[
https://issues.apache.org/jira/browse/CARBONDATA-2926?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ravindra Pesala resolved CARBONDATA-2926.
-----------------------------------------
Resolution: Fixed
Fix Version/s: 1.5.0
> ArrayIndexOutOfBoundException if varchar column is present before dictionary
> columns along with empty sort_columns.
> -------------------------------------------------------------------------------------------------------------------
>
> Key: CARBONDATA-2926
> URL: https://issues.apache.org/jira/browse/CARBONDATA-2926
> Project: CarbonData
> Issue Type: Bug
> Reporter: Ajantha Bhat
> Assignee: Ajantha Bhat
> Priority: Major
> Fix For: 1.5.0
>
> Time Spent: 2h 20m
> Remaining Estimate: 0h
>
> ArrayIndexOutOfBoundException if varchar column is present before dictionary
> columns along with empty sort_columns.
> cause:
> CarbonFactDataHandlerColumnar.isVarcharColumnFull() method uses
> model.getVarcharDimIdxInNoDict()
> and index of varchar column in no dictonary array became negative.
> currently index was calculated based on ordinal-number of dictionary columns.
> This can go negative in no_sort column case,
> solution:
> take the varchar dimension index from no dictionary array from at runtime
> based on schema.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
