[SPARK-48498][SQL][3.5] Always do char padding in predicates #46958
Conversation
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
### What changes were proposed in this pull request? For some data sources, CHAR type padding is not applied on both the write and read sides (by disabling `spark.sql.readSideCharPadding`), as a different SQL flavor, which is similar to MySQL: https://dev.mysql.com/doc/refman/8.0/en/char.html However, there is a bug in Spark that we always pad the string literal when comparing CHAR type and STRING literals, which assumes the CHAR type columns are always padded, either on the write side or read side. This is not always true. This PR makes Spark always pad the CHAR type columns when comparing with string literals, to satisfy the CHAR type semantic. ### Why are the changes needed? bug fix if people disable read side char padding ### Does this PR introduce _any_ user-facing change? Yes. After this PR, comparing CHAR type with STRING literals follows the CHAR semantic, while before it mostly returns false. ### How was this patch tested? new tests ### Was this patch authored or co-authored using generative AI tooling? no Closes apache#46832 from cloud-fan/char. Authored-by: Wenchen Fan <wenchen@databricks.com> Signed-off-by: Wenchen Fan <wenchen@databricks.com>
jackylee-ch
commented
Jun 12, 2024
| .internal() | ||
| .doc("When true, Spark will not apply char type padding for CHAR type columns in string " + | ||
| s"comparison predicates, when '${READ_SIDE_CHAR_PADDING.key}' is false.") | ||
| .version("4.0.0") |
There was a problem hiding this comment.
@cloud-fan Any idea about how should we deal with the started version for this config since we backport it to 3.5?
There was a problem hiding this comment.
I think we can change it to 3.5.2 here
unigof
reviewed
Jun 17, 2024
|
|
||
| test("SPARK-48498: always do char padding in predicates") { | ||
| import testImplicits._ | ||
| withSQLConf(SQLConf.READ_SIDE_CHAR_PADDING.key -> "false") { |
There was a problem hiding this comment.
cloud you add a test for conf LEGACY_NO_CHAR_PADDING_IN_PREDICATE?
It maybe would fail when LEGACY_NO_CHAR_PADDING_IN_PREDICATE = true
https://github.com/apache/spark/pull/46832/files?diff=split&w=1#r1642368622
There was a problem hiding this comment.
It will fail, as the legacy behavior is wrong. This is why we make this fix...
|
@jackylee-ch can you rebase and re-trigger the test? |
…_padding_fix_to_3.5
|
3.5 is broken and will be fixed by #47022 . Please rebase after the fix is merged |
…_padding_fix_to_3.5
Done. It is ready to be merged now. @cloud-fan |
cloud-fan
approved these changes
Jun 24, 2024
|
thanks, merging to 3.5! |
cloud-fan
added a commit
that referenced
this pull request
Jun 24, 2024
### What changes were proposed in this pull request? For some data sources, CHAR type padding is not applied on both the write and read sides (by disabling `spark.sql.readSideCharPadding`), as a different SQL flavor, which is similar to MySQL: https://dev.mysql.com/doc/refman/8.0/en/char.html However, there is a bug in Spark that we always pad the string literal when comparing CHAR type and STRING literals, which assumes the CHAR type columns are always padded, either on the write side or read side. This is not always true. This PR makes Spark always pad the CHAR type columns when comparing with string literals, to satisfy the CHAR type semantic. ### Why are the changes needed? bug fix if people disable read side char padding ### Does this PR introduce _any_ user-facing change? Yes. After this PR, comparing CHAR type with STRING literals follows the CHAR semantic, while before it mostly returns false. ### How was this patch tested? new tests ### Was this patch authored or co-authored using generative AI tooling? no Closes #46958 from jackylee-ch/backport_char_padding_fix_to_3.5. Lead-authored-by: jackylee-ch <lijunqing@baidu.com> Co-authored-by: Wenchen Fan <wenchen@databricks.com> Signed-off-by: Wenchen Fan <wenchen@databricks.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
What changes were proposed in this pull request?
For some data sources, CHAR type padding is not applied on both the write and read sides (by disabling
spark.sql.readSideCharPadding), as a different SQL flavor, which is similar to MySQL: https://dev.mysql.com/doc/refman/8.0/en/char.htmlHowever, there is a bug in Spark that we always pad the string literal when comparing CHAR type and STRING literals, which assumes the CHAR type columns are always padded, either on the write side or read side. This is not always true.
This PR makes Spark always pad the CHAR type columns when comparing with string literals, to satisfy the CHAR type semantic.
Why are the changes needed?
bug fix if people disable read side char padding
Does this PR introduce any user-facing change?
Yes. After this PR, comparing CHAR type with STRING literals follows the CHAR semantic, while before it mostly returns false.
How was this patch tested?
new tests
Was this patch authored or co-authored using generative AI tooling?
no