Open
Description
Most of the querying from the language server client happens solely in terms of a source code location (e.g. Position, or Range). Currently, we're storing source-level spans, along with relevant information in a vector, which is obviously not ideal. Usage of a good data structure should provide better performance and scalability