[
https://issues.apache.org/jira/browse/FLINK-31312?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17696083#comment-17696083
]
xzw0223 commented on FLINK-31312:
---------------------------------
[~Jiang Xin] Because enableObjectReuse is enabled, the row generated by source
is the same as the row object in map. Because objects are reused, the row
constructed by source adds field data by location. Therefore, the specified
field data needs to be retrieved by location in map,if enableObjectReuse is not
enabled, a row object will be re-copied every time you emit. At this time, the
row object will construct its fields according to RowTypeInfo, so there is no
problem.
> EnableObjectReuse cause different behaviors
> -------------------------------------------
>
> Key: FLINK-31312
> URL: https://issues.apache.org/jira/browse/FLINK-31312
> Project: Flink
> Issue Type: Bug
> Components: API / DataStream
> Reporter: Jiang Xin
> Priority: Major
>
> I have the following test code which fails with the exception `Accessing a
> field by name is not supported in position-based field mode`, however, if I
> remove the `enableObjectReuse`, it works.
> {code:java}
> public static void main(String[] args) throws Exception {
> StreamExecutionEnvironment env =
> StreamExecutionEnvironment.getExecutionEnvironment();
> env.setParallelism(1);
> // The test fails with enableObjectReuse
> env.getConfig().enableObjectReuse();
> final SourceFunction<Row> rowGenerator =
> new SourceFunction<Row>() {
> @Override
> public final void run(SourceContext<Row> ctx) throws
> Exception {
> Row row = new Row(1);
> row.setField(0, "a");
> ctx.collect(row);
> }
> @Override
> public void cancel() {}
> };
> final RowTypeInfo typeInfo =
> new RowTypeInfo(new TypeInformation[] {Types.STRING}, new
> String[] {"col1"});
> DataStream<Row> dataStream = env.addSource(rowGenerator, typeInfo);
> DataStream<Row> transformedDataStream =
> dataStream.map(
> (MapFunction<Row, Row>) value ->
> Row.of(value.getField("col1")), typeInfo);
> transformedDataStream.addSink(new PrintSinkFunction<>());
> env.execute("Mini Test");
> } {code}
> The `SourceFunction` generates rows without field names, but the return type
> info is assigned by `env.addSource(rowGenerator, typeInfo)`.
> With object-reuse enabled, rows would be passed to the MapFunction directly,
> so the exception raises. While if the object-reuse is disabled, rows would
> be reconstructed and given field names when passing to the next operator and
> the test works well.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
