Github user twalthr commented on a diff in the pull request:
https://github.com/apache/flink/pull/6323#discussion_r202506574
--- Diff:
flink-libraries/flink-table/src/main/scala/org/apache/flink/table/catalog/ExternalTableSourceUtil.scala
---
@@ -44,32 +43,27 @@ object ExternalTableSourceUtil extends Logging {
val properties = new DescriptorProperties()
externalCatalogTable.addProperties(properties)
val javaMap = properties.asMap
- val source = TableFactoryService.find(classOf[TableSourceFactory[_]],
javaMap)
- .asInstanceOf[TableSourceFactory[_]]
- .createTableSource(javaMap)
tableEnv match {
// check for a batch table source in this batch environment
case _: BatchTableEnvironment =>
- source match {
- case bts: BatchTableSource[_] =>
- new TableSourceSinkTable(Some(new BatchTableSourceTable(
- bts,
- new FlinkStatistic(externalCatalogTable.getTableStats))),
None)
- case _ => throw new TableException(
- s"Found table source '${source.getClass.getCanonicalName}' is
not applicable " +
- s"in a batch environment.")
- }
+ val source = TableFactoryService
--- End diff --
Usually it is very uncommon to define both a batch and streaming source in
the same factory. Your proposed change would require all future sources to
implement a check for the environment before which is unnecessary in 80% of the
cases. Separating by environment is a concept that can be find throughout the
entire `flink-table` module because both sources/sinks behave quite different.
---
