chore(sync native): Attempt to get the DB connection UUIDs via API

[Vitor-Avila]

During the sync native command, the CLI has to identify all existing DB connections in the target instance/Workspace (and their corresponding UUIDs). This is currently achieved by exporting each DB connection, and getting the uuid from the YAML file.

This process was adopted because the uuid is not exposed in the API response for older Superset versions. However, it affects performance for two reasons:

• Each DB connection is exported (the more connections created, the longer it would take).
• There was a change in Superset that exporting a DB connection also includes all datasets, which can take time for connections with hundreds/thousands of datasets.

This PR defaults the process to get the uuid from the API response, and fallback in the old approach in case the API response doesn't have the uuid field.

[betodealmeida]

Ohhh, nice!