Package org.apache.flink.table.gateway.service.result

Class ResultFetcher

java.lang.Object

org.apache.flink.table.gateway.service.result.ResultFetcher

public class ResultFetcher
extends Object

A fetcher to fetch result from submitted statement.

The fetcher uses the Iterator model. It means every time fetch the result with the current token, the fetcher will move forward and retire the old data.

After closes, the fetcher will not fetch the results from the remote but is able to return all data in the local cache.

Method Summary

Modifier and Type	Method	Description
void	close()
org.apache.flink.table.gateway.api.results.ResultSet	fetchResults(long token, int maxFetchSize)	Fetch results from the result store.
org.apache.flink.table.gateway.api.results.ResultSet	fetchResults(org.apache.flink.table.gateway.api.results.FetchOrientation orientation, int maxFetchSize)
static ResultFetcher	fromResults(org.apache.flink.table.gateway.api.operation.OperationHandle operationHandle, org.apache.flink.table.catalog.ResolvedSchema resultSchema, List<org.apache.flink.table.data.RowData> results)
static ResultFetcher	fromResults(org.apache.flink.table.gateway.api.operation.OperationHandle operationHandle, org.apache.flink.table.catalog.ResolvedSchema resultSchema, List<org.apache.flink.table.data.RowData> results, org.apache.flink.api.common.JobID jobID, org.apache.flink.table.api.ResultKind resultKind)
static ResultFetcher	fromTableResult(org.apache.flink.table.gateway.api.operation.OperationHandle operationHandle, org.apache.flink.table.api.internal.TableResultInternal tableResult, boolean isQueryResult)
org.apache.flink.table.catalog.ResolvedSchema	getResultSchema()
ResultStore	getResultStore()
ResultFetcher	withResourceManager(org.apache.flink.table.resource.ResourceManager resourceManager)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

fromTableResult

public static ResultFetcher fromTableResult(org.apache.flink.table.gateway.api.operation.OperationHandle operationHandle,
                                            org.apache.flink.table.api.internal.TableResultInternal tableResult,
                                            boolean isQueryResult)

fromResults

public static ResultFetcher fromResults(org.apache.flink.table.gateway.api.operation.OperationHandle operationHandle,
                                        org.apache.flink.table.catalog.ResolvedSchema resultSchema,
                                        List<org.apache.flink.table.data.RowData> results)

fromResults

public static ResultFetcher fromResults(org.apache.flink.table.gateway.api.operation.OperationHandle operationHandle,
                                        org.apache.flink.table.catalog.ResolvedSchema resultSchema,
                                        List<org.apache.flink.table.data.RowData> results,
                                        @Nullable
                                        org.apache.flink.api.common.JobID jobID,
                                        org.apache.flink.table.api.ResultKind resultKind)

withResourceManager

public ResultFetcher withResourceManager(org.apache.flink.table.resource.ResourceManager resourceManager)

close

public void close()

getResultSchema

public org.apache.flink.table.catalog.ResolvedSchema getResultSchema()

fetchResults

public org.apache.flink.table.gateway.api.results.ResultSet fetchResults(org.apache.flink.table.gateway.api.results.FetchOrientation orientation,
                                                                         int maxFetchSize)

fetchResults

public org.apache.flink.table.gateway.api.results.ResultSet fetchResults(long token,
                                                                         int maxFetchSize)

Fetch results from the result store. It tries to return the data cached in the buffer first. If the buffer is empty, then fetch results from the ResultStore. It's possible multiple threads try to fetch results in parallel. To keep the data integration, use the synchronized to allow only one thread can fetch the result at any time.

getResultStore

@VisibleForTesting
public ResultStore getResultStore()