[CT-2666] Review Test Coverage of SemanticModels

[peterallenwebb]

During initial implementation of SemanticModels, we were moving quickly and did not add as much testing as is needed to fully verify functionality and prevent regressions. Let's use this issue to do that.

[jtcohen6]

This issue can be closed when:

• reviewed the set of capabilities (and tests) we've added since initial implementation
• created tickets for test cases that are still missing
• added to current sprint for execution, or put into backlog if lower-priority

[QMalcolm]

Semantic Model tests are a fair bit more sparse than Metric tests. We currently have 4 tests in tests/functional/semantic_models/. There are 2 additional which assert anything about semantic_models outside of tests/functional/semantic_models. Additionally, more testing needs to be done on the sub protocols of semantic models in test_semantic_layer_nodes_satisfy_protocols.py.

• Test SemanticModel.depends_on is properly populated
• Test partial parsing in regards to SemanticModels (we have a test_pp_semantic_models.py but it's empty)
• Additional protocol satisfication tests
  • SemanticModelDefaults
  • NodeRelation
  • MeasureAggregationParameters
  • DimensionTypeParams
  • DimensionValidityParams
• Add tests to test_contract_graph_parsed.py specifically for semantic models

[QMalcolm]

Added #8139, #8140, #8141, and #8142. They are being tracked as part of the Semantic Layer 1.7 Epic