[Rcpp-devel] potential new way to expose constructors

Hello,

I've just commited some code that will potentially make it simpler to expose
constructors.

Where previously we would do something like

.constructor( init_2<double,double>() )

we can now do:

.ctor<double,double>()

We probably don't want to keep both, so I'd like to keep the second solution
but to call it constructor, so that we will do:

.constructor<double,double>()

Is this ok for everybody ? I guess this is only mainly relevant for Doug,
Andrew and John anyway at the moment :-)

Romain

--
Romain Francois
Professional R Enthusiast
+33(0) 6 28 91 30 30
http://romainfrancois.blog.free.fr
|- http://bit.ly/9VOd3l : ZAT! 2010
|- http://bit.ly/c6DzuX : Impressionnism with R
`- http://bit.ly/czHPM7 : Rcpp Google tech talk on youtube

The init_2 is unnatural.  I would prefer the new syntax.  As I have to
code that is relying in the init_2 at the moment I'm fine with
switching it out.

Does this mean that we will be able to expose
multiple constructors?

-Andrew

On Sun, Nov 21, 2010 at 12:00 PM, Romain Francois
<romain at r-enthusiasts.com>  wrote:

Hello,

I've just commited some code that will potentially make it simpler to expose
constructors.

Where previously we would do something like

.constructor( init_2<double,double>() )

we can now do:

.ctor<double,double>()

We probably don't want to keep both, so I'd like to keep the second solution
but to call it constructor, so that we will do:

.constructor<double,double>()

Is this ok for everybody ? I guess this is only mainly relevant for Doug,
Andrew and John anyway at the moment :-)

Romain

Le 21/11/10 20:42, Andrew Redd a ?crit :

The init_2 is unnatural. ?I would prefer the new syntax. ?As I have to
code that is relying in the init_2 at the moment I'm fine with
switching it out.

Great. I'll wait for Doug's go and switch them.

Does this mean that we will be able to expose
multiple constructors?

definitely. as before, but with the new syntax:

.constructor<double,double>()
.constructor<int>()

... etc ...

-Andrew

On Sun, Nov 21, 2010 at 12:00 PM, Romain Francois
<romain at r-enthusiasts.com> ?wrote:

Hello,

I've just commited some code that will potentially make it simpler to
expose
constructors.

Where previously we would do something like

.constructor( init_2<double,double>() )

we can now do:

.ctor<double,double>()

We probably don't want to keep both, so I'd like to keep the second
solution
but to call it constructor, so that we will do:

.constructor<double,double>()

Is this ok for everybody ? I guess this is only mainly relevant for Doug,
Andrew and John anyway at the moment :-)

Romain

--
Romain Francois
Professional R Enthusiast
+33(0) 6 28 91 30 30
http://romainfrancois.blog.free.fr
|- http://bit.ly/9VOd3l : ZAT! 2010
|- http://bit.ly/c6DzuX : Impressionnism with R
`- http://bit.ly/czHPM7 : Rcpp Google tech talk on youtube

That's fine with me.

Sorry for the slow response.  I'm down with a
cold or the flu today.

On Sun, Nov 21, 2010 at 1:52 PM, Romain Francois
<romain at r-enthusiasts.com>  wrote:

Le 21/11/10 20:42, Andrew Redd a ?crit :

The init_2 is unnatural.  I would prefer the new syntax.  As I have to
code that is relying in the init_2 at the moment I'm fine with
switching it out.

Great. I'll wait for Doug's go and switch them.

Does this mean that we will be able to expose
multiple constructors?

definitely. as before, but with the new syntax:

.constructor<double,double>()
.constructor<int>()

... etc ...

-Andrew

On Sun, Nov 21, 2010 at 12:00 PM, Romain Francois
<romain at r-enthusiasts.com>    wrote:

Hello,

I've just commited some code that will potentially make it simpler to
expose
constructors.

Where previously we would do something like

.constructor( init_2<double,double>() )

we can now do:

.ctor<double,double>()

We probably don't want to keep both, so I'd like to keep the second
solution
but to call it constructor, so that we will do:

.constructor<double,double>()

Is this ok for everybody ? I guess this is only mainly relevant for Doug,
Andrew and John anyway at the moment :-)

Romain