org.apache.spark.sql.execution.columnar.impl
Extension to standard enumeration value to hold the StructField.
Extension to standard enumeration value to hold the StructField. Also exposes the name as "name" field rather than having to use toString.
The initial size of delta column (the smallest delta in the hierarchy).
The maximum depth of the hierarchy of deltas for column starting with smallest delta, which is merged with larger delta, then larger, ...
The maximum depth of the hierarchy of deltas for column starting with smallest delta, which is merged with larger delta, then larger, ... till the full column value.
As of now only one-level of depth is used and deltas always merged immediately.
These enumeration values are the virtual columns that are injected in the select plan for update/delete so that those operations can actually apply the changes.
This is the currently used maximum depth which must be <= MAX_DEPTH.
This is the currently used maximum depth which must be <= MAX_DEPTH. It should only be used by transient execution-time structures and never in storage.
Returns the delta column index as store in region key given the 0 based table column index (table column index stored in region key is 1 based).
Returns 1 based table column index for given delta or table column index (table column index stored in region key is 1 based).