Implement CDC data synchronization hudi based on Debezium Server

As of now, for debezium, we have a kafka based source. But we can try to leverage the debezium server directly and avoid additional maintanence costs. 

Ref:

 
# Does not rely on kafka connector, reducing maintenance costs，Flink cdc is more successful because it does not depend on kafka and is easy to use
# Support database or table level data synchronously written to hudi, support schema evolution
# Support flink or spark at the same time

This feature helps to expand the usage scenarios of hudi

[https://github.com/memiiso/debezium-server-spark]
[https://debezium.io/documentation/reference/operations/debezium-server.html#_implementation_of_a_new_sink]

## JIRA info

- Link: https://issues.apache.org/jira/browse/HUDI-5073
- Type: Improvement

Start with the Debezium Server documentation for implementing a new sink and the linked memiiso/debezium-server-spark reference; no Hudi files or tests are named. Scope the design around database- or table-level synchronization into Hudi, schema evolution, and support for Flink or Spark, with completion confirmed against those stated goals.

Written by the indexing model from the issue text.