Quelle est la signification de "operator bool () const"

Réponses:

141

Fonctions des membres du formulaire 

operator TypeName()

sont des opérateurs de conversion. Ils permettent aux objets de type classe d'être utilisés comme s'ils étaient de type TypeNameet lorsqu'ils le sont, ils sont convertis en TypeNameutilisant la fonction de conversion.

Dans ce cas particulier, operator bool()permet d'utiliser un objet de type classe comme s'il s'agissait d'un bool. Par exemple, si vous avez un objet du type de classe nommé obj, vous pouvez l'utiliser comme

if (obj)

Cela appellera le operator bool(), retournera le résultat et utilisera le résultat comme condition du if.

Il convient de noter que operator bool()c'est une très mauvaise idée et que vous ne devriez vraiment jamais l'utiliser. Pour une explication détaillée des raisons pour lesquelles il est mauvais et pour la solution au problème, consultez «L'idiome Safe Bool».

(C ++ 0x, la prochaine révision de la norme C ++, ajoute la prise en charge des opérateurs de conversion explicites. Ceux-ci vous permettront d'écrire un coffre explicit operator bool()- fort qui fonctionne correctement sans avoir à sauter à travers les obstacles de l'implémentation de Safe Bool Idiom.)

James McNellis
la source
1
"utilisé comme s'il s'agissait d'un booléen" implique à tort que vous pourriez, par exemple, lui attribuer une valeur booléenne. Au contraire, dans le code de l'affiche, il génère une variable temporaire de type bool qui se rapporte à la valeur momentanée de col mais qui est par la suite indépendante de l'objet qui l'a créée. De plus, la mention du Safe Bool Idiom est excellente, mais juste pour enregistrer que des opinions contraires existent: à mon humble avis, le conseil "ne jamais vraiment l'utiliser" est exagéré - il donne des contrôles plus stricts du compilateur contre une mauvaise utilisation stupide au prix d'une API plus obscurcie cela peut conduire à une mauvaise utilisation accidentelle.
Tony Delroy
1
@Tony: Eh bien, il peut être utilisé comme s'il s'agissait d'un booléen; puisque le résultat de la conversion est une rvalue ( bool), non, vous ne pouvez pas lui attribuer. S'il s'agissait d'une lvalue modifiable (par exemple bool&), vous pourriez lui assigner. En ce qui concerne l'exactitude, je soutiens que an operator bool()est toujours incorrect car il permet l'utilisation d'un objet de type classe dans un grand nombre de situations où vous ne voulez jamais qu'il soit utilisé. Safe Bool est une alternative bien supérieure.
James McNellis
1
Donc, selon le dernier paragraphe d'aujourd'hui, il est parfaitement OK à utiliser explicit operator bool(). Est-ce que je comprends bien?
Zingam
1
Le comité C ++ semble être en désaccord avec vous sur l'opérateur bool (). Au moins pour la dernière version de la norme (par exemple en.cppreference.com/w/cpp/utility/optional ). Ou peut-être vouliez-vous dire que seul le code STL devrait être autorisé à l'utiliser?
Joe Steele
1
@JoeSteele - Voir stackoverflow.com/a/16615725/2492801 . Les opérateurs de conversion explicite sont sûrs!
Benjamin Bihler
9 
operator bool() const 
{
    return col != 0;
}

Définit comment la classe peut être convertie en valeur booléenne, constaprès le ()est utilisé pour indiquer que cette méthode ne mute pas (changez les membres de cette classe).

Vous utiliserez généralement ces opérateurs comme suit:

airplaysdk sdkInstance;
if (sdkInstance) {
    std::cout << "Instance is active" << std::endl;
} else {
    std::cout << "Instance is in-active error!" << std::endl;
}
langerra.com
la source
7

Je voudrais donner plus de codes pour que ce soit clair.

struct A
{
    operator bool() const { return true; }
};

struct B
{
    explicit operator bool() const { return true; }
};

int main()
{
    A a1;
    if (a1) cout << "true" << endl; // OK: A::operator bool()
    bool na1 = a1; // OK: copy-initialization selects A::operator bool()
    bool na2 = static_cast<bool>(a1); // OK: static_cast performs direct-initialization

    B b1;     
    if (b1) cout << "true" << endl; // OK: B::operator bool()
    // bool nb1 = b1; // error: copy-initialization does not consider B::operator bool()
    bool nb2 = static_cast<bool>(b1); // OK: static_cast performs direct-initialization
}
démarrage automatique
la source
3

C'est implicitune fonction de conversion définie par l'utilisateur pour convertir votre classe en soit trueou false.

//usage
bool value = yourclassinstance; //yourclassinstance is converted into bool!
Nawaz
la source
1

C'est une conversion implicite en bool. Autrement dit, partout où les conversions implicites sont autorisées, votre classe peut être convertie boolen appelant cette méthode.

