Correctly handle duplicates
Opened this issue · 1 comments
Use owl:equivalentClass to make sure we respect synonyms when answering Brick queries.
When ingesting the schema, look out for these edges. Add duplicate entries in the hash/entity indices so that the entity strings point to the same hash key, which is used internally. Doing the duplicate pointers at this stage will avoid unnecessary bloat in the database (because we would have to add duplicate entries into the node index for each of these synonyms).
The trick will be keeping track of which entity names we rename/rewrite when we execute a query. A user is going to be wanting the original entities back; it wouldn't be nice if a user uses Air_Handler_Unit and gets back a node named AHU.
Probably targeting 0.5.1 for this