Implement get_metadata_field_names() and get_tensor_names() in Fortran

[billschereriii]

Description

DataSet objects store multiple metadata fields that are accessible by a name parameter. It would be convenient to be able to get a list of the names of metadata fields in Fortran, corresponding to the following C++ function:

• std::vector<std::string> get_metadata_field_names()

Similarly, DataSet objects store multiple tensors that are accessible by a name parameter. It would be convenient to be able to get a list of the names of tensors in Fortran, corresponding to the following C++ function:

• std::vector<std::string> get_tensor_names()

Justification

This will improve the ability of users and developers to inspect and compare the contents of a DataSet

Implementation Strategy

Particular care will need to be made to be made to the array processing to ensure that no memory is leaked. An example of returning a single string may be found in the get_last_error() method recently implemented; an extension to that functionality should be able to accomodate an array of strings.

Acceptance Criteria

• Implement get_metadata_field_names() and get_tensor_names() in Fortran
• Update unit and integral test to validate it