<d>
bool b() { return true; }
bool quit;
if( quit = b() ) // <---------
    return 0;
</d>

DMD 0.114 reports:
.\test.d(17): '=' does not give a boolean result

I understand the intention but statement
does have boolean result.

Am I on the right side?

Andrew Fedoniouk.

On Sun, 6 Mar 2005 17:24:12 -0800, Andrew Fedoniouk wrote:

> <d>
> bool b() { return true; }
> bool quit;
> if( quit = b() ) // <---------
>     return 0;
> </d>
> 
> DMD 0.114 reports:
> .\test.d(17): '=' does not give a boolean result
> 

Just a comment, but you can fix this the long way ...

#  if( (quit = b()) == true  )

or the short way ...

#  if( 0, quit = b() )

-- 
Derek
Melbourne, Australia
7/03/2005 12:45:27 PM

This is a very annoying "feature" of dmd -- a blatant discrimination of <AssignExpression>!


Andrew Fedoniouk wrote:
> <d>
> bool b() { return true; }
> bool quit;
> if( quit = b() ) // <---------
>     return 0;
> </d>
> 
> DMD 0.114 reports:
> .\test.d(17): '=' does not give a boolean result
> 
> I understand the intention but statement
> does have boolean result.
> 
> Am I on the right side?
> 
> Andrew Fedoniouk.

Personally I like the feature, just not the error message, which, as Andrew demonstrates, claims something that is clearly not true.

Nick

In article <d0ggi4$1b9v$1@digitaldaemon.com>, zwang says...
>
>This is a very annoying "feature" of dmd -- a blatant discrimination of <AssignExpression>!
>
>
>Andrew Fedoniouk wrote:
>> <d>
>> bool b() { return true; }
>> bool quit;
>> if( quit = b() ) // <---------
>>     return 0;
>> </d>
>> 
>> DMD 0.114 reports:
>> .\test.d(17): '=' does not give a boolean result
>> 
>> I understand the intention but statement
>> does have boolean result.
>> 
>> Am I on the right side?
>> 
>> Andrew Fedoniouk.

Nick wrote:

>>>>.\test.d(17): '=' does not give a boolean result

>>This is a very annoying "feature" of dmd -- a blatant discrimination of <AssignExpression>!

> Personally I like the feature, just not the error message, which, as Andrew
> demonstrates, claims something that is clearly not true.

As long as 'true' is an integer 1, it's actually somewhat correct ;-)

And you should almost always write those "if" expressions out anyway...
If not by using two statements, then at least with an extra parenthesis.

--anders

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Derek Parnell schrieb am Mon, 7 Mar 2005 12:45:51 +1100:
> On Sun, 6 Mar 2005 17:24:12 -0800, Andrew Fedoniouk wrote:
>
>> <d>
>> bool b() { return true; }
>> bool quit;
>> if( quit = b() ) // <---------
>>     return 0;
>> </d>
>> 
>> DMD 0.114 reports:
>> .\test.d(17): '=' does not give a boolean result
>> 
>
> Just a comment, but you can fix this the long way ...
>
> #  if( (quit = b()) == true  )
>
> or the short way ...
>
> #  if( 0, quit = b() )

Where is this use of "," documented?

Thomas


-----BEGIN PGP SIGNATURE-----

iD8DBQFCOaIh3w+/yD4P9tIRAsMGAJsG5CLaiVIfC9Ud3OppNXec5h/MnQCeOmYC
7jpaCW5NEMg++MBnJbIOab0=
=l6yg
-----END PGP SIGNATURE-----

Thomas Kuehne wrote:

>>Just a comment, but you can fix this the long way ...
>>
>>#  if( (quit = b()) == true  )
>>
>>or the short way ...
>>
>>#  if( 0, quit = b() )
> 
> Where is this use of "," documented?

http://www.digitalmars.com/d/expression.html#Expression
> The left operand of the , is evaluated, then the right operand is
> evaluated. The type of the expression is the type of the right operand,
> and the result is the result of the right operand.

The "short" usage above is more like... gross, though ? :-)

--anders

On Thu, 17 Mar 2005 16:32:16 +0100, Anders F Björklund wrote:

> Thomas Kuehne wrote:
> 
>>>Just a comment, but you can fix this the long way ...
>>>
>>>#  if( (quit = b()) == true  )
>>>
>>>or the short way ...
>>>
>>>#  if( 0, quit = b() )
>> 
>> Where is this use of "," documented?
> 
> http://www.digitalmars.com/d/expression.html#Expression
>> The left operand of the , is evaluated, then the right operand is evaluated. The type of the expression is the type of the right operand, and the result is the result of the right operand.
> 
> The "short" usage above is more like... gross, though ? :-)
> 

Oh absolutely. I'd shoot anyone in my teams who'd try to get away with
that. I just wanted to highlight that DMD is not catching all the possible
ways you can (ab)use the 'if (A = B)' idiom.

-- 
Derek Parnell
Melbourne, Australia
18/03/2005 7:50:41 AM