spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Matt Boegner (Jira)" <j...@apache.org>
Subject [jira] [Updated] (SPARK-30130) Hardcoded numeric values in common table expressions which utilize GROUP BY are interpreted as ordinal positions
Date Wed, 04 Dec 2019 17:54:00 GMT

     [ https://issues.apache.org/jira/browse/SPARK-30130?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Matt Boegner updated SPARK-30130:
---------------------------------
    Description: 
Hardcoded numeric values in common table expressions which utilize GROUP BY are interpreted
as ordinal positions. 
{code:java}
val df = spark.sql("""
 with a as (select 0 as test, count group by test)
 select * from a
 """)
 df.show(){code}

 This results in an error message like {color:#e01e5a}GROUP BY position 0 is not in select
list (valid range is [1, 2]){color} .

 

However, this error does not appear in a traditional subselect format. For example, this query
executes correctly:
{code:java}
val df = spark.sql("""
 select * from (select 0 as test, count group by test) a
 """)
 df.show(){code}

  

  was:
Hardcoded numeric values in common table expressions which utilize GROUP BY are interpreted
as ordinal positions. 
val df = spark.sql("""
with a as (select 0 as test, count(*) group by test)
select * from a
""")
df.show()
This results in an error message like {color:#e01e5a}GROUP BY position 0 is not in select
list (valid range is [1, 2]){color} .

 

However, this error does not appear in a traditional subselect format. For example, this query
executes correctly:
val df = spark.sql("""
select * from (select 0 as test, count(*) group by test) a
""")
df.show()
 


> Hardcoded numeric values in common table expressions which utilize GROUP BY are interpreted
as ordinal positions
> ----------------------------------------------------------------------------------------------------------------
>
>                 Key: SPARK-30130
>                 URL: https://issues.apache.org/jira/browse/SPARK-30130
>             Project: Spark
>          Issue Type: Bug
>          Components: SQL
>    Affects Versions: 2.4.4
>            Reporter: Matt Boegner
>            Priority: Minor
>
> Hardcoded numeric values in common table expressions which utilize GROUP BY are interpreted
as ordinal positions. 
> {code:java}
> val df = spark.sql("""
>  with a as (select 0 as test, count group by test)
>  select * from a
>  """)
>  df.show(){code}
>  This results in an error message like {color:#e01e5a}GROUP BY position 0 is not in select
list (valid range is [1, 2]){color} .
>  
> However, this error does not appear in a traditional subselect format. For example, this
query executes correctly:
> {code:java}
> val df = spark.sql("""
>  select * from (select 0 as test, count group by test) a
>  """)
>  df.show(){code}
>   



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org


Mime
View raw message