Billy ONeal
la source
1

Comme les autres l'ont dit, c'est pour la conversion de type, dans ce cas en a bool. Par exemple:

class A {
    bool isItSafe;

public:
    operator bool() const
    {
        return isItSafe;
    }

    ...
};

Maintenant, je peux utiliser un objet de cette classe comme s'il s'agissait d'un booléen:

A a;
...
if (a) {
    ....
}
chrisaycock
la source
1

En écrivant mon propre unique_ptr, j'ai trouvé ce cas. Compte tenu de std::unique_ptr« soperator== :

template<class T1, class D1, class T2, class D2>
bool operator==(const unique_ptr<T1, D1>& x, const unique_ptr<T2, D2>& y);

template <class T, class D>
bool operator==(const unique_ptr<T, D>& x, nullptr_t) noexcept;

template <class T, class D>
bool operator==(nullptr_t, const unique_ptr<T, D>& x) noexcept;

Et ce cas de test de libstdcxx :

  std::unique_ptr<int> ptr;
  if (ptr == 0)
    { }
  if (0 == ptr)
    { }
  if (ptr != 0)
    { }
  if (0 != ptr)
    { }

Notez parce que cela ptra un explicit operator bool() const noexcept;, donc operator overload resolutionfonctionne bien ici, par exemple, ptr == 0choisit

 template <class T, class D>
 bool operator==(const unique_ptr<T, D>& x, nullptr_t) noexcept;`.

S'il n'y a pas de explicitmot-clé ici, ptrin ptr == 0sera converti en bool, puis boolsera converti en int, car bool operator==(int, int)est intégré et 0est int. Ce qui nous attend, c'est une erreur de résolution de surcharge ambiguë.

Voici un exemple minimal, complet et vérifiable :

#include <cstddef>
struct A
{
    constexpr A(std::nullptr_t) {}
    operator bool() 
    {
        return true;
    }
};

constexpr bool operator ==(A, A) noexcept
{
    return true;
}

constexpr bool operator ==(A, std::nullptr_t) noexcept
{
    return true;
}

constexpr bool operator ==(std::nullptr_t, A) noexcept
{
    return true;
}

int main()
{
    A a1(nullptr);
    A a2(0);
    a1 == 0;
}

gcc :

prog.cc: In function 'int main()':
prog.cc:30:8: error: ambiguous overload for 'operator==' (operand types are 'A' and 'int')
   30 |     a1 == 0;
      |     ~~ ^~ ~
      |     |     |
      |     A     int
prog.cc:30:8: note: candidate: 'operator==(int, int)' <built-in>
   30 |     a1 == 0;
      |     ~~~^~~~
prog.cc:11:16: note: candidate: 'constexpr bool operator==(A, A)'
   11 | constexpr bool operator ==(A, A) noexcept
      |                ^~~~~~~~
prog.cc:16:16: note: candidate: 'constexpr bool operator==(A, std::nullptr_t)'
   16 | constexpr bool operator ==(A, std::nullptr_t) noexcept
      |                ^~~~~~~~

clang :

