Mailing-List: contact dev-help@phoenix.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@phoenix.apache.org
Date: Wed, 1 Apr 2015 07:28:53 +0000 (UTC)
From: "Alicia Ying Shu (JIRA)" <jira@apache.org>
To: dev@phoenix.incubator.apache.org
Message-ID: <JIRA.12767246.1421179567000.89850.1427873333543@Atlassian.JIRA>
In-Reply-To: <JIRA.12767246.1421179567000@Atlassian.JIRA>
References: <JIRA.12767246.1421179567000@Atlassian.JIRA>
 <JIRA.12767246.1421179567568@arcas>
Subject: [jira] [Comment Edited] (PHOENIX-1580) Support UNION ALL
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/PHOENIX-1580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14390020#comment-14390020 ] 

Alicia Ying Shu edited comment on PHOENIX-1580 at 4/1/15 7:28 AM:
------------------------------------------------------------------

[~jamestaylor] [~maryannxue]  The temp schema is to work around Aggregate projection to replace rowProjector. OrderBy expression is normally column name, it should be the same no matter in the server or in the client. The temp schema is already built with the projection names (column names) as specified in the SQL. All UnionAllIT tests passed with Plan compiled Order By and Limit. 

Wrapping another Select over current Union All selects we need to specify a table. A dummy table or Null table did not work since parser needs a real table to resolve column information etc. I have to say if we have taken my very early approach of passing down rowProjector, we can simplify the approach quite bit. Given that column names are used to construct the temp schema, no need to wrap another select over. 


was (Author: aliciashu):
[~jamestaylor] [~maryannxue]  The temp schema is to work around Aggregate projection to replace rowProjector. OrderBy expression is normally column name, it should be the same no matter in the server or in the client. It applies to alias as well. The temp schema is already built with the projection names (column names) as specified in the SQL. I did not see an issue here. All UnionAllIT tests passed with Plan compiled Order By and Limit. 

Wrapping another Select over current Union All selects we need to specify a table. A dummy table or Null table did not work since parser needs a real table to resolve column information etc. I have to say if we have taken my very early approach of passing down rowProjector, we can simplify the approach quite bit. Given that column names are used to construct the temp schema, no need to wrap another select over. 


> Support UNION ALL
> -----------------
>
>                 Key: PHOENIX-1580
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-1580
>             Project: Phoenix
>          Issue Type: Bug
>            Reporter: Alicia Ying Shu
>            Assignee: Alicia Ying Shu
>         Attachments: PHOENIX-1580-grammar.patch, Phoenix-1580-v1.patch, phoenix-1580-v1-wipe.patch, phoenix-1580.patch, unionall-wipe.patch
>
>
> Select * from T1
> UNION ALL
> Select * from T2


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