Drop table/view statements that involve more than one table generate one event for each table/view. Previously, each of those statements had the original multi-table/view statement. Now, each event has a statement that applies to only that table (generated from the original with all the same clauses).
The previous change did not correctly capture the statements for a `RENAME TO` that renamed multiple tables, so fixed the code so that it generates a single `RENAME TO` for each table rename.
Refactored the mechanism by which components can listen to the activities of a DDL parser. The new approach
should be significantly more flexible for additional types of DDL events while making it easier to maintain
backward compatibility. It also will enable passing event-specific information on each DDL event.
There is no point in sending a record that contains a null key and null value. While this may not be likely for insert or update cases (since at least the value should not be null), it is possible when a row is deleted (meaning the record value will be null) but the table has no primary/unique key (meaning the record key will be null).
Corrects a bug where a deleted row was written to Kafka in the same as an insert, making them indistinguishable. Now, a deleted row is written with the row's primary/unique key as the record key, and a null record value. Note that if the row has no primary/unique key, no record is written to Kafka.
Depending upon how the source MySQL database is configured, its binlog might not contain all the history for the database. In particular, it might not have the CREATE TABLE statements for some tables, which are then "unknown" to the connector. When the connector reads the binlog and comes across a change event for a row in one of those "unknown" tables, it previously resulted in a NPE. With this change, the condition results in a warning message in the log, and all subsequent change events on that table will be skipped.
Modified the 'docs' profile to build and attach JavaDoc JARs for each module's source and test source artifacts. The profile will be automatically used when releasing.
Changed the EmbeddedConnector framework to initialize all major components via configuration properties rather than through the public builder. This increases the size of the configurations, but it simplifies what embedding applications must do to obtain an EmbeddedConnector instance.
The DatabaseHistory framework was also changed to be configurable in similar ways to the OffsetBackingStore. Essentially, connectors that want to use it (like the MySqlConnector) will describe it as part of the connector's configuration, allowing more flexibility in which DatabaseHistory implementation is used and how it is configured whether in Kafka Connector or as part of the EmbeddedConnector.
Added a README.md to `debezium-embedded` to provide documentation and sample code showing how to use the EmbeddedConnector.
