Optimize performance for SPARQL backend

[LesterLyu]

Owlready2 sends queries as soon as something is modified:

organization = cids.Organization()  # This sends a query
organization.hasName = 'some name'   # This sends a query
organizatiion.address = ...    # This sends a query

This could slow down performance when using SPARQL backend.

Proposed optimizations

Option 1

• Combine consecutive INSERT DATA queries into one.
  • Debounce the insertions, i.e. 5ms, or send the query as soon as other types of queries are requested.
  • Involves multi-threading. A separate thread to schedule the query.

Option 2

• Change the logic of when queries are sent. (Can be turned off)
  • Introduce a save() method for individuals:

  default_world.set_autosave(false)  # Turn off real time modification
  organization = cids.Organization()
  organization.hasName = 'some name'
  organizatiion.address = ...  
  organizatiion.save()   # This sends some optimized queries.

Option 3

• Implement Option1 and Option 2