Support fishing all resource types from dependency packages and input/* #1548
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
If the FHIRDefinitions fish functions are called specifying Type.Instance as a requested type, treat it like no types were passed in at all (i.e. wildcard type search) in order to search all types.
…cal assignments Now that SUSHI loads additional resource types from dependencies, there is a greater chance of naming collisions when resolving names. In one case, this caused a canoonical assignment that used to work to stop working (i.e. it used to resolve to a PlanDefinition from the tank, which was good; but then it started resolving to a Library from predefined resources, which didn't match the needed Canonical type). Now we can search for all metadata results from a given search term and, if necessary, choose the one that is a best match during canonical assignment. This may be relevant to other parts of the code as well, but for now we only do it for Canonical assignment in instances.
mint-thompson
previously approved these changes
Dec 26, 2024
jafeltra
previously approved these changes
Dec 26, 2024
mint-thompson
approved these changes
Dec 30, 2024
This was referenced Jan 6, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description: In the past, SUSHI only loaded specific resource types from dependency packages and predefined resources in input/* (i.e., StructureDefinitions, ValueSets, and CodeSystems). The recent integration of FPL 2.0.0 now loads all resource types, but SUSHI still did not return them in its results since it only fished (i.e. searched) for the previously supported types. This PR adds support for fishing those types by interpreting a fishing call for the
Type.Instancetype as a wildcard type (since a FSHInstancecan be literally any type).This also adds a new
fishForMetadatasfunction that will return all the matchingMetadatas instead of just the first one. This allows for smarter resolution ofCanonical(...)keywords to ensure that a matched type is resolved when one is available. This change, however, required that I add support forfishForMetadatasin all theFishableimplementations (not just FHIRDefinitions, but also Tank, Package, some exporters, etc.).Backstory for
fishForMetadataschange: Regression failed when I added support for fishing on all resource types because it introduced new types that could now come up in search results before the types it returned in previous versions. In one specific case, an IG was assigning CarePlan.instantiatesCanonical toCanonical(SomeGuideline). In previous versions of SUSHI,Canonical(SomeGuideline)resolved to aPlanDefinitiondefined ininput/fsh-- but with support for finding additional resource types in packages and predefined inputs, it was now resolving to aLibraryininput/resources. This was correct behavior because SUSHI prioritizes predefined resource above FSH definitions (to allow for "patched" resources to override FSH resources) -- but it caused errors becauseCarePlan.instantiatesCanonicalallowed forcanonical<PlanDefinition>but notcanonical<Library>, so it threw a mismatched type exception (Canonical(Library) is not one of the allowed types...). This seemed like an unfortunate error since there was a valid type in scope (thePlanDefinitiondefined in FSH), so I updated SUSHI to look through all the matches if the first match does not satisfy the canonical requirements. Now the regression passes with no changes.Testing Instructions: I've created a simple project that tests both aspects of this: (1) That additional resource types are now loaded (in this case, from predefined resource), and (2) that SUSHI intelligently chooses the right resource based on required Canonical type. You can try running the project using SUSHI 3.13.1, and then this PR branch. The project: AllTypesTest.zip
Related Issues: