$cascade operation round 2, incl. recursion

[paynejd]

Following up on #1088, the next round of work on the $cascade operator will support recursion and hierarchy, and will improve alignment with FHIR.

• Add ability to cascade a concept's hierarchy in addition to mappings -- this will be enabled by default, and the user can control it with the cascadeHierarchy=true|false parameter
• Add option for the user to control whether mappings are cascaded -- this is enabled by default: cascadeMappings=true|false
• Add option to exclude mappings from the list of resources returned, eg includeMappings=false
• Add option to exclude map types, eg excludeMapTypes=SAME-AS -- note that this parameter should be ignored if mapTypes parameter is present
• Add ability to recursively walk thru a concept's hierarchy and mappings
  • User should have the option to limit the number of levels -- eg cascadeLevels=0 is the current behavior of no recursion; cascadeLevels=1 goes to the next level; cascadeLevels=all or * means do all levels and should be the default behavior
  • Recursion needs to prevent infinite loops, eg. if a concept has already been processed, then don't process it again
  • Recursion should apply the same mapType filters

[snyaggarwal]

@paynejd This is deployed on QA

[paynejd]

This query returns 211 resources:

• https://api.qa.openconceptlab.org/orgs/CIEL/sources/CIEL/concepts/159615/$cascade/?cascadeLevels=1

These two queries (just including/excluding the same map types) should return a total of 212 resources, right? (Because both are going to include the original concept) But the first one is 173 and the second one is 4, for a total of 177 resources.

• https://api.qa.openconceptlab.org/orgs/CIEL/sources/CIEL/concepts/159615/$cascade/?cascadeLevels=1&mapTypes=CONCEPT-SET,Q-AND-A
• https://api.qa.openconceptlab.org/orgs/CIEL/sources/CIEL/concepts/159615/$cascade/?cascadeLevels=1&excludeMapTypes=CONCEPT-SET,Q-AND-A

What is causing the mismatch?

[paynejd]

@snyaggarwal walked us thru an analysis of cascade recursion that showed that the results above were actually correct. See the image below-- we'll want to add a similar diagram to our documentation. Closing this out and further work will take place in #1131