[ 
https://issues.apache.org/jira/browse/FLINK-16606?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Hequn Cheng updated FLINK-16606:
--------------------------------
    Release Note: 
Currently, DataTypes can be configured with some parameters, e.g., precision. 
However, the precision provided by users takes no effect now but a default 
value will always be used. This would bring some confusion to users. In this 
task, exceptions will be thrown if the value is not supported to make it more 
visible to users.  Changes include:
 - the precision for TimeType could only be 0
 - the length for VarBinaryType/VarCharType could only be 0x7fffffff
 - the precision/scale for DecimalType could only be 38/18
 - the precision for TimestampType/LocalZonedTimestampType could only be 3
 - the resolution for DayTimeIntervalType could only be `SECOND` and the 
fractionalPrecision could only be 3
 - the resolution for YearMonthIntervalType could only be `MONTH` and the 
`yearPrecision` could only be 2
 - the CharType/BinaryType/ZonedTimestampType is not supported

  was:Currently, DataTypes can be configured with some parameters, e.g., 
precision. However, the precision provided by users takes no effect now but a 
default value will always be used. This would bring some confusion to users. In 
this task, we throw exceptions if the value is not supported to make it more 
visible to users. 


> Throw exceptions for the data types which are not currently supported
> ---------------------------------------------------------------------
>
>                 Key: FLINK-16606
>                 URL: https://issues.apache.org/jira/browse/FLINK-16606
>             Project: Flink
>          Issue Type: Improvement
>          Components: API / Python
>    Affects Versions: 1.10.0
>            Reporter: Dian Fu
>            Assignee: Dian Fu
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 1.11.0
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> Currently, there are still cases where a data type isn't supported as the 
> Python data type will be firstly converted to TypeInformation which will lose 
> a few information, e.g, 
>  - the precision for TimeType could only be 0
>  - the length for VarBinaryType/VarCharType could only be 0x7fffffff
>  - the precision/scale for DecimalType could only be 38/18
>  - the precision for TimestampType/LocalZonedTimestampType could only be 3
>  - the resolution for DayTimeIntervalType could only be `SECOND` and the 
> fractionalPrecision could only be 3
>  - the resolution for YearMonthIntervalType could only be `MONTH` and the 
> `yearPrecision` could only be 2
>  - the CharType/BinaryType/ZonedTimestampType is not supported
> We should throw exceptions for the cases not supported.



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

Reply via email to