Adding a procedure getting all keys in the hashmaps

[degawa]

Motivation

The hashmap in the stdlib does not provide a way to retrieve all keys in the map. As shown in the Prior Art section, a way to retrieve all keys in a hashmap is required in other programming languages, and it would significantly expands the use of the hashmap.

Prior Art

The Map in Java has the SetKey() method that returns the keys in this map.
The list() function for Python's dict returns a list of all the keys used in the dictionary.
The unordered_map in C++ does not have a similar method, but a way to get all the keys is required. (for example, see https://stackoverflow.com/questions/8483985/obtaining-list-of-keys-and-values-from-unordered-map)

Additional Information

The hashmaps in the stdlib, including open_hash_map_type and chaining_hash_map_type, store both the entered keys and values to the component inverse. So it is not challenging work to retrieve the keys as a key_type array.
The conversion from the key_type to the original type is left to the users.

[awvwgk]

In TOML Fortran the map types support returning a list of keys, which can be used to iterate through the data structure. In general using stdlib' s map type in TOML Fortran requires this functionality, and one way of implementing this would be to just store the keys in a list beside the map.

[jvdp1]

Thank you @degawa for the proposal.

The hashmaps in the stdlib, including open_hash_map_type and chaining_hash_map_type, store both the entered keys and values to the component inverse. So it is not challenging work to retrieve the keys as a key_type array.

It seems you already have an idea on how to implement such an option. Would you like to open a PR for its implemenation?