[FLINK-38808][pipeline-connector][iceberg] Support partition transforms in Iceberg sink

[fcfangcc]

Purpose

This PR adds support for Iceberg partition transforms in the Iceberg sink. Previously, only identity partitioning was supported. With this change, users can utilize standard Iceberg transforms like year, month, day, hour, bucket, and truncate when defining partition keys.

• iceberg-doc

Changes

• IcebergDataSinkFactory: Updated to parse partition keys containing transforms from the sink.partition-key configuration.
• IcebergMetadataApplier: Implemented logic to apply partition transforms when creating new Iceberg tables.
• IcebergDataSinkOptions: Updated documentation to include examples of partition transforms.
• Tests: Added IcebergDataSinkFactoryTest cases to verify that partition specs are correctly generated for various transforms.

Usage

Users can now specify partition keys with transforms in the sink.partition-key option.
Example:

sink.partition-key = "testdb.table1:year(create_time);testdb.table2:bucket[16](id)"