Faster/better addition of dataset entries

[bennybp]

Description

Improve the efficiency of adding large numbers of entries to a dataset

• Fix expensive SQL queries. No need to retrieve all existing entries when adding a single entry. Use batching for large numbers
• Improve usability of add_entries functions in dataset models

For the first, local tests of adding 10,000 entries to a singlepoint dataset one-by-one show a reduction of 75%, with consistent (rather than steadily increasing) insertion time.

See openmm/spice-dataset#82 (comment) for the initial motivation

Changelog description

Improve the efficiency of adding large numbers of entries to a dataset

Status

• Code base linted
• Ready to go

[codecov]

Codecov Report

Merging #769 (6085e73) into main (d1252ac) will decrease coverage by 0.02%.
The diff coverage is 45.00%.

Additional details and impacted files