Skip to content

Support runend canonicalization for lists #9708

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

Merged
danking merged 2 commits into from
Sep 3, 2026
+169 −0
Merged

Support runend canonicalization for lists #9708

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
692c7e6
Support runend canonicalization for lists
danking Aug 31, 2026
d7ddd10
Push runend list metadata into listview slots
danking Sep 3, 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
169 changes: 169 additions & 0 deletions encodings/runend/src/array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,8 +23,11 @@ use vortex_array::TypedArrayRef;
use vortex_array::VortexSessionExecute;
use vortex_array::array_slots;
use vortex_array::arrays::DecimalArray;
use vortex_array::arrays::ListViewArray;
use vortex_array::arrays::Primitive;
use vortex_array::arrays::PrimitiveArray;
use vortex_array::arrays::VarBinViewArray;
use vortex_array::arrays::listview::ListViewArraySlotsExt;
use vortex_array::buffer::BufferHandle;
use vortex_array::dtype::DType;
use vortex_array::dtype::Nullability;
Expand Down Expand Up @@ -502,26 +505,72 @@ pub(super) fn run_end_canonicalize(
.execute_as::<VarBinViewArray>("values", ctx)?;
runend_decode_varbinview(pends, values, array.offset(), array.len(), ctx)?.into_array()
}
DType::List(..) => {
let values = array
.values()
.clone()
.execute_as::<ListViewArray>("values", ctx)?;
runend_decode_listview(pends, values, array.offset(), array.len())?.into_array()
}
_ => vortex_bail!("Unsupported RunEnd value type: {}", array.dtype()),
})
}

fn runend_decode_listview(
ends: PrimitiveArray,
values: ListViewArray,
offset: usize,
length: usize,
) -> VortexResult<ListViewArray> {
let validity = match values.validity()? {
Validity::NonNullable => Validity::NonNullable,
Validity::AllValid => Validity::AllValid,
Validity::AllInvalid => Validity::AllInvalid,
Validity::Array(validity) => Validity::Array(unsafe {
RunEnd::new_unchecked(ends.clone().into_array(), validity, offset, length).into_array()
}),
};

// SAFETY: the `RunEndArray`s re-express valid per-run ListView metadata over the logical output
// length. The original `elements` child is reused, so every view still points at a valid range.
Ok(unsafe {
ListViewArray::new_unchecked(
values.elements().clone(),
RunEnd::new_unchecked(
ends.clone().into_array(),
values.offsets().clone(),
offset,
length,
)
.into_array(),
RunEnd::new_unchecked(ends.into_array(), values.sizes().clone(), offset, length)
.into_array(),
validity,
)
})
}

#[cfg(test)]
mod tests {
use std::sync::Arc;
use std::sync::LazyLock;

use vortex_array::IntoArray;
use vortex_array::VortexSessionExecute;
use vortex_array::arrays::DecimalArray;
use vortex_array::arrays::DictArray;
use vortex_array::arrays::ListArray;
use vortex_array::arrays::ListViewArray;
use vortex_array::arrays::VarBinViewArray;
use vortex_array::arrays::listview::ListViewArraySlotsExt;
use vortex_array::assert_arrays_eq;
use vortex_array::builders::VarBinBuilder;
use vortex_array::dtype::DType;
use vortex_array::dtype::DecimalDType;
use vortex_array::dtype::Nullability;
use vortex_array::dtype::PType;
use vortex_array::dtype::i256;
use vortex_array::validity::Validity;
use vortex_buffer::buffer;
use vortex_error::VortexResult;
use vortex_session::VortexSession;
Expand Down Expand Up @@ -605,6 +654,37 @@ mod tests {
assert_arrays_eq!(arr.into_array(), expected, &mut ctx);
}

