-
Notifications
You must be signed in to change notification settings - Fork 28.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[SPARK-32526][SQL]Fix some test cases of sql/catalyst
module in scala 2.13
#29370
[SPARK-32526][SQL]Fix some test cases of sql/catalyst
module in scala 2.13
#29370
Conversation
@dongjoon-hyun, this pr try to fix |
cc @srowen from the numbers alone, this pr fixed about 70% failed cases of |
I will update failure list to JIRA after this pr later |
details of remaining failed cases already updated to the attachment and added comments in SPARK-32526 |
This change itself looks fine. If you think the rest of the changes in sql / catalyst are similar, I'd make them in one go. If there are difficult or logically distinct other fixes, that can be another PR. They can all be attached to one JIRA |
Yeah, let's try to do it in one go. |
@HyukjinKwon @srowen ok ~ let me give it a try~ |
ok to test |
The main change of Address beb3928 is call The effect as follow:
Before
After
|
Test build #127179 has finished for PR 29370 at commit
|
…etSuite and InferFiltersFromConstraintsSuite
The main change of Address eeeb28a impl The effect as follow:
Before
After
|
Test build #127188 has finished for PR 29370 at commit
|
Looking good. I think you're welcome to keep going and we can 'checkpoint' and commit some fixes whenever there is a logical and large enough group of them to commit. That could be now too, up to you. |
The main change of Address 4018b1f as follow:
The effect as follow:
Before
After
|
@srowen agree with u, we can make a 'checkpoint' if 4018b1f no problem, and I will update the PR description. I found that It is difficult to fix the cases caused by For example, |
sql/catalyst
module in scala 2.13
Test build #127260 has finished for PR 29370 at commit
|
Jenkins retest this please |
Test build #127283 has finished for PR 29370 at commit
|
If there aren't any more fixes that are similar here, I can merge this. We can add more now if there are others that are easily fixable in SQL. |
Test build #127356 has started for PR 29370 at commit |
@srowen There are still some fixes left to commit and will be done today |
The main change of Address db90b1a as follow:
The effect as follow:
Before
After
|
The main change of Address fca4fa7 as follow:
The effect as follow:
Before
After
|
Test build #127360 has started for PR 29370 at commit |
Test build #127364 has started for PR 29370 at commit |
The main change of Address 5328d33 is add a sort
The effect as follow:
Before
After
|
@srowen After 5328d33, I think that there are no more similar problems that can be fixed. The reasons for the remaining failure cases can be divided into two categories:
|
Test build #127367 has finished for PR 29370 at commit
|
Jenkins retest this please |
Test build #127385 has finished for PR 29370 at commit
|
Merged to master. I didn't resolve the JIRA yet because there are a few more tests to fix. |
Nice to get this being fixed! |
@srowen @HyukjinKwon I will try to give a new pr to resolve rest problems |
What changes were proposed in this pull request?
The purpose of this pr is to partial resolve SPARK-32526, total of 88 failed and 2 aborted test cases were fixed, the related suite as follow:
DataSourceV2AnalysisBaseSuite
related test cases (71 FAILED -> Pass)TreeNodeSuite
(1 FAILED -> Pass)MetadataSuite
(1 FAILED -> Pass)InferFiltersFromConstraintsSuite
(3 FAILED -> Pass)StringExpressionsSuite
(1 FAILED -> Pass)JacksonParserSuite
(1 FAILED -> Pass)HigherOrderFunctionsSuite
(1 FAILED -> Pass)ExpressionParserSuite
(1 FAILED -> Pass)CollectionExpressionsSuite
(6 FAILED -> Pass)SchemaUtilsSuite
(2 FAILED -> Pass)ExpressionSetSuite
(ABORTED -> Pass)ArrayDataIndexedSeqSuite
(ABORTED -> Pass)The main change of this pr as following:
Optimizer
andAnalyzer
are changed to pass compile,ArrayBuffer
is not aSeq
in scala 2.13, calltoSeq
method manually to compatible with Scala 2.12m.mapValues().view.force
pattern return aMap
in scala 2.12 but return aIndexedSeq
in scala 2.13, calltoMap
method manually to compatible with Scala 2.12.TreeNode
are changed to passDataSourceV2AnalysisBaseSuite
related test cases andTreeNodeSuite
failed case.call
toMap
method ofMetadata#hash
methodcase map
branch becausemap.mapValues
returnMap
in Scala 2.12 and returnMapView
in Scala 2.13.impl
contact method ofExpressionSet
in Scala 2.13 version refer toExpressionSet
in Scala 2.12 to support+ +
method conform toExpressionSet
semanticsGenericArrayData
not acceptArrayBuffer
input, calltoSeq
when useArrayBuffer
constructionGenericArrayData
for Scala version compatibilityCall
toSeq
inRandomDataGenerator#randomRow
method to ensure contents offields
isSeq
notArrayBuffer
Call
toSeq
LetJacksonParser#parse
still return aSeq
because the check method ofJacksonParserSuite#"skipping rows using pushdown filters"
dependence onSeq
typeCall
toSeq
inAstBuilder#visitFunctionCall
, otherwisectx.argument.asScala.map(expression)
isBuffer
in Scala 2.13Add a
LongType
match toArraySetLike.nullValueHolder
Add a
sorted
to ensureduplicateColumns
string inSchemaUtils.checkColumnNameDuplication
method error message have a deterministic orderWhy are the changes needed?
We need to support a Scala 2.13 build.
Does this PR introduce any user-facing change?
No
How was this patch tested?
Scala 2.12: Pass the Jenkins or GitHub Action
Scala 2.13: Do the following:
Before
After