prog.cc:30:8: error: use of overloaded operator '==' is ambiguous (with operand types 'A' and 'int')
    a1 == 0;
    ~~ ^  ~
prog.cc:16:16: note: candidate function
constexpr bool operator ==(A, std::nullptr_t) noexcept
               ^
prog.cc:11:16: note: candidate function
constexpr bool operator ==(A, A) noexcept
               ^
prog.cc:30:8: note: built-in candidate operator==(int, int)
    a1 == 0;
       ^
prog.cc:30:8: note: built-in candidate operator==(float, int)
prog.cc:30:8: note: built-in candidate operator==(double, int)
prog.cc:30:8: note: built-in candidate operator==(long double, int)
prog.cc:30:8: note: built-in candidate operator==(__float128, int)
prog.cc:30:8: note: built-in candidate operator==(int, float)
prog.cc:30:8: note: built-in candidate operator==(int, double)
prog.cc:30:8: note: built-in candidate operator==(int, long double)
prog.cc:30:8: note: built-in candidate operator==(int, __float128)
prog.cc:30:8: note: built-in candidate operator==(int, long)
prog.cc:30:8: note: built-in candidate operator==(int, long long)
prog.cc:30:8: note: built-in candidate operator==(int, __int128)
prog.cc:30:8: note: built-in candidate operator==(int, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(int, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(int, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(int, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(long, int)
prog.cc:30:8: note: built-in candidate operator==(long long, int)
prog.cc:30:8: note: built-in candidate operator==(__int128, int)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, int)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, int)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, int)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, int)
prog.cc:30:8: note: built-in candidate operator==(float, float)
prog.cc:30:8: note: built-in candidate operator==(float, double)
prog.cc:30:8: note: built-in candidate operator==(float, long double)
prog.cc:30:8: note: built-in candidate operator==(float, __float128)
prog.cc:30:8: note: built-in candidate operator==(float, long)
prog.cc:30:8: note: built-in candidate operator==(float, long long)
prog.cc:30:8: note: built-in candidate operator==(float, __int128)
prog.cc:30:8: note: built-in candidate operator==(float, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(float, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(float, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(float, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(double, float)
prog.cc:30:8: note: built-in candidate operator==(double, double)
prog.cc:30:8: note: built-in candidate operator==(double, long double)
prog.cc:30:8: note: built-in candidate operator==(double, __float128)
prog.cc:30:8: note: built-in candidate operator==(double, long)
prog.cc:30:8: note: built-in candidate operator==(double, long long)
prog.cc:30:8: note: built-in candidate operator==(double, __int128)
prog.cc:30:8: note: built-in candidate operator==(double, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(double, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(double, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(double, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(long double, float)
prog.cc:30:8: note: built-in candidate operator==(long double, double)
prog.cc:30:8: note: built-in candidate operator==(long double, long double)
prog.cc:30:8: note: built-in candidate operator==(long double, __float128)
prog.cc:30:8: note: built-in candidate operator==(long double, long)
prog.cc:30:8: note: built-in candidate operator==(long double, long long)
prog.cc:30:8: note: built-in candidate operator==(long double, __int128)
prog.cc:30:8: note: built-in candidate operator==(long double, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(long double, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(long double, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(long double, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(__float128, float)
prog.cc:30:8: note: built-in candidate operator==(__float128, double)
prog.cc:30:8: note: built-in candidate operator==(__float128, long double)
prog.cc:30:8: note: built-in candidate operator==(__float128, __float128)
prog.cc:30:8: note: built-in candidate operator==(__float128, long)
prog.cc:30:8: note: built-in candidate operator==(__float128, long long)
prog.cc:30:8: note: built-in candidate operator==(__float128, __int128)
prog.cc:30:8: note: built-in candidate operator==(__float128, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(__float128, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(__float128, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(__float128, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(long, float)
prog.cc:30:8: note: built-in candidate operator==(long, double)
prog.cc:30:8: note: built-in candidate operator==(long, long double)
prog.cc:30:8: note: built-in candidate operator==(long, __float128)
prog.cc:30:8: note: built-in candidate operator==(long, long)
prog.cc:30:8: note: built-in candidate operator==(long, long long)
prog.cc:30:8: note: built-in candidate operator==(long, __int128)
prog.cc:30:8: note: built-in candidate operator==(long, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(long, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(long, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(long, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(long long, float)
prog.cc:30:8: note: built-in candidate operator==(long long, double)
prog.cc:30:8: note: built-in candidate operator==(long long, long double)
prog.cc:30:8: note: built-in candidate operator==(long long, __float128)
prog.cc:30:8: note: built-in candidate operator==(long long, long)
prog.cc:30:8: note: built-in candidate operator==(long long, long long)
prog.cc:30:8: note: built-in candidate operator==(long long, __int128)
prog.cc:30:8: note: built-in candidate operator==(long long, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(long long, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(long long, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(long long, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(__int128, float)
prog.cc:30:8: note: built-in candidate operator==(__int128, double)
prog.cc:30:8: note: built-in candidate operator==(__int128, long double)
prog.cc:30:8: note: built-in candidate operator==(__int128, __float128)
prog.cc:30:8: note: built-in candidate operator==(__int128, long)
prog.cc:30:8: note: built-in candidate operator==(__int128, long long)
prog.cc:30:8: note: built-in candidate operator==(__int128, __int128)
prog.cc:30:8: note: built-in candidate operator==(__int128, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(__int128, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(__int128, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(__int128, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, float)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, double)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, long double)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, __float128)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, long)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned int, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, float)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, double)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, long double)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, __float128)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, float)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, double)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, long double)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, __float128)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned long long, unsigned __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, float)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, double)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, long double)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, __float128)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, long)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, __int128)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, unsigned int)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, unsigned long)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, unsigned long long)
prog.cc:30:8: note: built-in candidate operator==(unsigned __int128, unsigned __int128)
1 error generated.
陳 力
la source
-4

Une autre utilisation courante est que les conteneurs std effectuent une comparaison d'égalité sur les valeurs clés à l'intérieur des objets personnalisés

class Foo
{
    public: int val;
};

class Comparer { public:
bool operator () (Foo& a, Foo&b) const {
return a.val == b.val; 
};

class Blah
{
std::set< Foo, Comparer > _mySet;
};
user2945260
la source
2
Ceci est un exemple utilise operator ()pas operator bool. Ils sont totalement différents. operator ()est l'opérateur d'appel, de sorte que a Comparerpeut être appelé en tant que fonction. Cela operator ()arrive juste à revenir bool, mais cela ne le rend pas identique à operator bool, ce qui permet simplement un cast implicite vers bool.
Michael Dorst