Skip to content

fix: Support scalar inputs for datetime expressions when constant folding is disabled #3448

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.

Sign up for GitHub

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

Open
0lai0 wants to merge 6 commits into
base: main
Choose a base branch
from
Open

fix: Support scalar inputs for datetime expressions when constant folding is disabled #3448

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
c3fc264
Native engine panics on all-scalar inputs for hour, minute, second, u…
0lai0 Feb 9, 2026
5bcba2f
add test
0lai0 Feb 9, 2026
07c7273
remove int32array
0lai0 Feb 9, 2026
566c981
fix fmt and test
0lai0 Feb 10, 2026
517d6a8
fix test assert
0lai0 Feb 10, 2026
a6b383f
remove unused import
0lai0 Feb 10, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 18 additions & 4 deletions native/spark-expr/src/datetime_funcs/extract_date_part.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@
use crate::utils::array_with_timezone;
use arrow::compute::{date_part, DatePart};
use arrow::datatypes::{DataType, TimeUnit::Microsecond};
use datafusion::common::{internal_datafusion_err, DataFusionError};
use datafusion::common::internal_datafusion_err;
use datafusion::logical_expr::{
ColumnarValue, ScalarFunctionArgs, ScalarUDFImpl, Signature, Volatility,
};
Expand Down Expand Up @@ -86,9 +86,23 @@ macro_rules! extract_date_part {
let result = date_part(&array, DatePart::$date_part_variant)?;
Ok(ColumnarValue::Array(result))
}
_ => Err(DataFusionError::Execution(
concat!($fn_name, "(scalar) should be fold in Spark JVM side.").to_string(),
)),
[ColumnarValue::Scalar(scalar)] => {
// When Spark's ConstantFolding is disabled, literal-only expressions like
// hour can reach the native engine as scalar inputs.
// Instead of failing and requiring JVM folding, we evaluate the scalar
// natively by broadcasting it to a single-element array.
let array = scalar.clone().to_array_of_size(1)?;
let array = array_with_timezone(
array,
self.timezone.clone(),
Some(&DataType::Timestamp(
Microsecond,
Some(self.timezone.clone().into()),
)),
)?;
let result = date_part(&array, DatePart::$date_part_variant)?;
Ok(ColumnarValue::Array(result))
}
}
}

Expand Down
184 changes: 109 additions & 75 deletions native/spark-expr/src/datetime_funcs/unix_timestamp.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,10 +16,10 @@
// under the License.

use crate::utils::array_with_timezone;
use arrow::array::{Array, AsArray, PrimitiveArray};
use arrow::array::{Array, AsArray, Int64Array, PrimitiveArray};
use arrow::compute::cast;
use arrow::datatypes::{DataType, Int64Type, TimeUnit::Microsecond};
use datafusion::common::{internal_datafusion_err, DataFusionError};
use datafusion::common::{internal_datafusion_err, DataFusionError, ScalarValue};
use datafusion::logical_expr::{
ColumnarValue, ScalarFunctionArgs, ScalarUDFImpl, Signature, Volatility,
};
Expand Down Expand Up @@ -77,79 +77,37 @@ impl ScalarUDFImpl for SparkUnixTimestamp {
.map_err(|_| internal_datafusion_err!("unix_timestamp expects exactly one argument"))?;

match args {
[ColumnarValue::Array(array)] => match array.data_type() {
DataType::Timestamp(_, _) => {
let is_utc = self.timezone == "UTC";
let array = if is_utc
&& matches!(array.data_type(), DataType::Timestamp(Microsecond, Some(tz)) if tz.as_ref() == "UTC")
{
array
} else {
array_with_timezone(
array,
self.timezone.clone(),
Some(&DataType::Timestamp(Microsecond, Some("UTC".into()))),
)?
};

let timestamp_array =
array.as_primitive::<arrow::datatypes::TimestampMicrosecondType>();

let result: PrimitiveArray<Int64Type> = if timestamp_array.null_count() == 0 {
timestamp_array
.values()
.iter()
.map(|&micros| micros / MICROS_PER_SECOND)
.collect()
} else {
timestamp_array
.iter()
.map(|v| v.map(|micros| div_floor(micros, MICROS_PER_SECOND)))
.collect()
};

Ok(ColumnarValue::Array(Arc::new(result)))
}
DataType::Date32 => {
let timestamp_array = cast(&array, &DataType::Timestamp(Microsecond, None))?;

let is_utc = self.timezone == "UTC";
let array = if is_utc {
timestamp_array
} else {
array_with_timezone(
timestamp_array,
self.timezone.clone(),
Some(&DataType::Timestamp(Microsecond, Some("UTC".into()))),
)?
};

let timestamp_array =
array.as_primitive::<arrow::datatypes::TimestampMicrosecondType>();

let result: PrimitiveArray<Int64Type> = if timestamp_array.null_count() == 0 {
timestamp_array
.values()
.iter()
.map(|&micros| micros / MICROS_PER_SECOND)
.collect()
} else {
timestamp_array
.iter()
.map(|v| v.map(|micros| div_floor(micros, MICROS_PER_SECOND)))
.collect()
};

