 |
C++Talk.NET
C++ language newsgroups
|
| View previous topic :: View next topic |
| Author |
Message |
Don Ziesig
Guest
|
 Posted: Thu Nov 10, 2005 5:56 pm Post subject: template needs to handle invalid floating point without comp |
 |
|
Hi!
I need a method of implementing the following concept without
compilation errors:
template <typename C>
C myclass<C>::and(C x, C y)
{
if( not floating point C ) // tested with if(0)
{
return x & y;
}
else
{
return 0;
}
}
Unfortunately, even though I get warned about unreachable code in the
test case, the compiler still rejects a floating point x & y;
Thanks,
Don Ziesig
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
 |
Jeffrey Schwab
Guest
|
| Posted: Thu Nov 10, 2005 8:24 pm Post subject: Re: template needs to handle invalid floating point without |
|
|
Don Ziesig wrote:
| Quote: | Hi!
I need a method of implementing the following concept without
compilation errors:
template <typename C
C myclass
{
if( not floating point C ) // tested with if(0)
{
return x & y;
}
else
{
return 0;
}
}
Unfortunately, even though I get warned about unreachable code in the
test case, the compiler still rejects a floating point x & y;
|
Rather than using if/else, just specialize the template for special
cases. E.g:
template< typename C >
C bitwise_and( C x, C y )
{
return x & y;
}
template< >
double bitwise_and( double x, double y )
{
return 0;
}
#include <iostream>
int main( )
{
std::cout << bitwise_and( 3, 5 ) << "n";
std::cout << bitwise_and( 3.0, 5.0 ) << "n";
}
And be careful: "and" is a keyword.
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
richard@ex-parrot.com
Guest
|
| Posted: Thu Nov 10, 2005 8:24 pm Post subject: Re: template needs to handle invalid floating point without |
|
|
Don Ziesig wrote:
| Quote: |
I need a method of implementing the following concept without
compilation errors:
template <typename C
C myclass
|
First of all, "and" cannot be used as a function name as it is an
alternative spelling of the token &&.
| Quote: | {
if( not floating point C ) // tested with if(0)
|
Do you mean that you need to test whether C is a valid floating point
type? If so, dispatching via an is_floating_point type traits such as
the one in boost or TR1 will do this:
template <typename C>
C myclass<C>::my_and_2(C x, C y, std::tr1::true_type) {
return 0; // It is a floating point
}
template <typename C>
C myclass<C>::my_and_2(C x, C y, std::tr1::false_type) {
return x & y; // It isn't a floating point
}
template <typename C>
C myclass<C>::my_and(C x, C y) {
return my_and_2(x, y, std::tr1::is_floating_point<C>() );
}
--
Richard Smith
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
Dan McLeran
Guest
|
| Posted: Thu Nov 10, 2005 8:24 pm Post subject: Re: template needs to handle invalid floating point without |
|
|
You can't parametize templates with floats.
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
mlimber
Guest
|
| Posted: Fri Nov 11, 2005 12:53 am Post subject: Re: template needs to handle invalid floating point without |
|
|
Don Ziesig wrote:
| Quote: | Hi!
I need a method of implementing the following concept without
compilation errors:
template <typename C
C myclass
{
if( not floating point C ) // tested with if(0)
{
return x & y;
}
else
{
return 0;
}
}
Unfortunately, even though I get warned about unreachable code in the
test case, the compiler still rejects a floating point x & y;
|
Try partial specialization:
template <typename C>
C and ( const C x, const C y )
{
return x & y;
}
template<>
float and<>( const float, const float )
{
return 0;
}
You can add specializations for other types (e.g. double), also, and
Boost and Loki each provide type lists and type traits, which might
make the job simpler for a larger number of specializations.
Cheers! --M
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
Dan McLeran
Guest
|
| Posted: Fri Nov 11, 2005 9:04 am Post subject: Re: template needs to handle invalid floating point without |
|
|
Oops, i misunderstood the question. I thought you were asking if you
could do this:
template<float f>
Sorry.
