[CH] Support CH backend with Spark 3.5.x

[zzcclp]

Description

Support CH backend with Spark3.5:

• [GLUTEN-6067][CH][Part 1] Support CH backend with Spark3.5 #6068
  • Upgrade Spark version to 3.5 and compile passed;
  • Upgrade Delta version to 3.2 and compile passed;
• Support Scala 2.13 #6311
• Support native write
  • [GLUTEN-6067][CH] [Part 2] Support CH backend with Spark3.5 - Prepare for supporting sink transform #6197
  • [GLUTEN-6067][VL] [Part 3-1] Refactor: Rename VeloxColumnarWriteFilesExec to ColumnarWriteFilesExec #6403
  • [GLUTEN-6067][CH] [Part 3-2] Basic support for Native Write in Spark 3.5 #6586
  • [CH] Fully Support writing parquet and mergetree in spark 3.5.x with delta protocol #7028
• CH backend UT passed: (now only the MergeTree + Delta UT passed);
  • [GLUTEN-6067][CH][Minor] Compile Spark-3.5 ut with backends-clickhouse #6114
  • [GLUTEN-6067][CH][MINOR][UT] Pass backends-clickhouse ut in Spark 3.5 #6623
    • [GLUTEN-6067][CH][MINOR][UT] Followup 6623, fix backends-clickhouse ut issse in CI #6891
    • Failed
      • test posexplode issue: https://github.com/oap-project/gluten/issues/ 1767 - failed
      • GlutenParquetFilterSuite q2 q9
    • Fallack
      • Support WindowGroupLimit , [CH] Support WindowGroupLimit for row_number, rank and dense_rank #7087 )
        • TPCDS 44, 67, 70
        • GLUTEN-1874 not null in one stream - fallback
        • GLUTEN-1874 not null in both streams - fallback
        • GLUTEN-4032: fix shuffle read coredump after union - fallback
      • GlutenClickHouseHiveTableSuite test 'hive udf' => fallback
    • GlutenClickhouseCountDistinctSuite check count distinct with expr fallback => why no fallback
    • GlutenClickHouseDecimalSuite TPCH Decimal(38,19) Q1 result is not matched if allowPrecisionLoss = true
• Spark 35 UT passed

[zzcclp]

there are some tasks need to be completed.

[zzcclp]

DON'T CLOSE THIS ISSUE, there are some tasks need to be completed.

[baibaichen]

Make GLUTEN-4316: fix crash on dynamic partition inserting work, need fix

1. [CH] transform with map_keys and map_values failed. #6561
2. [CH] Insert wrong result for struct field #6588

Make test table bucketed by all typed columns work, we need fix

1. [CH] ORC read wrong timestamp value #6506

[melin]

upgrade spark 3.5.2、3.4.3？

[baibaichen]

upgrade spark 3.5.2、3.4.3？

We are planing not to support 3.4.x, as for 3.5.2, let's support 3.5.1 first.