Concept Map

→ data structure to handle sets of language concept objects

• maps language concept type IDs to lists of language concept objects of that type
  • optionally, additional information about where in the AST certain language concept objects were extracted can be used as key as well
• used by traversers, processors, and post processors to pass around language concept data
• central data structure contained within a project object

Implementation

Two possible structures:

• Map<String, Concept[]>: maps language concept type ID to a list containing all language concepts of that type extracted so far
  • recommended, as it is easier to handle than the version below
• Map<String, Map<String, Concept[]>>: maps the identifier of the child AST node under which a number of language concept were extracted to an instance of the data structure described above
  • this is useful when trying to discern between multiple extracted concepts of the same type, but in different contexts
  • e.g. class MyClass extends BaseType implements SomeOtherType, may lead to two type reference concept objects being extracted. To compose them (along with other concepts) to a class declaration they need to be differentiated via the first map key.
  • this structure isn’t strictly needed and can be avoided by using local contexts for providing a partially initialized parent concept object to processors further down the tree