Why is std::move named std::move despite not actually moving data?

Why is std::move named std::move?

The std::move(x) function does not actually move anything in terms of physical data. It is essentially a cast to an rvalue, specifically an xvalue. This casting can lead to confusion, so it's important to understand the intended purpose behind the naming convention.

The original concept of move dates back to the initial move proposal in 2002. This proposal introduced rvalue references and showed how to write a more efficient std::swap. At the time, the only meaning associated with "&&" was logical and, so a more readable method was needed to convey the intent of the function.

Instead of using "cast_to_rvalue" or a similar name, the developers chose "move" to highlight the purpose of the cast: to enable move semantics. This naming convention helps clarify the intent of the code without getting into technical details about value categories.

Today, the journeyman C programmer should be aware that std::move is essentially a cast, while beginner programmers can understand its purpose is to move from a source rather than copy from it. Other developers can implement similar functionality under a different name if desired, demonstrating that std::move does not possess a monopoly on this feature.

In terms of generated object code, std::move has no impact on trivially movable objects. However, for objects with custom move assignment operators, std::move can influence which overload is called, potentially resulting in a performance improvement.

Overall, the naming of std::move as "move" serves a pedagogical purpose, emphasizing the intent of the function rather than its technical implementation.

The above is the detailed content of Why is std::move named std::move despite not actually moving data?. For more information, please follow other related articles on the PHP Chinese website!