Skip to content

fix(proto): preserve empty projection when ser/de MemoryScanExec #24087

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
adriangb merged 2 commits into from
Aug 5, 2026
+31 −15
Merged

fix(proto): preserve empty projection when ser/de MemoryScanExec #24087

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d8fe64d
preserve empty projection
buraksenn Aug 4, 2026
6e5213f
use roundtrip_test_sql_with_context in test
buraksenn Aug 4, 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
27 changes: 12 additions & 15 deletions datafusion/proto/src/physical_plan/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1179,15 +1179,11 @@ pub trait PhysicalPlanNodeExt: Sized {
})?;
let schema: SchemaRef = SchemaRef::new(proto_schema.try_into()?);

let projection = if !scan.projection.is_empty() {
Some(
scan.projection
.iter()
.map(|i| *i as usize)
.collect::<Vec<_>>(),
)
} else {
None
// Preserve the empty-projection sentinel written by `try_from_data_source_exec`.
let projection = match scan.projection.as_slice() {
[] => None,
[u32::MAX] => Some(Vec::new()),
indices => Some(indices.iter().map(|i| *i as usize).collect()),
};

let mut sort_information = vec![];
Expand Down Expand Up @@ -2364,12 +2360,13 @@ pub trait PhysicalPlanNodeExt: Sized {
let proto_schema: protobuf::Schema =
source_conf.original_schema().as_ref().try_into()?;

let proto_projection = source_conf
.projection()
.as_ref()
.map_or_else(Vec::new, |v| {
v.iter().map(|x| *x as u32).collect::<Vec<u32>>()
});
// Proto3 can't tell `None` from `Some(vec![])`; encode the latter
// as the `[u32::MAX]` sentinel, matching the join/filter nodes.
let proto_projection = match source_conf.projection().as_ref() {
None => Vec::new(),
Some(v) if v.is_empty() => vec![u32::MAX],
Some(v) => v.iter().map(|x| *x as u32).collect(),
};
Comment on lines +2363 to +2369

@adriangb adriangb Aug 5, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Note this is a wire format change but it only impacts the already broken empty projection case, so it seems justified.


let proto_sort_information = source_conf
.sort_information()
Expand Down
19 changes: 19 additions & 0 deletions datafusion/proto/tests/cases/roundtrip_physical_plan.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2678,6 +2678,25 @@ async fn roundtrip_empty_projection() -> Result<()> {
roundtrip_test_sql_with_context(sql, &ctx).await
}

#[tokio::test]
async fn roundtrip_memory_source_empty_projection() -> Result<()> {
// Memory scan: `Some(vec![])` must not decode back as `None`
let ctx = SessionContext::new();
let batch = RecordBatch::try_new(
Arc::new(Schema::new(vec![
Field::new("a", DataType::Utf8, false),
Field::new("b", DataType::Int64, false),
])),
vec![
Arc::new(arrow::array::StringArray::from(vec!["Tom"])),
Arc::new(arrow::array::Int64Array::from(vec![18i64])),
],
)?;
ctx.register_batch("tmem", batch)?;
let sql = "select 1 from tmem";
roundtrip_test_sql_with_context(sql, &ctx).await
}

#[tokio::test]
async fn roundtrip_physical_plan_node() {
use datafusion::prelude::*;
Expand Down
Loading