[KPhotoAlbum] Inefficiency in ImageSearchInfo::match()

[Robert Krawitz]

On Tue, 11 Sep 2012 14:26:21 +0200, Jesper K. Pedersen wrote:
> Ahh  x &= y is equal to x = x & y NOT x = x && y;
>
> That can be fixed with rewriting the code to
>
>   b = b && trueFunction()
>
> and in the KPA code doing similar. Robert feel free to fix.

I'd still rather have it explicitly short-circuit, but at least this
would be better.

> On Tuesday 11 September 2012 13:58:21 Johannes Zarl wrote:
>> On Tuesday 11 September 2012 12:53:57 Jesper K. Pedersen wrote:
>> > On Sunday 09 September 2012 20:07:34 Robert Krawitz wrote:
>> > > Is there any reason it does this rather than simply return if any of
>> > > these tests fail?
>> > 
>> > I'm 99% sure that &= will not evaluate the right side if the left side
>> > already is false.
>> 
>> C++ actually has to evaluate the right side of the code. The arguments of
>> the operator&= are evaluated before operator&= is called.
>> 
>> Short example:
>> 
>> cat > test.cpp <<EOF
>> #include <iostream>
>> 
>> bool trueFunction()
>> {
>>     std::cout << "trueFunction called." <<std::endl;
>>     return true;
>> }
>> 
>> int main() {
>>     bool b= false;
>>     b &= trueFunction();
>>     std::cout << "b is: " << ( b ? "true" : "false" ) << std::endl;
>>     return 0;
>> }
>> EOF
>> 
>> g++ -o test test.cpp
>> ./test
>> trueFunction called.
>> b is: false
>> 
>> Cheers,
>>   Johannes

-- 
Robert Krawitz                                     <rlk at alum.mit.edu>

MIT VI-3 1987 - Congratulations MIT Engineers men's hoops Final Four!
Tall Clubs International  --  http://www.tall.org/ or 1-888-IM-TALL-2
Member of the League for Programming Freedom  --  http://ProgFree.org
Project lead for Gutenprint   --    http://gimp-print.sourceforge.net

"Linux doesn't dictate how I work, I dictate how Linux works."
--Eric Crampton