drill-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Humair M (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (DRILL-5575) Select column range
Date Wed, 07 Jun 2017 13:34:18 GMT

     [ https://issues.apache.org/jira/browse/DRILL-5575?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Humair M updated DRILL-5575:
----------------------------
    Description: 
Ability to select column ranges would be useful. When dealing with wide files being able to
select sequential range of columns without specifying each column will improve the usability.


So instead of this:
select columns[1], columns[2], columns[3], columns[8], columns[9], columns[10] from `data.csv`

use this:
select columns[1-3], columns[8-10] from `data.csv`

On a sidenote for extremely wide files (10K columns) the only way to query the data or load
it into native tables is to chop it up first as JVM is running out of constant limits (error:
too many constants) when querying the file directly or creating a table from it. This issue
has been reported elsewhere.

  was:
Ability to select column ranges would be useful. When dealing with wide files being able to
select sequential range of columns without specifying each column will improve the usability.


So instead of this:
select columns[1], columns[2], columns[3], columns[8], columns[9], columns[10] from `data.csv`

use this:
select columns[1-3], columns[8-10] from `data.csv`

On a sidenote for extremely wide files (10K columns) the only way to query the data or load
it into native tables is to chop it up first as Java is runnint out of the constant limits
(error: too many constants) when querying the file directly or creating a table from it. This
issue has been reported elsewhere.


> Select column range
> -------------------
>
>                 Key: DRILL-5575
>                 URL: https://issues.apache.org/jira/browse/DRILL-5575
>             Project: Apache Drill
>          Issue Type: Improvement
>          Components: SQL Parser
>    Affects Versions: 1.10.0
>         Environment: All
>            Reporter: Humair M
>            Priority: Trivial
>
> Ability to select column ranges would be useful. When dealing with wide files being able
to select sequential range of columns without specifying each column will improve the usability.

> So instead of this:
> select columns[1], columns[2], columns[3], columns[8], columns[9], columns[10] from `data.csv`
> use this:
> select columns[1-3], columns[8-10] from `data.csv`
> On a sidenote for extremely wide files (10K columns) the only way to query the data or
load it into native tables is to chop it up first as JVM is running out of constant limits
(error: too many constants) when querying the file directly or creating a table from it. This
issue has been reported elsewhere.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message