Use more effective record count when JDBC source is used with an SQL query

[yruslan]

Background

This idea is reported by @filiphornak

Currently the record count is calculated this way if SQL expression (rather than table name) is used as an input to the JDBC source: https://github.com/AbsaOSS/pramen/blob/0f8040a8bad151eccf5a6ee3403b2ae9c6a24b9e/pramen/core/src/main/scala/za/co/absa/pramen/core/reader/TableReaderJdbc.scala#L129-L129

This is not always effective since Spark does not always can get the record count without fetching all records.

Feature

Use more effective record count when JDBC source is used with an SQL query.

Proposed Solution

SELECT COUNT(*) AS CNT FROM (${query})