[lucic71]

The manual Init marker on rules such as vector::emplace_back was used to encode semantic information about the rule: the forward pack is used to construct an element type.

For automatically generated rules this is not a good approach. The signature does not tell how forwarded arguments are used, this information exists as text in the description of the standard, for example:

std::map<Key,T,Compare,Allocator>::emplace

template< class... Args > std::pair<iterator, bool> emplace( Args&&... args )

Inserts a new element into the container constructed in-place with the given args, if there is no element with the key in the container. The constructor of the new element (i.e. std::pair<const Key, T>) is called with exactly the same arguments as supplied to emplace, forwarded via std::forward(args)....

---

This PR moves the Init marker in the Rust translation:

fn f67<T1: CppLess + Clone + ByteRepr + 'static, T2: Default + Clone + ByteRepr + 'static>(
    a0: Ptr<Map<T1, (Value<T1>, Value<T2>)>>,
    // init is the pack argument which builds typename std::map<T1, T2, T3, T4>::value_type
    // as the text of the standard says
    init: cpp_type!("typename std::map<T1, T2, T3, T4>::value_type", (Value<T1>, Value<T2>)),
) -> (Value<BidirectionalIterator<(Value<T1>, Value<T2>)>>, Value<bool>)

cpp_type has no semantic meaning for the rule, it expands to the second argument, i.e. (Value<T1>, Value<T2>).

---

I also added rules for std::any which builds a decay_t<T> using the forwarded arguments. I wanted to make sure that this also works.

[nunoplopes]

(conflict)