Tumbling window aggregation :)

[turbolytics]

Getting back up to speed after 6 months.

https://duckdb.org/docs/connect/concurrency.html

DuckDB uses optimistic concurrency control, an approach generally considered to be the best fit for read-intensive analytical database systems as it speeds up read query processing. As a result any transactions that modify the same rows at the same time will cause a transaction conflict error:

• Initialize the tumbling window manager.
• Setup the tumbling window manager to poll on a loop for closed windows.
• Update sqlflow to remove conf dep and inject pipeline directly
• Add more structure around config - every yaml property should be converted to a typed config object
• Add invoke test cases. For windowed it should process message batch then flush the window.
• Add Support for configurable out (including kafka)
• Add tests to ensure windowing close logic - 0 flushes, partial flush and total flushes.
• Add benchmark
• Update config tables.sql.window -> tables.sql.manager and update window package -> handlers .Tumbling window

[turbolytics]

closes #27