[ 
https://issues.apache.org/jira/browse/FLINK-9444?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16531061#comment-16531061
 ] 

ASF GitHub Bot commented on FLINK-9444:
---------------------------------------

Github user twalthr commented on a diff in the pull request:

    https://github.com/apache/flink/pull/6218#discussion_r199736070
  
    --- Diff: 
flink-formats/flink-avro/src/test/java/org/apache/flink/formats/avro/typeutils/BackwardsCompatibleAvroSerializerTest.java
 ---
    @@ -26,7 +26,7 @@
     import org.apache.flink.api.java.typeutils.runtime.PojoSerializer;
     import 
org.apache.flink.api.java.typeutils.runtime.PojoSerializer.PojoSerializerConfigSnapshot;
     import org.apache.flink.core.memory.DataInputViewStreamWrapper;
    -import org.apache.flink.formats.avro.generated.User;
    +import org.apache.flink.formats.avro.generated.SimpleUser;
    --- End diff --
    
    The problem is that {{BackwardsCompatibleAvroSerializer}} does not support 
records with logical types. Logical types need a Kryo configuration that the 
serializer does not set correctly. This might be a bug or at least a missing 
feature. Given that this serializer only exists for backwards compatibility for 
1.3 (which used Avro 1.7 without logical type), I added a simple user for this 
test. I will add a comment about this to the code.


> KafkaAvroTableSource failed to work for map and array fields
> ------------------------------------------------------------
>
>                 Key: FLINK-9444
>                 URL: https://issues.apache.org/jira/browse/FLINK-9444
>             Project: Flink
>          Issue Type: Bug
>          Components: Kafka Connector, Table API & SQL
>    Affects Versions: 1.6.0
>            Reporter: Jun Zhang
>            Assignee: Jun Zhang
>            Priority: Blocker
>              Labels: patch, pull-request-available
>             Fix For: 1.6.0
>
>         Attachments: flink-9444.patch
>
>
> When some Avro schema has map/array fields and the corresponding TableSchema 
> declares *MapTypeInfo/ListTypeInfo* for these fields, an exception will be 
> thrown when registering the *KafkaAvroTableSource*, complaining like:
> Exception in thread "main" org.apache.flink.table.api.ValidationException: 
> Type Map<String, Integer> of table field 'event' does not match with type 
> GenericType<java.util.Map> of the field 'event' of the TableSource return 
> type.
>  at org.apache.flink.table.api.ValidationException$.apply(exceptions.scala:74)
>  at 
> org.apache.flink.table.sources.TableSourceUtil$$anonfun$validateTableSource$1.apply(TableSourceUtil.scala:92)
>  at 
> org.apache.flink.table.sources.TableSourceUtil$$anonfun$validateTableSource$1.apply(TableSourceUtil.scala:71)
>  at 
> scala.collection.IndexedSeqOptimized$class.foreach(IndexedSeqOptimized.scala:33)
>  at scala.collection.mutable.ArrayOps$ofRef.foreach(ArrayOps.scala:186)
>  at 
> org.apache.flink.table.sources.TableSourceUtil$.validateTableSource(TableSourceUtil.scala:71)
>  at 
> org.apache.flink.table.plan.schema.StreamTableSourceTable.<init>(StreamTableSourceTable.scala:33)
>  at 
> org.apache.flink.table.api.StreamTableEnvironment.registerTableSourceInternal(StreamTableEnvironment.scala:124)
>  at 
> org.apache.flink.table.api.TableEnvironment.registerTableSource(TableEnvironment.scala:438)



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to