Support for multiple FHIR profile extensions

[chandrashekar-s]

Description of what I changed

Fixes #980

This is an extension to the PR #924. In this PR we provide support for multiple FHIR profile extension for a given resource type. This is achieved by creating a Master AvroConverter for the given resource type by merging all the individual AvroConverters against the configured extensions profiles including the base profile. The merging includes a deep merge of the nested elements under the AvroConverter.

E2E test

Tested the changes with the default US Core FHIR profile extensions and verified that the extension fields are properly populated in the parquet files.

TESTED:

Please replace this with a description of how you tested your PR beyond the
automated e2e/unit tests.

Checklist: I completed these to help reviewers :)

• I have read and will follow the review process.

• I am familiar with Google Style Guides for the language I have coded in.

  No? Please take some time and review Java and Python style guides.

• My IDE is configured to follow the Google code styles.

  No? Unsure? -> configure your IDE.

• I have added tests to cover my changes. (If you refactored existing code that was well tested you do not have to add tests)

• I ran mvn clean package right before creating this pull request and added all formatting changes to my commit.

• All new and existing tests passed.

• My pull request is based on the latest changes of the master branch.

  No? Unsure? -> execute command git pull --rebase upstream master

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 59.20578% with 113 lines in your changes are missing coverage. Please review.

Project coverage is 50.52%. Comparing base (b354beb) to head (178366b).

Files	Patch %	Lines
...r/bunsen/avro/tools/GenerateAggregatedSchemas.java	0.00%	55 Missing ⚠️
...unsen/avro/converters/DefinitionToAvroVisitor.java	88.18%	9 Missing and 4 partials ⚠️
.../cerner/bunsen/definitions/PrimitiveConverter.java	38.46%	6 Missing and 2 partials ⚠️
...m/cerner/bunsen/definitions/HapiConverterUtil.java	12.50%	5 Missing and 2 partials ⚠️
...ner/bunsen/definitions/LeafExtensionConverter.java	40.00%	3 Missing and 3 partials ⚠️
.../com/cerner/bunsen/avro/tools/GenerateSchemas.java	44.44%	4 Missing and 1 partial ⚠️
.../com/google/fhir/analytics/AvroConversionUtil.java	44.44%	4 Missing and 1 partial ⚠️
...java/com/cerner/bunsen/ProfileMappingProvider.java	66.66%	3 Missing ⚠️
...m/cerner/bunsen/avro/converters/NoOpConverter.java	0.00%	2 Missing ⚠️
...a/com/cerner/bunsen/ProfileMapperFhirContexts.java	50.00%	2 Missing ⚠️
... and 6 more

Additional details and impacted files

@@             Coverage Diff              @@
##             master     #986      +/-   ##
============================================
+ Coverage     50.36%   50.52%   +0.16%     
- Complexity      644      663      +19     
============================================
  Files            87       91       +4     
  Lines          5266     5465     +199     
  Branches        669      701      +32     
============================================
+ Hits           2652     2761     +109     
- Misses         2366     2444      +78     
- Partials        248      260      +12     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[chandrashekar-s]

@bashir2 The PR is ready for the first round of review, can you please have a look and let me know if the approach taken is good. In the meanwhile, I will add unit tests.

[bashir2]

Thanks @chandrashekar-s for the changes. I think this is a nice solution for the multiple extension issue and I like the merge() idea. I am sharing my first batch of comments. I have not yet finished my first pass and may share another batch later today, if I can (otherwise tomorrow).

[chandrashekar-s]

Thanks for the review @bashir2. In the latest commits, I have added more unit tests and have addressed your comments.

[chandrashekar-s]

@bashir2 Please have a look at the latest commits containing more unit tests and some minor fixes.

[bashir2]

Thanks @chandrashekar-s for the changes. Most of my comments are minor and I think we can close this PR this week. Beside the comments below, there is also one unresolved comment about ParquetMerger which I cannot reply to anymore (I guess because you have force pushed over an old commit); it is about the experiments you did with compatible but different schemas (e.g., when a new extension is added). Can you please document your finding somewhere? Even if it is in a javadoc that is fine, but it would be nice to have a record of that.

[chandrashekar-s]

Thanks @bashir2 for the detailed review. I have addressed most of the comments, there are only 2 comments left which we can discuss in our tomorrow's meeting.

[bashir2]

Thanks @chandrashekar-s for the changes; all remaining comments are minor and/or we discussed them. So please feel free to merge this once they are resolved.

[chandrashekar-s]

Thanks @bashir2 for the review. I have addressed the comments in the latest commits. I will be merging the changes. Please let me know if there are any concerns.