[SPARK-59411][SQL] Support the TIME data type in the width_bucket function - #58706
Closed
stevomitric wants to merge 1 commit into
Closed
[SPARK-59411][SQL] Support the TIME data type in the width_bucket function#58706stevomitric wants to merge 1 commit into
stevomitric wants to merge 1 commit into
Conversation
…ction `width_bucket` accepted DOUBLE and interval (year-month / day-time) operands but rejected TIME. TIME is physically a Long (nanoseconds since midnight), so it slots into the existing double-based bucket math the same way the Long-backed day-time interval does: - Add `AnyTimeType` to the three value/min/max `TypeCollection`s. - Accept `(TIME, TIME, TIME)` in `checkInputDataTypes` (allowing mixed precisions, like the interval cases); operands are compared on the shared nanos-of-day value. No change to `computeBucketNumber` / eval / codegen: operands already flow through `Number.doubleValue()` (interpreted) and long->double widening (codegen). Tests: `MathExpressionsSuite` unit cases and golden cases in `time.sql` (bucket edges, reversed range, mixed precision, and null/degenerate inputs). Co-authored-by: Isaac <no-reply@databricks.com>
uros-b
approved these changes
Sep 11, 2026
Member
|
Thank you @stevomitric! |
uros-b
pushed a commit
that referenced
this pull request
Sep 11, 2026
…ction ### What changes were proposed in this pull request? Make `width_bucket` accept the TIME data type. It already supports DOUBLE and interval (year-month / day-time) operands. This PR adds TIME, which is physically a Long (nanoseconds since midnight) and reuses the existing double-based bucket math (no eval/codegen change). Mixed TIME precisions are allowed and compared on the shared nanos-of-day value. ### Why are the changes needed? DATE/TIMESTAMP get equiwidth bucketing through intervals, but a TIME column could not be bucketed (e.g. bucketing event times across a 09:00–17:00 workday). ### Does this PR introduce _any_ user-facing change? Yes. `width_bucket(TIME'12:00:00', TIME'09:00:00', TIME'17:00:00', 8)` now returns 4 instead of raising an error. ### How was this patch tested? `MathExpressionsSuite` unit cases and golden cases in `time.sql` (bucket edges, reversed range, mixed precision, and null/degenerate inputs). ### Was this patch authored or co-authored using generative AI tooling? Co-Authored-By: Claude Opus 4.8 Closes #58706 from stevomitric/stevomitric/spark-widthbucket-time. Authored-by: Stevo Mitric <stevomitric2000@gmail.com> Signed-off-by: Uros Bojanic <221401595+uros-b@users.noreply.github.com> (cherry picked from commit efa63e6) Signed-off-by: Uros Bojanic <221401595+uros-b@users.noreply.github.com>
Member
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
2 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?
Make
width_bucketaccept the TIME data type. It already supports DOUBLE and interval (year-month / day-time) operands. This PR adds TIME, which is physically a Long (nanoseconds since midnight) and reuses the existing double-based bucket math (no eval/codegen change).Mixed TIME precisions are allowed and compared on the shared nanos-of-day value.
Why are the changes needed?
DATE/TIMESTAMP get equiwidth bucketing through intervals, but a TIME column could not be bucketed (e.g. bucketing event times across a 09:00–17:00 workday).
Does this PR introduce any user-facing change?
Yes.
width_bucket(TIME'12:00:00', TIME'09:00:00', TIME'17:00:00', 8)now returns 4 instead of raising an error.How was this patch tested?
MathExpressionsSuiteunit cases and golden cases intime.sql(bucket edges, reversed range, mixed precision, and null/degenerate inputs).Was this patch authored or co-authored using generative AI tooling?
Co-Authored-By: Claude Opus 4.8