Skip to content
Snippets Groups Projects
Commit 5363ed71 authored by Liang-Chi Hsieh's avatar Liang-Chi Hsieh Committed by Yin Huai
Browse files

[SPARK-9361] [SQL] Refactor new aggregation code to reduce the times of checking compatibility

JIRA: https://issues.apache.org/jira/browse/SPARK-9361

Currently, we call `aggregate.Utils.tryConvert` in many places to check it the logical.Aggregate can be run with new aggregation. But looks like `aggregate.Utils.tryConvert` will cost considerable time to run. We should only call `tryConvert` once and keep it value in `logical.Aggregate` and reuse it.

In `org.apache.spark.sql.execution.aggregate.Utils`, the codes involving with `tryConvert` should be moved to catalyst because it actually doesn't deal with execution details.

Author: Liang-Chi Hsieh <viirya@appier.com>

Closes #7677 from viirya/refactor_aggregate and squashes the following commits:

babea30 [Liang-Chi Hsieh] Merge remote-tracking branch 'upstream/master' into refactor_aggregate
9a589d7 [Liang-Chi Hsieh] Fix scala style.
0a91329 [Liang-Chi Hsieh] Refactor new aggregation code to reduce the times to call tryConvert.
parent 7bbf02f0
No related branches found
No related tags found
No related merge requests found
Loading
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment