add flink source and sink ci test

[F-PHantam]

main work in this pr :

1. On the basis of the original MySQL CDC, a basic table has been added, which is mostly full data type (except for JSON, time, and YAER that require special processing);
2. Flink source reads the LakeSoul cdc table and then writes it to the LakeSoul table. Then, the data written to the LakeSoul table by Flink is compared with MySQL data. The data in MySQL table will be added, deleted, and updated, and data comparison is performed every time;
3. Flink DataGen generates some data on its own and writes it to CSV and LakeSoul table, without primary keys. Then Spark reads the CSV data and compares the LakeSoul table data.