Are C 14 Generic Lambdas Based on Templates or Type Erasure?

Unveiling the Mystery of Generic Lambdas in C 14

In C 14, the introduction of generic lambdas brought about a new dimension of flexibility in code. They utilize the power of auto as an argument type, but is their mechanism rooted in C templates or Java-like type erasure?

The Mechanics of Generic Lambdas

Generic lambdas leverage C templates to define a closure type that possesses a templated call operator. This differs from the non-templated call operator in C 11 lambdas. Consider the example:

auto glambda = [](auto a) { return a; };

Here, the closure type of glambda is defined as follows:

class /* unnamed */
{
public:
    template<typename T>
    T operator () (T a) const { return a; }
};

Comparison to C Templates and Java Generics

Generic lambdas share similarities with C templates, where the compiler generates distinct functions for different argument types. However, there's a crucial difference: generic lambdas don't create multiple function instances at compile time. Instead, they employ a single unnamed functor that defines a templated call operator.

In contrast to Java generics, which employ type erasure, generic lambdas preserve type information. The type template parameters in the call operator allow the lambda to deduce and manipulate the actual argument types at runtime.

Conclusion

Generic lambdas in C 14 are a powerful tool for writing concise and flexible code. While they draw inspiration from C templates, their mechanism is distinctly different, offering a unique blend of compile-time and runtime behavior.

The above is the detailed content of Are C 14 Generic Lambdas Based on Templates or Type Erasure?. For more information, please follow other related articles on the PHP Chinese website!