TRAPI response JSON bug release

• Removed 'provided_by' and 'members' properties from TRAPI response message Knowledge Graph nodes

Improve SemSimian Server result similarity match node annotation

• use NodeNormalizer to add name and category to SemSimian Server similarity match nodes
• some technical debt cleanup in documentation

Patch to fix 0.0.12; partial implementation of SemSimian Server support

• Constraints relaxed for the filtering of Query Node specification of set_interpretation such that incomplete nodes (which may be specified as a part of multistep MCQ) generate a warning but are ignored (they don't now raise an exception).
• minor release includes partial implementation (including SEMSIMIAN related environment variables, including a new SEMSIMIAN_MODE flag to tag two distinct layers of SemSimian access: server and (indirect) Monarch)

Proper detection of set_interpretation directive singularity for MCQ nodes

Input Query Graph list of nodes more robustly checked for constrain of having one and only one well-formed MCQ node with a set_interpretation directive

fix usage of TRAPI is_set

• removed 'is_set' expectation from QNode validation of MCQ inputs since 'set_interpretation' only is now used
• fixed unit tests broken from release 0.0.10
• also added new 'test.yml' workflow for CI validation of commits to the main branch

v0.0.10

• added "directionality": "object_to_subject" to SemSimian API call (as per recommendation of MMCQ team)

Application errors now reported to TRAPI Response logs list of LogEntry entries

Repaired MMCQ error reporting to report all validation state as TRAPI Response logs array of LogEntry items, returning empty TRAPI Response Message knowledge graph and results if there are errors (but not for just warnings, debug or info messages)

Additional Phenotype-to-Disease query term sets added to Jupyter Notebook

v0.0.8

Multi-Curie Queries Jupyter Notebook

• Added a Multi-Curie Queries Jupyter Notebook, with some basic Phenotype-to-Disease queries.
• Some tiny code fixes to make more robust. Note Python 3.11 constraint and some extra (optional) poetry dependencies to run the notebook!

Code differentiates "set_interpretation" values "MANY", versus "ALL"

• Algorithm now differentiates between 'MANY' and 'ALL' SemSimian matches filtering out duplicate matches by taking the highest scoring (phenotypic) query term to (disease-related phenotype) match pairs