reporting

The pipeline has been streamlined.

* Add stock adjustments to connectors

* Add XML template

* Add properties file to autoload from Nifi registry

* Add schema mapping file

* Add adjustment reason mapping file

* Add status changes tables to connectors

* Add XML template

* Add properties file to autoload from Nifi registry

* Add schema mapping file

* Technically RELEASED_NO_ORDER is not in the schema, but in the payload, but it is easier to include in the schema mapping file.

* Fix demo data to match new eLMIS demo data.

* Give Nifi access to OpenLMIS db, so that it can check a requisition exists before inserting line items. Also, redis is renamed to not conflict with redis running on Ref Distro.

* Transform process groups are added to Nifi registry, so add properties to autoload them into Nifi on startup.

* Modify the preload script to add controller services for Redis

* Add updated transform templates

This requires some additional logic in the requisition transform. Requisitions are saved in a cache with the V3 UUID mapping, in order to ensure coordination (requisitions saved before line items to avoid foreign key violations). This is saved after publishing to Kafka, in order for requisitions to go through the pipeline before line items. Each line item checks for existence of its parent requisition in the cache before continuing through the pipeline. Line items are also saved in the cache for their children (losses and adjustments).

Redis with persistence and a mapping volume has been added to docker compose for this cache.

Instead of building it each time at startup.

This was due to a change in the failure to load the client secrets and I changed the name from Check for Measures to Get Measures so we are able to use the autoload script.