Check list size before concat in ScalarValue#10329
Merged
alamb merged 2 commits intoapache:mainfrom May 2, 2024
Merged
Conversation
…d. If not, do not attempt to call concat on empty arrays.
Contributor
|
Thanks @timsaucer -- I started the CI checks on this PR and hope to check it out later today if no one beats me to it |
Contributor
|
Thanks @timsaucer and @jayzhan211 |
Contributor
|
BTW I think this would benefit from some end to end testing as well, so I filed #10345 |
richox
pushed a commit
to auron-project/datafusion
that referenced
this pull request
May 8, 2024
* Add check in list_to_array_of_size to see if any data will be returned. If not, do not attempt to call concat on empty arrays. * Formatting
richox
pushed a commit
to auron-project/datafusion
that referenced
this pull request
May 8, 2024
* Add check in list_to_array_of_size to see if any data will be returned. If not, do not attempt to call concat on empty arrays. * Formatting
richox
pushed a commit
to apache/auron
that referenced
this pull request
May 9, 2024
revert "split huge batch into smaller ones during projection (#448)", This reverts commit 7d33df1. fix issue "Nondeterministic expression xxx should be initialized before eval". fix issue "IO error: Bad file descriptor" while accessing spill file. fix jvm coredump by removing chrono dependency for printing logs.
Merged
lihao712
pushed a commit
to apache/auron
that referenced
this pull request
May 9, 2024
revert "split huge batch into smaller ones during projection (#448)", This reverts commit 7d33df1. fix issue "Nondeterministic expression xxx should be initialized before eval". fix issue "IO error: Bad file descriptor" while accessing spill file. fix jvm coredump by removing chrono dependency for printing logs. Co-authored-by: zhangli20 <zhangli20@kuaishou.com>
joroKr21
pushed a commit
to coralogix/arrow-datafusion
that referenced
this pull request
May 24, 2024
* Add check in list_to_array_of_size to see if any data will be returned. If not, do not attempt to call concat on empty arrays. * Formatting
joroKr21
added a commit
to coralogix/arrow-datafusion
that referenced
this pull request
May 24, 2024
* Add check in list_to_array_of_size to see if any data will be returned. If not, do not attempt to call concat on empty arrays. * Formatting Co-authored-by: Tim Saucer <timsaucer@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR resolves an issue in which we get a panic attempting to concatenate empty arrays. This prevents lead and lag functions from working on these data types.
Which issue does this PR close?
Closes #10328.
Rationale for this change
When we call
WindowAggState::newwith a data type of a list, we will callScalarValue::try_from(out_type)?.to_array_of_size(0)which in turn callslist_to_array_of_size. Within this function there is no check on the return size. It could be zero due to either an empty array input or a request for 0 elements. In either of these conditions, you will get a panic fromconcat.What changes are included in this PR?
This is a small change to check the return size. When zero it will just take an empty slice of the original array.
Are these changes tested?
I have included an expansion to an existing unit test to cover this. In my bug report I have also shown minimal example to reproduce the problem and the updated results from it's inclusion.
Are there any user-facing changes?
There are no user facing changes.