[ 
https://issues.apache.org/jira/browse/FLINK-9963?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Ruidong Li updated FLINK-9963:
------------------------------
    Description: 
Sometimes it might be useful to just read or write a single value into Kafka or 
other connectors. We should add a single-value {{SerializationSchemaFactory}} 
and single-value {{DeserializationSchemaFactory}}, the types below and their 
array types shall be considered.

{{byte, short, int, long, float, double, string}}

For the numeric types, we might want to specify the endian format.

  was:
Sometimes it might be useful to just read or write a string into Kafka or other 
connectors. We should add a simple string {{SerializationSchemaFactory}} and 
{{DeserializationSchemaFactory}}. How we want to represent all data types and 
nested types is still up for discussion. We could also support just a single 
string field?

Schema derivation should be supported by the factories.

See {{org.apache.flink.formats.json.JsonRowFormatFactory}} for a reference.


> Add a single value table format factory
> ---------------------------------------
>
>                 Key: FLINK-9963
>                 URL: https://issues.apache.org/jira/browse/FLINK-9963
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Table API & SQL
>            Reporter: Timo Walther
>            Assignee: Ruidong Li
>            Priority: Major
>              Labels: pull-request-available
>
> Sometimes it might be useful to just read or write a single value into Kafka 
> or other connectors. We should add a single-value 
> {{SerializationSchemaFactory}} and single-value 
> {{DeserializationSchemaFactory}}, the types below and their array types shall 
> be considered.
> {{byte, short, int, long, float, double, string}}
> For the numeric types, we might want to specify the endian format.



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

Reply via email to