[FEA] Need Deletion Vector read support for Databricks 14.3

[razajafri]

Is your feature request related to a problem? Please describe.
Databricks 14.3 uses deletion vectors and currently, we don't support deletion vectors which leads to errors while reading a delta table

The following error can be reproduced by running TEST_MODE=DELTA_LAKE_ONLY WITH_DEFAULT_UPSTREAM_SHIM=0 TEST_PARALLEL=0 TESTS=delta_lake_auto_compact_test.py::test_auto_compact_min_num_files ./jenkins/databricks/test.sh

E                   Caused by: org.apache.spark.SparkException: Exception thrown in awaitResult: Part of the plan is not columnar class com.nvidia.spark.rapids.delta.RapidsDeltaWriteExec
E                   RapidsDeltaWrite
E                   +- Coalesce 1
E                      +- Project [id#3157L]
E                         +- Filter if (isnotnull(_databricks_internal_edge_computed_column_skip_row#3180)) (_databricks_internal_edge_computed_column_skip_row#3180 = false) else isnotnull(raise_error(DELTA_SKIP_ROW_COLUMN_NOT_FILLED, map(keys: [], values: []), NullType))
E                            +- ColumnarToRow
E                               +- FileScan parquet [id#3157L,_databricks_internal_edge_computed_column_skip_row#3180] Batched: true, DataFilters: [], Format: Parquet, Location: TahoeBatchFileIndex(1 paths)[file:/tmp/pyspark_tests/0905-174917-z50tzgqt-10-59-227-36-master-822..., PartitionFilters: [], PushedFilters: [], ReadSchema: struct<id:bigint,_databricks_internal_edge_computed_column_skip_row:boolean>
E

Describe the solution you'd like
We need to add support for Deletion Vectors but to do this it may make sense to break the functionality down into smaller parts and do the write and read

Describe alternatives you've considered
None

Additional context
#8654
#11541