feat: better parsing of glue tables; enable directory file structure

[rsyi]

Closes #86.

With this:

• Glue table names are parsed more effectively (rather than schema_table or catalog_schema_table, we write catalog.schema.table by parsing the provided glue name and returning the properly-formatted payload).
• Use the database/instance name as a separate directory, by using the formatting "connection_name/database/catalog.schema.table". If name isn't given in the connections.yaml file, you'll instead get: "database/catalog.schema.table". If exclusively using Glue, this reduces the complexity of the file structure nicely.

I don't believe anyone other than @rubenssoto is using this extractor (at the moment), but if you are, please open an issue and I'd be happy to provide a migration script! Breaking semantic versioning here, b/c this change is coming out so soon after the release of the feature.

[codecov-io]

Codecov Report

Merging #89 into master will increase coverage by 0.44%.
The diff coverage is 78.57%.

@@            Coverage Diff             @@
##           master      #89      +/-   ##
==========================================
+ Coverage   67.56%   68.01%   +0.44%     
==========================================
  Files          30       31       +1     
  Lines        1520     1588      +68     
==========================================
+ Hits         1027     1080      +53     
- Misses        493      508      +15     

Impacted Files	Coverage Δ
pipelines/whale/loader/whale_loader.py	58.11% <66.66%> (-0.15%)	⬇️
pipelines/whale/extractor/glue_extractor.py	78.78% <78.78%> (ø)
pipelines/whale/utils/extractor_wrappers.py	76.66% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0d618ab...961699b. Read the comment docs.