Implement basic Bulk Data Access export server

[johngrimes]

This server will be a Spring Boot application, and will live within the server directory.

It will use Maven for dependency management and building the project, but it will not inherit the pom.xml from the root directory. Versioning of the server will be independent of the Pathling library. The version will be
1.0.0-SNAPSHOT until we are ready to release it.

The server will depend upon the libary-runtime (v8.0.0) module of the Pathling library, which will be included as a dependency in the pom.xml file. It will not depend upon any of the other modules in the Pathling library, or on any non-public APIs.

It will inherit some of the structure and functionality of the original Pathling server (v7).

It will use HAPI FHIR as the framework for building the FHIR server. This will create a servlet that can be mounted using the @WebServlet annotation in the same way as the original Pathling server.

Initially, the server will only support one operation - the Bulk Data Export operation.

In scope:

• System level export
• NDJSON output format
• _type parameter
• Bulk data status request, output manifest (required fields only)
• Capability statement requirements

Out of scope:

• Authorisation
• _since and _until parameter
• _elements parameter
• _includeAssociatedData parameter
• _typeFilter parameter
• organizeOutputBy parameter
• allowPartialManifests parameter
• Binary to DocumentReference conversion
• Group-level export
• Bulk Cohort API
• Deletion bundles
• _list parameter

[johngrimes]

Let's aim for support of _since, _until and _elements.

Also, we will need to add authorisation, see: https://pathling.csiro.au/docs/7.2.0/server/authorization

We will also need to look at caching.

[johngrimes]

We should also implement the Bulk Data Delete Request.

[fhnaumann]

There are some open questions regarding the caching + bulk delete request behavior.
The current implementation has some problems, i.e. deleting something that was cached and then requesting that again still returns the data.