Fix insertion macro for Databricks Unity Catalog

[TannerHopkins]

This PR fixes an issue for Databricks projects with Unity Catalog enabled.
Currently, the insert_into_metadata_table macro only specifies the schema_name.table_name, but with catalogs (databases) introduced in Unity Catalog, the database_name is also required or the insertion fails.
This should be backwards compatible with non-UC implementations of databricks since they can now also be fully qualified with a database_name of hive_metastore - assuming this value gets populated automatically, of course.

If there are any issues with non-UC implementations of databricks, or spark since it also shares this macro, an alternate solution would be to first check that the database_name has a value, and if so use it, otherwise just keep schema_name.table_name

[TannerHopkins]

Looks like this first attempt failed (UC not enabled on this cluster error).
Going to try it with an if statement - looks like this is actually what @NiallRees originally did to make dbt_artifacts compatible with unity catalog.
I think this macro was just missed in that original PR

[VDFaller]

Is this in version 2.2.2? If not, would it be the reason I'm getting

23:32:37  Database error while running on-run-end
23:32:37  
23:32:37  Finished running 2 table models in 0 hours 0 minutes and 36.15 seconds (36.15s).
23:32:37  
23:32:37  Completed with 1 error and 0 warnings:
23:32:37
23:32:37  on-run-end failed, error:
23:32:37   Table not found: dev_vince.dbt_artifacts.model_executions; line 4 pos 16

My relevant dbt_project.yml looks like

models:
  dbt_artifacts:
    +database: "{{target.catalog}}"
    +schema: dbt_artifacts

and my relevant profile looks like

databricks:
  target: dev
  outputs:
    dev:
      catalog: dev_vince
      schema: literally_anything_else
      type: databricks

[jaypeedevlin]

@VDFaller Have you done a full dbt run or dbt run -s dbt_artifacts? If not the tables wouldn't have been created for the on-run-end to insert the data into...

[VDFaller]

Nope. much better. Thank you much.