Support BMM Indexed Container meta-type, enabling Hash<K,V> structures

[wolandscat]

BmmParseRoundTrip tests pass.
This allows the properties of type Hash<K,V>, which are common in openEHR and other RMs to be correctly represented, rather than using the current generic workaround. This approach enables the type of such properties to be more naturally represented and interrogated, and the BMM files are more comprehensible. This also fills a hole in Archie's conformance to the published BMM and P_BMM specifications.
Note: only the BMM / PBMM test version of openehr_base_1.1.0 schema was changed; not the one used in the main part of Archie.

[codecov]

Codecov Report

Attention: Patch coverage is 60.37736% with 21 lines in your changes are missing coverage. Please review.

Project coverage is 71.73%. Comparing base (ca42c1a) to head (713ad18).
Report is 13 commits behind head on master.

Files	Patch %	Lines
...r/bmm/v2/persistence/PBmmIndexedContainerType.java	57.14%	7 Missing and 2 partials ⚠️
...m/v2/persistence/PBmmIndexedContainerProperty.java	53.33%	5 Missing and 2 partials ⚠️
.../openehr/bmm/core/BmmIndexedContainerProperty.java	50.00%	3 Missing ⚠️
...in/java/org/openehr/bmm/core/BmmContainerType.java	0.00%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##             master     #581      +/-   ##
============================================
- Coverage     71.75%   71.73%   -0.02%     
- Complexity     6943     6975      +32     
============================================
  Files           663      667       +4     
  Lines         22680    22779      +99     
  Branches       3672     3689      +17     
============================================
+ Hits          16273    16341      +68     
- Misses         4670     4692      +22     
- Partials       1737     1746       +9     

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

[wolandscat]

@MattijsK not sure how strictly you take the CodeCov reports - let me know if you need to get it over the 71% or whatever. It means adding tests that (theoretically) should have been there already. Also CodeCov can't detect (AFAIK) some usage of classes that is not strictly in JUnit tests - e.g. PBmmXx classes - they are exercised by BMM schema parsing and if the test on the related BmmXxx class works, it's pretty much a guarantee that the PBmmXxx classes are correct.