[ 
https://issues.apache.org/jira/browse/HBASE-25444?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Junsheng Tan updated HBASE-25444:
---------------------------------
    Description: 
Say we want to check whether a cell value is within a set of values.
If the size of the set is 1000, we need to create 1000 SingleColumnValueFilter 
with CompareOperator.EQUAL and wrap them with a FilterListWithOR. The scan 
performance of these filters would be very bad.
But if we have a set filter, we just need to create only one filter with 1000 
values.  And the containing operation could be implemented as O(1) or O(log( n 
)).
This `SetFilter` is general and useful in my opinion, so would the community 
consider to implement a `SetFilter`?

  was:
Say we want to check whether a cell value is within a set of values.
If the size of the set is 1000, we need to create 1000 SingleColumnValueFilter 
with CompareOperator.EQUAL and wrap them with a FilterListWithOR. The scan 
performance of these filters would be very bad.
But if we have a set filter, we just need to create only one filter with 1000 
values.  And the containing operation could be implemented as O(1) or O(log(n)).
This `SetFilter` is general and useful in my opinion, so would the community 
consider to implement a `SetFilter`?


> Add a filter with containing semantic
> -------------------------------------
>
>                 Key: HBASE-25444
>                 URL: https://issues.apache.org/jira/browse/HBASE-25444
>             Project: HBase
>          Issue Type: New Feature
>          Components: Filters
>    Affects Versions: 3.0.0-alpha-1
>            Reporter: Junsheng Tan
>            Priority: Major
>
> Say we want to check whether a cell value is within a set of values.
> If the size of the set is 1000, we need to create 1000 
> SingleColumnValueFilter with CompareOperator.EQUAL and wrap them with a 
> FilterListWithOR. The scan performance of these filters would be very bad.
> But if we have a set filter, we just need to create only one filter with 1000 
> values.  And the containing operation could be implemented as O(1) or O(log( 
> n )).
> This `SetFilter` is general and useful in my opinion, so would the community 
> consider to implement a `SetFilter`?



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

Reply via email to