#[test]
fn test_runend_list_i64() {
let mut ctx = SESSION.create_execution_ctx();
let values = ListArray::from_iter_slow::<u32, _>(
vec![vec![1i64, 2], vec![3], vec![4, 5, 6]],
Arc::new(DType::Primitive(PType::I64, Nullability::NonNullable)),
)
.unwrap()
.into_array();
let arr = RunEnd::new(buffer![2u32, 5, 10].into_array(), values, &mut ctx);

let expected = ListArray::from_iter_slow::<u32, _>(
vec![
vec![1i64, 2],
vec![1, 2],
vec![3],
vec![3],
vec![3],
vec![4, 5, 6],
vec![4, 5, 6],
vec![4, 5, 6],
vec![4, 5, 6],
vec![4, 5, 6],
],
Arc::new(DType::Primitive(PType::I64, Nullability::NonNullable)),
)
.unwrap()
.into_array();
assert_arrays_eq!(arr.into_array(), expected, &mut ctx);
}

#[test]
fn test_runend_nullable_decimal() {
let mut ctx = SESSION.create_execution_ctx();
Expand Down Expand Up @@ -638,6 +718,95 @@ mod tests {
assert_arrays_eq!(arr.into_array(), expected, &mut ctx);
}

#[test]
fn test_runend_list_bool() {
let mut ctx = SESSION.create_execution_ctx();
let values = ListArray::from_iter_slow::<u32, _>(
vec![vec![true, false], vec![false], vec![true, true, false]],
Arc::new(DType::Bool(Nullability::NonNullable)),
)
.unwrap()
.into_array();
let arr = RunEnd::new(buffer![2u32, 5, 10].into_array(), values, &mut ctx);

let expected = ListArray::from_iter_slow::<u32, _>(
vec![
vec![true, false],
vec![true, false],
vec![false],
vec![false],
vec![false],
vec![true, true, false],
vec![true, true, false],
vec![true, true, false],
vec![true, true, false],
vec![true, true, false],
],
Arc::new(DType::Bool(Nullability::NonNullable)),
)
.unwrap()
.into_array();
assert_arrays_eq!(arr.into_array(), expected, &mut ctx);
}

#[test]
fn test_runend_list_utf8() {
let mut ctx = SESSION.create_execution_ctx();
let values = ListArray::try_new(
VarBinViewArray::from_iter_str(["a", "b", "c", "d", "e", "f"]).into_array(),
buffer![0u32, 2, 3, 6].into_array(),
Validity::NonNullable,
)
.unwrap()
.into_array();
let arr = RunEnd::new(buffer![2u32, 5, 10].into_array(), values, &mut ctx);

let expected = ListArray::try_new(
VarBinViewArray::from_iter_str([
"a", "b", "a", "b", "c", "c", "c", "d", "e", "f", "d", "e", "f", "d", "e", "f",
"d", "e", "f", "d", "e", "f",
])
.into_array(),
buffer![0u32, 2, 4, 5, 6, 7, 10, 13, 16, 19, 22].into_array(),
Validity::NonNullable,
)
.unwrap()
.into_array();
assert_arrays_eq!(arr.into_array(), expected, &mut ctx);
}

#[test]
fn test_runend_list_canonicalizes_to_runend_listview_slots() -> VortexResult<()> {
let mut ctx = SESSION.create_execution_ctx();
let values = ListArray::try_new(
buffer![1i64, 2, 3, 4, 5, 6].into_array(),
buffer![0u32, 2, 3, 6].into_array(),
Validity::from_iter([true, false, true]),
)?
.into_array();
let arr = RunEnd::try_new(buffer![2u32, 5, 6].into_array(), values, &mut ctx)?;

let listview = arr
.clone()
.into_array()
.execute::<ListViewArray>(&mut ctx)?;
assert!(listview.offsets().is::<RunEnd>());
assert!(listview.sizes().is::<RunEnd>());
match listview.validity()? {
Validity::Array(validity) => assert!(validity.is::<RunEnd>()),
validity => panic!("expected array-backed validity, got {validity:?}"),
}

let expected = ListArray::try_new(
buffer![1i64, 2, 1, 2, 3, 3, 3, 4, 5, 6].into_array(),
buffer![0u32, 2, 4, 5, 6, 7, 10].into_array(),
Validity::from_iter([true, true, false, false, false, true]),
)?
.into_array();
assert_arrays_eq!(arr.into_array(), expected, &mut ctx);
Ok(())
}

#[test]
fn test_runend_dict() {
let mut ctx = SESSION.create_execution_ctx();
Expand Down
Loading