You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@drill.apache.org by "Jaroslaw Sosnicki (JIRA)" <ji...@apache.org> on 2015/12/02 19:03:10 UTC

[jira] [Created] (DRILL-4151) CSV Support with multiline header

Jaroslaw Sosnicki created DRILL-4151:
----------------------------------------

             Summary: CSV Support with multiline header
                 Key: DRILL-4151
                 URL: https://issues.apache.org/jira/browse/DRILL-4151
             Project: Apache Drill
          Issue Type: Wish
          Components: Functions - Drill
            Reporter: Jaroslaw Sosnicki


The modern data sources produce CSV files with two header lines:
first line contains field descriptions while second line contains filed types
Would be feasible to implement such a format in to DRILL as additional storage format type?

This example demonstrates an output CSV header from one of the data sources.


LDEV_COUNT,MONITORED_LDEV_COUNT,READ_IO_COUNT,READ_IO_RATE,READ_HIT_IO_COUNT,READ_HIT_RATE,WRITE_IO_COUNT,WRITE_IO_RATE,WRITE_HIT_IO_COUNT,WRITE_HIT_RATE,READ_MBYTES,READ_XFER_RATE,WRITE_MBYTES,WRITE_XFER_RATE,INTERVAL,INPUT_RECORD_TYPE,RECORD_TIME
ulong,ulong,double,double,double,double,double,double,double,double,double,double,double,double,ulong,string(8),time_t







--
This message was sent by Atlassian JIRA
(v6.3.4#6332)