[
https://issues.apache.org/jira/browse/SPARK-34638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17877595#comment-17877595
]
Jiri Humpolicek edited comment on SPARK-34638 at 8/29/24 6:03 AM:
------------------------------------------------------------------
[~viirya] Do you think it would be possible to do that? I think it will be
great feature when spark reads only necessary fields from query in general way.
In case of rich nested structures it could save huge amount of resources. I
found unresolved improvement for this more general case from last year
https://issues.apache.org/jira/browse/SPARK-42879 .
was (Author: yuryn):
[~viirya] Do you think it would be possible to do that? I think it will be
great feature when spark reads only necessary fields from query in general way.
In case of rich nested structures it could safe huge amount of resources. I
found unresolved improvement for this more general case from last year
https://issues.apache.org/jira/browse/SPARK-42879 .
> Spark SQL reads unnecessary nested fields (another type of pruning case)
> ------------------------------------------------------------------------
>
> Key: SPARK-34638
> URL: https://issues.apache.org/jira/browse/SPARK-34638
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.1.1
> Reporter: Jiri Humpolicek
> Assignee: L. C. Hsieh
> Priority: Major
> Fix For: 3.2.0
>
>
> Based on this [SPARK-29721|https://issues.apache.org/jira/browse/SPARK-29721]
> I found another nested fields pruning case.
> Example:
> 1) Loading data
> {code:scala}
> val jsonStr = """{
> "items": [
> {"itemId": 1, "itemData": "a"},
> {"itemId": 2, "itemData": "b"}
> ]
> }"""
> val df = spark.read.json(Seq(jsonStr).toDS)
> df.write.format("parquet").mode("overwrite").saveAsTable("persisted")
> {code}
> 2) read query with explain
> {code:scala}
> val read = spark.table("persisted")
> spark.conf.set("spark.sql.optimizer.nestedSchemaPruning.enabled", true)
> read.select(explode($"items").as('item)).select($"item.itemId").explain(true)
> // ReadSchema: struct<items:array<struct<itemData:string,itemId:bigint>>>
> {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
