[
https://issues.apache.org/jira/browse/SPARK-32721?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Chao Sun updated SPARK-32721:
-----------------------------
Description:
The following if clause:
{code:sql}
if(p, null, false)
{code}
can be simplified to:
{code:sql}
and(p, null)
{code}
And similarly, the following clause:
{code:sql}
if(p, null, true)
{code}
can be simplified to:
{code:sql}
or(not(p), null)
{code}
iff predicate {{p}} is deterministic, i.e., can be evaluated to either true or
false, but not null.
{{and}} and {{or}} clauses are more optimization friendly. For instance, by
converting {{if(col > 42, null, false)}} to {{and(col > 42, null)}}, we can
potentially push the filter {{col > 42}} down to data sources to avoid
unnecessary IO.
was:
The following if clause:
{code:sql}
if(p, null, false)
{code}
can be simplified to:
{code:sql}
and(p, null)
{code}
And similarly, the following clause:
{code:sql}
if(p, null, true)
{code}
can be simplified to:
{code:sql}
or(not(p), null)
{code}
iff predicate {{p}} is deterministic, i.e., can be evaluated to either true or
false, but not null.
> Simplify if clauses with null and boolean
> -----------------------------------------
>
> Key: SPARK-32721
> URL: https://issues.apache.org/jira/browse/SPARK-32721
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.0.0
> Reporter: Chao Sun
> Priority: Major
>
> The following if clause:
> {code:sql}
> if(p, null, false)
> {code}
> can be simplified to:
> {code:sql}
> and(p, null)
> {code}
> And similarly, the following clause:
> {code:sql}
> if(p, null, true)
> {code}
> can be simplified to:
> {code:sql}
> or(not(p), null)
> {code}
> iff predicate {{p}} is deterministic, i.e., can be evaluated to either true
> or false, but not null.
> {{and}} and {{or}} clauses are more optimization friendly. For instance, by
> converting {{if(col > 42, null, false)}} to {{and(col > 42, null)}}, we can
> potentially push the filter {{col > 42}} down to data sources to avoid
> unnecessary IO.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
