geode-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "John Blum (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (GEODE-1929) Specifying ORDER BY in a OQL Query should not require the SELECT to be DISTINCT
Date Fri, 23 Sep 2016 17:46:21 GMT

     [ https://issues.apache.org/jira/browse/GEODE-1929?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

John Blum updated GEODE-1929:
-----------------------------
    Description: 
Currently, when a user writes a sorted OQL query, the user would specify an {{ORDER BY}} clause
indicating which object properties to sort on along with the direction of the sort.  Additionally,
for some non-apparent reason, Geode also requires the result of the query to be unique (a.k.a.
{{DISTINCT}}).

Therefore, the following query would considered to be invalid by Geode...

{code:sql}
SELECT * FROM /People p WHERE p.lastname = 'Doe' ORDER BY p.lastname ASC, p.birthDate DESC
{code}

To correct this query, a user much specify the {{DISTINCT}} OQL keyword...

{code:sql}
SELECT DISTINCT * FROM /People p WHERE p.lastname = 'Doe' ORDER BY p.lastname ASC, p.birthDate
DESC
{code}

However, sorting has nothing to do with uniqueness as it is still possible to sort results
even when they contain duplicates.

This seems to be a technical limitation in Geode rather than an apparent limitation of data
grid technology, which has be inappropriately exposed to the end user.

  was:
Currently, when a user writes a sorted OQL query, the user would specify an {{ORDER BY}} clause
indicating which object properties to sort on along with the direction of the sort.  Additionally,
for some non-apparent reason, Geode also requires the result of the query to be unique (a.k.a.
DISTINCT).

Therefore, the following query would considered to be invalid by Geode...

{code:sql}
SELECT * FROM /People p WHERE p.lastname = 'Doe' ORDER BY p.lastname ASC, p.birthDate DESC
{code}

To correct this query, a user much specify the {{DISTINCT}} OQL keyword...

{code:sql}
SELECT DISTINCT * FROM /People p WHERE p.lastname = 'Doe' ORDER BY p.lastname ASC, p.birthDate
DESC
{code}

However, sorting has nothing to do with uniqueness as it is still possible to sort results
even when they contain duplicates.

This seems to be a technical limitation in Geode rather than an apparent limitation of data
grid technology, which has be inappropriately exposed to the end user.


> Specifying ORDER BY in a OQL Query should not require the SELECT to be DISTINCT
> -------------------------------------------------------------------------------
>
>                 Key: GEODE-1929
>                 URL: https://issues.apache.org/jira/browse/GEODE-1929
>             Project: Geode
>          Issue Type: Bug
>          Components: querying
>            Reporter: John Blum
>
> Currently, when a user writes a sorted OQL query, the user would specify an {{ORDER BY}}
clause indicating which object properties to sort on along with the direction of the sort.
 Additionally, for some non-apparent reason, Geode also requires the result of the query to
be unique (a.k.a. {{DISTINCT}}).
> Therefore, the following query would considered to be invalid by Geode...
> {code:sql}
> SELECT * FROM /People p WHERE p.lastname = 'Doe' ORDER BY p.lastname ASC, p.birthDate
DESC
> {code}
> To correct this query, a user much specify the {{DISTINCT}} OQL keyword...
> {code:sql}
> SELECT DISTINCT * FROM /People p WHERE p.lastname = 'Doe' ORDER BY p.lastname ASC, p.birthDate
DESC
> {code}
> However, sorting has nothing to do with uniqueness as it is still possible to sort results
even when they contain duplicates.
> This seems to be a technical limitation in Geode rather than an apparent limitation of
data grid technology, which has be inappropriately exposed to the end user.



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

Mime
View raw message