[HIVE-10031] Modify the using of jobConf variable in ParquetRecordReaderWrapper constructor

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Assignee: Dong Chen
• Reporter: Dong Chen
• Created: 20/Mar/15 04:55
• Updated: 15/Apr/15 16:39
• Resolved: 15/Apr/15 16:39

Description

In ParquetRecordReaderWrapper constructor, it create splits, set projections and filters in conf, create task context, and then create Parquet record reader. In this procedure, we could improve the logic of conf usage:
1. the updated jobConf is not passed to Parquet in one case. (https://github.com/apache/hive/blob/trunk/ql/src/java/org/apache/hadoop/hive/ql/io/parquet/read/ParquetRecordReaderWrapper.java#L103)
I think we should use the instance variable jobConf, which is updated when projectioning and filtering, instead of passed in oldJobConf.