Refactoring

[kravets-levko]

1. Get rid of additional wrapper for client and merge DBSQLClient and HiveClient classes
2. Rename some of other Hive** classes
3. Update docs and tests
4. Use import everywhere to get as much type hinting as possible

[susodapop]

LGTM.

How can we get the tests check to pass?

[kravets-levko]

@susodapop thanks!

How can we get the tests check to pass?

Currently no way for external PRs 😢 I still make sure they pass locally, though

[moderakh]

1. Get rid of additional wrapper for client and merge DBSQLClient and HiveClient classes
2. Rename some of other Hive** classes

@kravets-levko could you please explain the rationale of why this is needed? (I am not saying it is not needed, I just would like to have more context the rationale behind the proposed change :-) )

also is there a jira ticket for this? could you please mention the PECO jira ticket number for this work?

[kravets-levko]

@moderakh I created https://databricks.atlassian.net/browse/PECO-216 some time ago for this. There were few reasons for these changes:

1. at the very beginning we agreed that there are too much "hive" stuff here, so I took an opportunity to rename some classes. There's still a HiveDriver class, but I decided to keep it for now;
2. DBSQLClient was just a very thin wrapper to make HiveClient easier to use and ensure that all required parameters (e.g. PAT) are passed. Now, when we have an idea where it goes, we don't need that wrapper and could update HiveClient itself;
3. Using import instead of require is just for better TS support: for some reason, TS doesn't pick up types for modules imported with require 🤷‍♂️
4. DBSQLClient.close is now promisified just for consistency with other methods.