FlussDeserializer (flink-cdc-parent 3.8-SNAPSHOT API)

Type Parameters:

T - The type of output records produced by this deserializer.

All Superinterfaces:

Serializable

All Known Implementing Classes:

FlussRecordDeserializer
```
public interface FlussDeserializer<T>
extends Serializable
```
A generic deserializer interface for converting Fluss ScanRecords into output records of type T. This interface is intentionally not bound to any specific CDC type, allowing the Fluss source to be reused with different output types.

Method Summary

All Methods Instance Methods Abstract Methods Default Methods
Modifier and Type	Method	Description
`List<T>`	`deserialize(FlussSourceRecord element, org.apache.fluss.metadata.TablePath tablePath)`	Deserializes a Fluss `ScanRecord` into a list of output records.
`default void`	`removeState(org.apache.fluss.metadata.TablePath tablePath)`	Removes all state retained for an unsubscribed table.
`default List<T>`	`restoreState(org.apache.fluss.metadata.TablePath tablePath, int schemaId, org.apache.fluss.types.RowType rowType)`	Restores internal state (e.g., schema caches) from a recovered split.

- Method Detail
  - deserialize
```
List<T> deserialize(FlussSourceRecord element,
                    org.apache.fluss.metadata.TablePath tablePath)
```
    Deserializes a Fluss ScanRecord into a list of output records.
    
    Parameters:
    
    element - The Fluss scan record to deserialize.
    
    tablePath - The Fluss table path (database.table).
    
    Returns:
    
    A list of deserialized output records.
  - restoreState
```
default List<T> restoreState(org.apache.fluss.metadata.TablePath tablePath,
                             int schemaId,
                             org.apache.fluss.types.RowType rowType)
```
    Restores internal state (e.g., schema caches) from a recovered split. Called during split initialization to enable correct schema change detection after failover.
    The default implementation is a no-op. Override in subclasses that maintain schema caches.
    
    Parameters:
    
    tablePath - The Fluss table path.
    
    schemaId - The schema ID from the checkpointed split.
    
    rowType - The RowType corresponding to the schemaId.
  - removeState
```
default void removeState(org.apache.fluss.metadata.TablePath tablePath)
```
    Removes all state retained for an unsubscribed table.