[
https://issues.apache.org/jira/browse/FLINK-12708?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jark Wu updated FLINK-12708:
----------------------------
Description:
In order to support Blink batch and temporal table join, we need some new
source&sink interfaces and changes about TableSchema.
- {{AsyncTableFunction}}: used as async lookup function, to support async
temporal table join (i.e. dimension table join.).
- {{LookupableTableSource}}: The {{LookupableTableSource}} interface adds
support for the table to be accessed via key column(s) in a lookup fashion in
order to support temporal table join.
- {{BoundedTableSource}}: used for batch table source, returns a bounded
DataStream, not a {{InputFormat}}, because we also want to reuse existing
streaming source implementation to support bounded source, for example: bounded
Kafka source.
- {{BoundedTableSink}}: used for batch table sink, emit a bounded DataStream,
not a {{OutputFormat}}. The reason is the same as above.
- {{TableSchema}}: The {{TableSchema}} is extended to carry more information
about primary key and unique keys. The key information will be used in
optimization.
was:
çIn order to support Blink batch and temporal table join, we need some new
source&sink interfaces and changes about TableSchema.
- {{AsyncTableFunction}}: used as async lookup function, to support async
temporal table join (i.e. dimension table join.).
- {{LookupableTableSource}}: The {{LookupableTableSource}} interface adds
support for the table to be accessed via key column(s) in a lookup fashion in
order to support temporal table join.
- {{BoundedTableSource}}: used for batch table source, returns a bounded
DataStream, not a {{InputFormat}}, because we also want to reuse existing
streaming source implementation to support bounded source, for example: bounded
Kafka source.
- {{BoundedTableSink}}: used for batch table sink, emit a bounded DataStream,
not a {{OutputFormat}}. The reason is the same as above.
- {{TableSchema}}: The {{TableSchema}} is extended to carry more information
about primary key and unique keys. The key information will be used in
optimization.
> Introduce new Interfaces for source and sink to make Blink runner work
> ----------------------------------------------------------------------
>
> Key: FLINK-12708
> URL: https://issues.apache.org/jira/browse/FLINK-12708
> Project: Flink
> Issue Type: New Feature
> Components: Table SQL / API
> Reporter: Jark Wu
> Assignee: Jark Wu
> Priority: Major
>
> In order to support Blink batch and temporal table join, we need some new
> source&sink interfaces and changes about TableSchema.
> - {{AsyncTableFunction}}: used as async lookup function, to support async
> temporal table join (i.e. dimension table join.).
> - {{LookupableTableSource}}: The {{LookupableTableSource}} interface adds
> support for the table to be accessed via key column(s) in a lookup fashion in
> order to support temporal table join.
> - {{BoundedTableSource}}: used for batch table source, returns a bounded
> DataStream, not a {{InputFormat}}, because we also want to reuse existing
> streaming source implementation to support bounded source, for example:
> bounded Kafka source.
> - {{BoundedTableSink}}: used for batch table sink, emit a bounded
> DataStream, not a {{OutputFormat}}. The reason is the same as above.
> - {{TableSchema}}: The {{TableSchema}} is extended to carry more information
> about primary key and unique keys. The key information will be used in
> optimization.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
