It would depend upon implementation, but I don't think getting the end
through the function is likely to be much slower than using a temporary
local store as a comparison. The idea is right, but it is more than likely
that in compiler optimisation that the call to l_int.end() would be inlined,
and take very little code. If end is of constant value it may even be
possible that the compiler would account for it, although I couldn't be
sure. The call to end would likely be just a return of a reference stored in
the container object, possibly a constant reference.
The only time it would be of any significance would be if you were creating
time critical code, and in that case it would be best to avoid using a list.
Also if the list was changed within the for loop then it's possible that the
stop value would become invalid, although that would be less likely in a
list than in other containers.
After saying all that I would agree that in theory the second version would
be very slightly more efficient, but on the other hand it involves more
code, creates an extra variable (iterator), which takes up some amount of
memory. It may even be that the creation of the const_iterator takes more
time than you gain in the for loop, although that is probably unlikely.
As usualy I've gone overkill in this reply, oops - sorry!
Jason.
----- Original Message -----
From: "Luigi Santillo" <gi...@li...>
To: <dev...@li...>
Sent: Tuesday, October 16, 2001 6:40 PM
Subject: [Dev-C++] a trivial question on for loops
> Hello everybody,
>
> I know this is an easy one, but what I've seen on a C++ book puzzled me.
>
> Consider the two pieces of code that follow:
>
> 1) as found on a recent C++ book
> list<int> l_int;
> // ... fill the list with a few intergers
> for (list<int>::const_iterator i = l_int.begin(), i != l_int.end(); ++i)
> // ... do something
>
> 2) as I used to know
> list<int> l_int;
> // ... fill the list with a few intergers
> list<int>::const_iterator stop = l_int.end();
> for (list<int>::const_iterator i = l_int.begin(), i != stop; ++i)
> // ... do something
>
> The two pieces are equivalent, but the second one is more efficient. Do
you
> agree with me?
>
> --
> gisan
> gi...@li...
>
>
> _______________________________________________
> Dev-cpp-users mailing list
> Dev...@li...
> https://lists.sourceforge.net/lists/listinfo/dev-cpp-users
|
