1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
|
This program was create so that the author Paul Elliott could learn
gtkmm programming.
While working on this project I found the following problem
and the pending SigC solution
To: Paul Elliott <pelliott@io.com>
Cc: libsigc-list@gnome.org
Subject: Re: [sigc] binding a arguement which is a reference
Am 2004.02.03 14:07 schrieb(en) Paul Elliott:
> I found in the archives the following message: See the end of this
> post:
>
>
> I am using sigc++-1.2 and I get a similar message to that mentioned
> below
> whenever the parameter I am trying to bind is a reference
> in the method mentioned in the "slot(...)" call.
>
> Is this a restriction or a bug?? Does the code need to be fixed
> or a note put in the documentation the the parameter that you bind
> can not be a reference in the called method?
It is a restriction that will be lifted in sigc++-2.0. You might want
to test the latest beta release sigc++-1.9.13. Of course, you are
welcome to patch the documentation for sigc++-1.2!
> Could this be an example of the famous reference to reference
> problem??
If I remember correctly, the problem with sigc++-1.2 is that the
bind adaptor explicitly calls the destructors of its bound objects.
This doesn't work when the objects are references.
> I made a workaround in my code by changing the called method,
> replace the referenced type T& to boost::reference_wrapper<T>.
> I got reference_wrapper from the boost libraries.
> The referenced_wrapper type has "value" semantics in that it
> has default constructors, copy constructors and assignment operators.
>
> What is the official libsigc+ approach to this problem?
In sigc++-2.0 you would pass the object through sigc::ref()
so that sigc::bind() can deduce that the value should be bound as
a reference and not as a value.
Cheers!
Martin
> Quoted message follows:
> ==================cut here with a chain saw==================
> From: gtkmm mailing list account <gtkmm-list bernard-hugueney org>
>
> To: Martin Schulze <martin-ml hippogriff de>
>
> Cc: gtkmm-list gnome org, libsigc-list gnome org
>
> Subject: [sigc] Re: binding by value, slot taking reference was Re:
> [gtkmm] binding arguments to slots
>
> Date: Tue, 18 Nov 2003 11:23:31 +0100
>
>
>
>
> Le Mardi 18 Novembre 2003 10:26, Martin Schulze a écrit :
> > >
> > > There is still something I don't understand with slot and bind :
> > >
> > > I would like to make a slot for a function taking arguments by
> > > reference:
> > > program :
> > > "
> > > using namespace std;
> > > using namespace SigC;
> > >
> > > void foo1(int& i){++i;}
> > >
> > > int main(int argc, char** argv){
> > > int i=1;
> > > Slot0<void>slot1=bind(slot(foo1),i);
> > > return 0;
> > > }
> >
> > Try to specify the argument type as template parameter, i.e.:
> >
> > Slot0<void> slot1 = bind<int&>(slot(foo1), i);
>
> I shoudl have said I already tried that :-(
> >
> > I'm not sure if it works, though - let us know!
>
> same error:
>
> /usr/include/sigc++-1.2/sigc++/bind.h: Dans static member function
> �
>
> static void SigC::AdaptorBindData1_<C1>::dtor(void*) [with C1 =
> int&] :
> /usr/include/sigc++-1.2/sigc++/bind.h:167: instantiated from
> `SigC::Slot0<R> SigC::bind(const SigC::Slot1<R, C1>&, A1) [with A1 =
> int&, R = void, C1 = int&]'
> testLibSigC.cxx:9: instantiated from here
> /usr/include/sigc++-1.2/sigc++/bind.h:116: error: destructor name
> `~int&' does not match type `int' of expression
>
> >
> > BTW in libsigc++-2.0 you will be able to write:
> >
> > sigc::slot<void> slot1 = sigc::bind(foo1, sigc::ref(i));
> >
> > instead.
>
> Cool, I'll be missing my boost libs less :-)
> (For the record, I also tried boost::ref(i) to no avail :
> /usr/include/sigc++-1.2/sigc++/bind.h: Dans static member function
> static void SigC::AdaptorBindData1_<C1>::dtor(void*) [with C1 =
> int&] :
> /usr/include/sigc++-1.2/sigc++/bind.h:167: instantiated from
> `SigC::Slot0<R> SigC::bind(const SigC::Slot1<R, C1>&, A1) [with A1 =
> boost::reference_wrapper<int>, R = void, C1 = int&]'
> testLibSigC.cxx:11: instantiated from here
> /usr/include/sigc++-1.2/sigc++/bind.h:116: error: destructor name
> `~int&' does not match type `int' of expression
>
> Now I'm really holding my breath for libsigc++-2.0 in gtkmm.
>
> Thanks again for your kind help !
>
> Bernard
>
> --
> Paul Elliott 1(512)837-1096
> pelliott@io.com PMB 181, 11900 Metric Blvd Suite J
> http://www.io.com/~pelliott/pme/ Austin TX 78758-3117
>
_______________________________________________
libsigc-list mailing list
libsigc-list@gnome.org
http://mail.gnome.org/mailman/listinfo/libsigc-list
We are all interested in the future because that is where you and I
are going to spend the rest of our lives. Criswell
My friend future events such as these will affect you in the future. Criswell.
Can you prove it did not happen? Criswell.
What do you think will be the next obstacle the Earth people will put
in our way? Janis.
Well as long as they can think we'll have our problems. But those
whom were using cannot think. They are the dead. Brought to simulated
life by our electrode guns. You know, it's an interesting thing when
you consider the Earth people who can think are so frightened by
those who cannot, the dead. Eros.
You see, you see, your stupid minds, stupid, stupid. Eros.
|
