[FEA] Make TPC-H derived datagen reuse random column generation functions already present in benchmark utils

**Is your feature request related to a problem? Please describe.**
The NDS-H data generator within `libcudf` allows generating TPC-H derived tables for running internal benchmarks. It is located [here](https://github.com/rapidsai/cudf/tree/branch-24.10/cpp/benchmarks/common/tpch_data_generator). This data generator has a [module ](https://github.com/rapidsai/cudf/blob/branch-24.10/cpp/benchmarks/common/tpch_data_generator/random_column_generator.hpp)that facilitates generating different type of columns randomly. But some of the functions defined here are already implemented in an adjacent data generator for `libcudf` microbenchmarks [here](https://github.com/rapidsai/cudf/blob/branch-24.10/cpp/benchmarks/common/generate_input.hpp) leading to duplicate implementations. One such example is the [create_random_column ](https://github.com/rapidsai/cudf/blob/branch-24.10/cpp/benchmarks/common/generate_input.hpp#L668) function. We should be able to reuse this function within the NDS-H data generator.