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
Don Ziesig
Guest
|
| Posted: Fri Nov 11, 2005 9:07 am Post subject: Re: template needs to handle invalid floating point without |
|
|
Jeffrey Schwab wrote:
| Quote: | Don Ziesig wrote:
Hi!
I need a method of implementing the following concept without
compilation errors:
template <typename C
C myclass
{
if( not floating point C ) // tested with if(0)
{
return x & y;
}
else
{
return 0;
}
}
Unfortunately, even though I get warned about unreachable code in the
test case, the compiler still rejects a floating point x & y;
Rather than using if/else, just specialize the template for special
cases. E.g:
template< typename C
C bitwise_and( C x, C y )
{
return x & y;
}
template
double bitwise_and( double x, double y )
{
return 0;
}
#include
int main( )
{
std::cout << bitwise_and( 3, 5 ) << "n";
std::cout << bitwise_and( 3.0, 5.0 ) << "n";
}
And be careful: "and" is a keyword.
|
Hello again!
I was using the pseudo code as an example of what I wanted to do [I
really don't have "and" as a function name  ].
As it turns out, what I thought might be a quick and easy fix (it looked
like there was a lot of code that would have to be changed if I couldn't
do what I wanted) turned out to be a quick and easier fix when the
suggestion of specialization was made earlier in the thread. I was able
to eliminate over 1000 lines of miserable hacking by re-ordering the
inheritance chain and eliminating the floating point logicals (and other
compilation-error producing code) by specializing logical ops in the
integer type descendants only!
Thanks all!!!!!!
Donz
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
wkaras@yahoo.com
Guest
|
| Posted: Fri Nov 11, 2005 9:12 am Post subject: Re: template needs to handle invalid floating point without |
|
|
Don Ziesig wrote:
| Quote: | Hi!
I need a method of implementing the following concept without
compilation errors:
template <typename C
C myclass
{
if( not floating point C ) // tested with if(0)
{
return x & y;
}
else
{
return 0;
}
}
.... |
Here is another approach.
template <typename C, bool Whole>
struct bit_and_
{ C operator () (C x, C y) { return(x & y); } };
template <typename C>
struct bit_and_<C, false>
{ C operator () (C x, C y) { return(0); } };
template <typename C>
C bit_and(C x, C y)
{ bit_and_<C, (((C(1) / 2) * 4) == C(0))> ba;
return(ba(x, y)); }
A simplified alternative:
template <typename C>
C bit_and(C x, C y)
{ return( bit_and_<C, ((C(1) / 2) * 4) == C(0)>()(x, y));
would only compile with a newer version of GCC.
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
E. Mark Ping
Guest
|
| Posted: Sat Nov 12, 2005 10:31 am Post subject: Re: template needs to handle invalid floating point without |
|
|
In article <1131646388.140873.106720 (AT) z14g2000cwz (DOT) googlegroups.com>,
<richard (AT) ex-parrot (DOT) com> wrote:
| Quote: | Do you mean that you need to test whether C is a valid floating point
type? If so, dispatching via an is_floating_point type traits such as
the one in boost or TR1 will do this:
template <typename C
C myclass
return my_and_2(x, y, std::tr1::is_floating_point<C>() );
|
Or you can use the existing numeric_limits<C>::is_integer
--
Mark Ping
[email]emarkp (AT) soda (DOT) CSUA.Berkeley.EDU[/email]
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
Richard Smith
Guest
|
| Posted: Sun Nov 13, 2005 3:35 pm Post subject: Re: template needs to handle invalid floating point without |
|
|
E. Mark Ping wrote:
| Quote: | template <typename C
C myclass
return my_and_2(x, y, std::tr1::is_floating_point<C>() );
Or you can use the existing numeric_limits<C>::is_integer
|
How will that help? I want to find out whether it's a floating point
number, not whether it's an integer. Plenty of types besides floating
points numbers and integers support operator& -- a std::bitset does,
for example; so do enums (after a fashion and they can be readily made
to work better).
If you don't want to rely on a TR1 component (or, alternatively, a
boost component), it's easy enough to roll your own is_floating_point
class.
--
Richard Smith
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]
|
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
