Cleanup of the postsynaptic trace transformer used in NEST code generation

[clinssen]

This is preliminary work for the corresponding presynaptic trace transformer that will be added for the SpiNNaker code generation target.

Change lists to sets and de-duplicate methods from neuron-synapse co-generation-related transformers into a separate utils class.

Additionally, refactor the way that extra attributes and information about the model is stored in an extra "metadata" data structure, which, for every model, maps named attributes to values. (This is currently done by setting extra attributes on the neuron and synapse ASTModel class instances, which is very dirty.) The idea is that each transformer accepts metadata and can add metadata for each model using its own set of unique key/value mappings.

[github-actions]

Bencher Report

Branch	1318/merge
Testbed	ubuntu-latest

🚨 1 Alert

Iteration	Benchmark	Measure Units	View	Benchmark Result (Result Δ%)	Upper Boundary (Limit %)
0	tests/nest_continuous_benchmarking/test_nest_continuous_benchmarking.py::TestNESTContinuousBenchmarking::test_stdp_nn_synapse	Latency seconds (s)	📈 plot 🚷 threshold 🚨 alert (🔔)	4.06 s (+20.13%) Baseline: 3.38 s	3.72 s (109.21%)

Click to view all benchmark results

Benchmark	Latency	Benchmark Result seconds (s) (Result Δ%)	Upper Boundary seconds (s) (Limit %)
tests/nest_continuous_benchmarking/test_nest_continuous_benchmarking.py::TestNESTContinuousBenchmarking::test_stdp_nn_synapse	📈 view plot 🚷 view threshold 🚨 view alert (🔔)	4.06 s (+20.13%) Baseline: 3.38 s	3.72 s (109.21%)

Benchmark	Latency	Benchmark Result seconds (s) (Result Δ%)	Upper Boundary seconds (s) (Limit %)
tests/nest_continuous_benchmarking/test_nest_continuous_benchmarking.py::TestNESTContinuousBenchmarking::test_stdp_nn_synapse	📈 view plot 🚷 view threshold 🚨 view alert (🔔)	3.27 s (-2.90%) Baseline: 3.37 s	3.71 s (88.27%)

Benchmark	Latency	Benchmark Result seconds (s) (Result Δ%)	Upper Boundary seconds (s) (Limit %)
tests/nest_continuous_benchmarking/test_nest_continuous_benchmarking.py::TestNESTContinuousBenchmarking::test_stdp_nn_synapse	📈 view plot 🚷 view threshold 🚨 view alert (🔔)	3.27 s (-2.89%) Baseline: 3.37 s	3.71 s (88.28%)

🐰 View full continuous benchmarking report in Bencher

[clinssen]

Or whould these methods better go in NESTCodeGeneratorUtils?

[pnbabu]

I think we can keep the post port-based methods here and move the vt port methods to NESTCodeGeneratorUtils.