[SPARK-48681][SQL] Use ICU in Lower/Upper expressions for UTF8_BINARY strings #47043
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
uros-db
changed the title
uros-db
force-pushed
the
change-lower-upper
branch
from
ca93b88 to
194177d
Compare
uros-db
changed the title
mkaravel
approved these changes
Jun 21, 2024
common/unsafe/src/test/java/org/apache/spark/unsafe/types/CollationSupportSuite.java
Show resolved
Hide resolved
Contributor
|
Please also update the PR description and explain why we are making this change. |
cloud-fan
approved these changes
Jun 24, 2024
Contributor
|
thanks, merging to master! |
cloud-fan
pushed a commit
that referenced
this pull request
Jun 28, 2024
### What changes were proposed in this pull request? Update `InitCap` Spark expressions to use ICU case mappings for UTF8_BINARY collation, instead of the currently used JVM case mappings. This behaviour is put under the `ICU_CASE_MAPPINGS_ENABLED` flag in SQLConf, which is true by default. Note: the same flag is used for `Lower` & `Upper` expressions, with changes introduced in: #47043. ### Why are the changes needed? To keep the consistency between collations - all collations shouls use ICU-based case mappings, including the UTF8_BINARY collation. ### Does this PR introduce _any_ user-facing change? Yes, the behaviour of `initcap` string function for UTF8_BINARY will now rely on ICU-based case mappings. However, by turning the `ICU_CASE_MAPPINGS_ENABLED` flag off, users can get the old JVM-based case mappings. Note that the difference between the two is really subtle. ### How was this patch tested? Existing tests, with extended `CollationSupport` unit tests for InitCap to verify both ICU and JVM behaviour. ### Was this patch authored or co-authored using generative AI tooling? No. Closes #47100 from uros-db/change-initcap. Authored-by: Uros Bojanic <[email protected]> Signed-off-by: Wenchen Fan <[email protected]>
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.
What changes were proposed in this pull request?
Update
Lower&UpperSpark expressions to use ICU case mappings for UTF8_BINARY collation, instead of the currently used JVM case mappings. This behaviour is put under theICU_CASE_MAPPINGS_ENABLEDflag in SQLConf, which istrueby default.Why are the changes needed?
To keep the consistency between collations - all collations shouls use ICU-based case mappings, including the UTF8_BINARY collation.
Does this PR introduce any user-facing change?
Yes, the behaviour of
lower&upperstring functions for UTF8_BINARY will now rely on ICU-based case mappings. However, by turning theICU_CASE_MAPPINGS_ENABLEDflag off, users can get the old JVM-based case mappings. Note that the difference between the two is really subtle.How was this patch tested?
Existing tests, with extended
CollationSupportunit tests for Lower/Upper to verify both ICU and JVM behaviour.Was this patch authored or co-authored using generative AI tooling?
No.