WIP Convert query to nodesets

[GianlucaFicarelli]

WIP

This is an initial experiment to convert a query to node_sets, to be able to use libsonata to get the ids.

Possible benefits:

• it seems faster than using multiple masks on a Pandas DataFrame, even when reading from hdf5 files
• it doesn't need to load in memory properties that are needed only for filtering

Limitations:

• some complex queries containing $and cannot be easily expressed using node_sets, unless converted to some longer and probably inefficient form.

Example:

{
    "$and": [
        {"node_id": [1, 2, 3]},
        {"$or": [{"mtype": "L6_Y"}, {"morphology": "morph-B"}]},
        {"$or": [{"region": "AAA"}, {"synapse_class": "EXC"}]},
    ]
}

[codecov-commenter]

Codecov Report

Merging #223 (3262bbd) into master (2da1c6d) will not change coverage.
The diff coverage is 100.00%.

@@            Coverage Diff            @@
##            master      #223   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           29        29           
  Lines         2163      2191   +28     
=========================================
+ Hits          2163      2191   +28     

Flag	Coverage Δ
pytest	100.00% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
bluepysnap/query.py	100.00% <100.00%> (ø)