[jira] [Created] (FLINK-19981) Improve dealing with Row in DataStream API

["Timo Walther (Jira)" <ji...@apache.org>]

Timo Walther created FLINK-19981:
------------------------------------

             Summary: Improve dealing with Row in DataStream API
                 Key: FLINK-19981
                 URL: https://issues.apache.org/jira/browse/FLINK-19981
             Project: Flink
          Issue Type: Sub-task
          Components: Table SQL / API
            Reporter: Timo Walther
            Assignee: Timo Walther


Working with indices for accessing a row's fields is cumbersome especially if a row consists of many columns (some pipelines have 500+ columns) and a column name needs to be mapped to an index before a field can be accessed. We can leverage the available schema information to make dealing with rows more convenient. The runtime impact should be neglectable if implemented correctly. If maximum performance is required people can fallback to index-based field access or even internal RowData structure at any time.

This introduces:
{code}
Row.withPositions(int length)
Row.withNames()
Row.getField(String)
Row.setField(String, T value)
{code}



--
This message was sent by Atlassian Jira
(v8.3.4#803005)