Package org.apache.iceberg.flink.source
Class IcebergSource.Builder<T>
java.lang.Object
org.apache.iceberg.flink.source.IcebergSource.Builder<T>
- Enclosing class:
- IcebergSource<T>
-
Method Summary
Modifier and TypeMethodDescriptionasOfTimestamp
(Long newAsOfTimestamp) assignerFactory
(SplitAssignerFactory assignerFactory) build()
caseSensitive
(boolean newCaseSensitive) endSnapshotId
(Long newEndSnapshotId) exposeLocality
(boolean newExposeLocality) filters
(List<Expression> newFilters) flinkConfig
(org.apache.flink.configuration.ReadableConfig config) includeColumnStats
(boolean newIncludeColumnStats) maxAllowedPlanningFailures
(int maxAllowedPlanningFailures) monitorInterval
(Duration newMonitorInterval) nameMapping
(String newNameMapping) planParallelism
(int planParallelism) project
(org.apache.flink.table.api.TableSchema newProjectedFlinkSchema) properties
(Map<String, String> properties) Deprecated.readerFunction
(ReaderFunction<T> newReaderFunction) Set the read properties for Flink source.Set the read properties for Flink source.splitComparator
(SerializableComparator<IcebergSourceSplit> newSplitComparator) splitLookback
(Integer newSplitLookback) splitOpenFileCost
(Long newSplitOpenFileCost) startSnapshotId
(Long newStartSnapshotId) startSnapshotTimestamp
(Long newStartSnapshotTimestamp) streaming
(boolean streaming) streamingStartingStrategy
(StreamingStartingStrategy newStartingStrategy) tableLoader
(TableLoader loader) useSnapshotId
(Long newSnapshotId) watermarkColumn
(String columnName) Emits watermarks once per split based on the min value of column statistics from files metadata in the given split.watermarkColumnTimeUnit
(TimeUnit timeUnit) When the type of thewatermarkColumn(java.lang.String)
isTypes.LongType
, then sets theTimeUnit
to convert the value.
-
Method Details
-
tableLoader
-
table
-
assignerFactory
-
splitComparator
public IcebergSource.Builder<T> splitComparator(SerializableComparator<IcebergSourceSplit> newSplitComparator) -
readerFunction
-
flinkConfig
-
caseSensitive
-
useSnapshotId
-
streamingStartingStrategy
public IcebergSource.Builder<T> streamingStartingStrategy(StreamingStartingStrategy newStartingStrategy) -
startSnapshotTimestamp
-
startSnapshotId
-
tag
-
branch
-
startTag
-
endTag
-
endSnapshotId
-
asOfTimestamp
-
splitSize
-
splitLookback
-
splitOpenFileCost
-
streaming
-
monitorInterval
-
nameMapping
-
project
-
project
public IcebergSource.Builder<T> project(org.apache.flink.table.api.TableSchema newProjectedFlinkSchema) -
filters
-
limit
-
includeColumnStats
-
planParallelism
-
exposeLocality
-
maxAllowedPlanningFailures
-
set
Set the read properties for Flink source. View the supported properties inFlinkReadOptions
-
setAll
Set the read properties for Flink source. View the supported properties inFlinkReadOptions
-
watermarkColumn
Emits watermarks once per split based on the min value of column statistics from files metadata in the given split. The generated watermarks are also used for ordering the splits for read. Accepted column types are timestamp/timestamptz/long. For long columns consider settingwatermarkColumnTimeUnit(TimeUnit)
.Consider setting `read.split.open-file-cost` to prevent combining small files to a single split when the watermark is used for watermark alignment.
-
watermarkColumnTimeUnit
When the type of thewatermarkColumn(java.lang.String)
isTypes.LongType
, then sets theTimeUnit
to convert the value. The default value isTimeUnit.MICROSECONDS
. -
properties
Deprecated. -
build
-
setAll(java.util.Map<java.lang.String, java.lang.String>)
instead.