Spline support for expand operation

## Background

The Spline agent does not support column-level lineage tracking in expand operations. So, if my sql contains "grouping", or "with cube", the column lineage will be lost after expand.
## Feature

Add OperationNodeBuilder to deal with spark expand operation.

## Example [Optional]


## Proposed Solution [Optional]

Solution Ideas

1. I add a ExpandNodeBuilder class in package za.co.absa.spline.harvester.builder.plan, which has almost the same code as UnionNodeBuilder
2. I modified the class OperationNodeBuilderFactory, to deal with expand operators.