fix(cubesql): Calculate proper limit and offset for CubeScan in nested limits case #3633

Job	Run time
latest-tag-sha	20s
Build native Linux 20 x86_64-unknown-linux-gnu Python fallback	2m 46s
build	6m 48s
tests (20.x, athena-export-bucket-s3)	18m 30s
tests (20.x, bigquery-export-bucket-gcs)	8m 30s
tests (20.x, clickhouse)	5m 30s
tests (20.x, clickhouse-export-bucket-s3)	6m 20s
tests (20.x, databricks-jdbc)	6m 7s
tests (20.x, databricks-jdbc-export-bucket-s3)	7m 3s
tests (20.x, databricks-jdbc-export-bucket-azure)	6m 18s
tests (20.x, mssql)	5m 46s
tests (20.x, mysql)	5m 24s
tests (20.x, postgres)	5m 27s
tests (20.x, redshift)	6m 25s
tests (20.x, redshift-export-bucket-s3)	5m 58s
tests (20.x, snowflake)	5m 40s
tests (20.x, snowflake-export-bucket-s3)	5m 52s
tests (20.x, snowflake-export-bucket-azure)	5m 38s
tests (20.x, snowflake-export-bucket-azure-via-storage-integration)	5m 58s
tests (20.x, snowflake-export-bucket-gcs)	5m 39s
	2h 5m 59s

Provide feedback