 2007 22:50:32 +1000, Derek Parnell wrote:
> 
> P.S. I finally got Bud to compile using V2 after about a thousand tiny
> changes and now I want to tweak so I can also compile the same source in
> V1. 
> 
> Needless to say, but there will be a slight delay in the next Bud release
> until I sort out this next bloody irritation.
> 

Can you work around it by separating the aliases out into separate v1 and v2 d-files?  Like

version(D_Version2)
{
   import v2types;
}
else
{
   import v1types;
}

--bb

Dnia Sun, 2 Sep 2007 22:50:32 +1000
Derek Parnell <derek@psych.ward> napisał/a:

> Okay, I just don't get it.
> 
> This code below doesn't compile in V1 dmd.
> 
> version(D_Version2)
> {
>     alias const(wchar)[] wstring;
>     alias const(dchar)[] dstring;
> }
> else
> {
>     alias char[] string;
>     alias wchar[] wstring;
>     alias dchar[] dstring;
> }
> 
> 
> I can understand why - because the first two alias declarations have invalid V1 syntax - but doesn't that make the whole D_Version2 thing pointless?

Unfortunetly this is correct behavior.
Even when D_Version2 is undefined, anything which is inside must be
syntacticly correct. const(wchar) is incorrect for D1.

:/
-- 
Witold Baryluk, aleph0
MAIL: baryluk@smp.if.uj.edu.pl
JID: movax@jabber.autocom.pl

Derek Parnell wrote:
> This code below doesn't compile in V1 dmd.
> [...]
> I can understand why - because the first two alias declarations have
> invalid V1 syntax - but doesn't that make the whole D_Version2 thing
> pointless?

Um, I see it's a pain. Surely you know the string mixin workaround?

Regards, Frank

0ffh wrote:
> Um, I see it's a pain. Surely you know the string mixin workaround?

Someone posted it somewhere at some time.
I currently use this for strings:

---<snip>---

version (D_Version2)
{
	mixin("alias invariant(char)* icptr;");
	alias char* cptr;
}
else
{
	alias char*  icptr;
	alias char*  cptr;
	alias char[] string;
}

---<snap>---

Or were you just complaining about the lack of elegance here?

Regards, Frank

0ffh wrote:
> Derek Parnell wrote:
> 
>> This code below doesn't compile in V1 dmd.
>> [...]
>> I can understand why - because the first two alias declarations have
>> invalid V1 syntax - but doesn't that make the whole D_Version2 thing
>> pointless?
> 
> 
> Um, I see it's a pain. Surely you know the string mixin workaround?
> 
> Regards, Frank
> 
> 

For the record, the string mixin workaround looks like this:

version (D_Version2) {
    mixin("alias const(char)* c_str;");
} else {
    alias char* c_str;
}

You can find this code near the top of the Python header.
http://dsource.org/projects/pyd/browser/trunk/infrastructure/python/python.d

So far as I can tell, I was the first person to come up with (or at least use) this workaround, but others may have come up with it on their own.

-- 
Kirk McDonald
http://kirkmcdonald.blogspot.com
Pyd: Connecting D and Python
http://pyd.dsource.org

Derek Parnell wrote:
> Okay, I just don't get it.
> 
> This code below doesn't compile in V1 dmd.
> 
> version(D_Version2)
> {
>     alias const(wchar)[] wstring;
>     alias const(dchar)[] dstring;
> }
> else
> {
>     alias char[] string;
>     alias wchar[] wstring;
>     alias dchar[] dstring;
> }
> 
> 
> I can understand why - because the first two alias declarations have
> invalid V1 syntax - but doesn't that make the whole D_Version2 thing
> pointless?
> 

Have you read this thread?  Thomas Kuehne explains how to write code that works in both D 1 and 2.

http://www.digitalmars.com/d/archives/digitalmars/D/supporting_DMD-1.016_and_DMD-2.000_with_the_same_source_code_54628.html