Ok(ColumnarValue::Array(Arc::new(result)))
}
_ => Err(DataFusionError::Execution(format!(
"unix_timestamp does not support input type: {:?}",
array.data_type()
))),
},
_ => Err(DataFusionError::Execution(
"unix_timestamp(scalar) should be fold in Spark JVM side.".to_string(),
)),
[ColumnarValue::Array(array)] => self.eval_array(&array),
[ColumnarValue::Scalar(scalar)] => {
// When Spark's ConstantFolding is disabled, literal-only expressions like
// unix_timestamp can reach the native engine
// as scalar inputs. Evaluate the scalar natively by broadcasting it to a
// single-element array and converting the result back to a scalar.
let array = scalar.clone().to_array_of_size(1)?;
let result = self.eval_array(&array)?;

let result_array = match result {
ColumnarValue::Array(array) => array,
ColumnarValue::Scalar(_) => {
return Err(DataFusionError::Internal(
"unix_timestamp: expected array result from eval_array".to_string(),
))
}
};

let int64_array = result_array
.as_any()
.downcast_ref::<Int64Array>()
.expect("unix_timestamp should return Int64Array");

let scalar_result = if int64_array.is_null(0) {
ScalarValue::Int64(None)
} else {
ScalarValue::Int64(Some(int64_array.value(0)))
};

Ok(ColumnarValue::Scalar(scalar_result))
}
}
}

Expand All @@ -158,6 +116,82 @@ impl ScalarUDFImpl for SparkUnixTimestamp {
}
}

impl SparkUnixTimestamp {
fn eval_array(&self, array: &Arc<dyn Array>) -> datafusion::common::Result<ColumnarValue> {
match array.data_type() {
DataType::Timestamp(_, _) => {
let is_utc = self.timezone == "UTC";
let array = if is_utc
&& matches!(array.data_type(), DataType::Timestamp(Microsecond, Some(tz)) if tz.as_ref() == "UTC")
{
Arc::clone(array)
} else {
array_with_timezone(
Arc::clone(array),
self.timezone.clone(),
Some(&DataType::Timestamp(Microsecond, Some("UTC".into()))),
)?
};

let timestamp_array =
array.as_primitive::<arrow::datatypes::TimestampMicrosecondType>();

let result: PrimitiveArray<Int64Type> = if timestamp_array.null_count() == 0 {
timestamp_array
.values()
.iter()
.map(|&micros| micros / MICROS_PER_SECOND)
.collect()
} else {
timestamp_array
.iter()
.map(|v| v.map(|micros| div_floor(micros, MICROS_PER_SECOND)))
.collect()
};

Ok(ColumnarValue::Array(Arc::new(result)))
}
DataType::Date32 => {
let timestamp_array =
cast(array.as_ref(), &DataType::Timestamp(Microsecond, None))?;

let is_utc = self.timezone == "UTC";
let array = if is_utc {
timestamp_array
} else {
array_with_timezone(
timestamp_array,
self.timezone.clone(),
Some(&DataType::Timestamp(Microsecond, Some("UTC".into()))),
)?
};

let timestamp_array =
array.as_primitive::<arrow::datatypes::TimestampMicrosecondType>();

let result: PrimitiveArray<Int64Type> = if timestamp_array.null_count() == 0 {
timestamp_array
.values()
.iter()
.map(|&micros| micros / MICROS_PER_SECOND)
.collect()
} else {
timestamp_array
.iter()
.map(|v| v.map(|micros| div_floor(micros, MICROS_PER_SECOND)))
.collect()
};

Ok(ColumnarValue::Array(Arc::new(result)))
}
_ => Err(DataFusionError::Execution(format!(
"unix_timestamp does not support input type: {:?}",
array.data_type()
))),
}
}
}

#[cfg(test)]
mod tests {
use super::*;
Expand Down
21 changes: 21 additions & 0 deletions spark/src/test/scala/org/apache/comet/CometExpressionSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -618,6 +618,27 @@ class CometExpressionSuite extends CometTestBase with AdaptiveSparkPlanHelper {
}
}

test("scalar hour/minute/second/unix_timestamp with ConstantFolding disabled") {
Copy link
Contributor

@hsiang-c hsiang-c Feb 9, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you for the contribution! Let's migrate these tests to SQL File Tests, see https://datafusion.apache.org/comet/contributor-guide/sql-file-tests.html#sql-file-tests

withSQLConf(
SQLConf.OPTIMIZER_EXCLUDED_RULES.key ->
"org.apache.spark.sql.catalyst.optimizer.ConstantFolding") {
val df = spark.sql("""
|SELECT
| hour(timestamp('2026-04-18 04:18:45')) AS h,
| minute(timestamp('2026-04-18 04:18:45')) AS m,
| second(timestamp('2026-04-18 04:18:45')) AS s,
| unix_timestamp(timestamp('2020-01-01 00:00:00')) AS u
|""".stripMargin)

val Row(h: Int, m: Int, s: Int, u: Long) = df.head()

assert(h == 4)
assert(m == 18)
assert(s == 45)
assert(u == 1577836800L)
}
}

test("hour on int96 timestamp column") {
import testImplicits._

Expand Down
Loading