miguel-porto / flora-on-server Goto Github PK

This algorithm would need to check that

• all biological entities existing in a territory have a name
• no names overlap in terms of their biological concept
• hierarchy of names follows taxonomic rules

The algorithm should detect conflicting/incompatible names and ask the user to choose one, when multiple options exist.

This is important!

For example, "this taxon corresponds to the populations inhabiting coastal cliffs that have x and y traits..."

Instead of an "accepted" field, the data model should allow to have any number of checklists, and allow the users to track the changes from one checklist version to another.

YlemParser should output a list of possible disambiguations.

Describe the existence of the taxa in user-defined territorries, and their endemic status. Territories can be hierarchically described e.g. Iberia includes Pt and Hs.

e.g. if Linkagrostis is PART_OF Agrostis [sensu Franco], which is PART_OF Poaceae, then we don't need Linkagrostis to be PART_OF Poaceae. These superfluous links could be removed.

Long processing queries (e.g. output checklist) make the client close socket.

Territories and taxa must have a "complete" boolean field so that inference of OccurrenceStatus and NativeStatus can be precisely made when it is true. For example, if a territory is complete, the AbundanceLevel of a given taxon in the territory can be inferred from its children; if it is not complete, it can't be inferred.
Similar situation for taxon endemism.

Consider the idea of migrating to GraphQL.