This is an automated email from the ASF dual-hosted git repository.
alamb pushed a commit to branch branch-50
in repository https://gitbox.apache.org/repos/asf/datafusion.git
The following commit(s) were added to refs/heads/branch-50 by this push:
new 14da942f81 feat: expose `udafs` and `udwfs` methods on
`FunctionRegistry` (#17650) (#17725)
14da942f81 is described below
commit 14da942f81153d7f10644fba16add6b298e5b536
Author: Marko Milenković <[email protected]>
AuthorDate: Mon Sep 22 18:19:44 2025 +0100
feat: expose `udafs` and `udwfs` methods on `FunctionRegistry` (#17650)
(#17725)
* expose udafs and udwfs method on `FunctionRegistry`
* fix doc test
* add default implementations not to trigger
backward incompatible change for others
---
datafusion/core/src/execution/context/mod.rs | 8 ++++++++
datafusion/core/src/execution/session_state.rs | 8 ++++++++
datafusion/execution/src/task.rs | 8 ++++++++
datafusion/expr/src/registry.rs | 26 +++++++++++++++++++++++++-
datafusion/proto/src/bytes/mod.rs | 8 ++++++++
datafusion/proto/src/bytes/registry.rs | 8 ++++++++
datafusion/spark/src/lib.rs | 2 ++
7 files changed, 67 insertions(+), 1 deletion(-)
diff --git a/datafusion/core/src/execution/context/mod.rs
b/datafusion/core/src/execution/context/mod.rs
index de1d40dda3..fb5e2e0223 100644
--- a/datafusion/core/src/execution/context/mod.rs
+++ b/datafusion/core/src/execution/context/mod.rs
@@ -1727,6 +1727,14 @@ impl FunctionRegistry for SessionContext {
) -> Result<()> {
self.state.write().register_expr_planner(expr_planner)
}
+
+ fn udafs(&self) -> HashSet<String> {
+ self.state.read().udafs()
+ }
+
+ fn udwfs(&self) -> HashSet<String> {
+ self.state.read().udwfs()
+ }
}
/// Create a new task context instance from SessionContext
diff --git a/datafusion/core/src/execution/session_state.rs
b/datafusion/core/src/execution/session_state.rs
index a7b3bdeeac..38212167f9 100644
--- a/datafusion/core/src/execution/session_state.rs
+++ b/datafusion/core/src/execution/session_state.rs
@@ -1875,6 +1875,14 @@ impl FunctionRegistry for SessionState {
self.expr_planners.push(expr_planner);
Ok(())
}
+
+ fn udafs(&self) -> HashSet<String> {
+ self.aggregate_functions.keys().cloned().collect()
+ }
+
+ fn udwfs(&self) -> HashSet<String> {
+ self.window_functions.keys().cloned().collect()
+ }
}
impl OptimizerConfig for SessionState {
diff --git a/datafusion/execution/src/task.rs b/datafusion/execution/src/task.rs
index b11596c4a3..19f97f9e79 100644
--- a/datafusion/execution/src/task.rs
+++ b/datafusion/execution/src/task.rs
@@ -201,6 +201,14 @@ impl FunctionRegistry for TaskContext {
fn expr_planners(&self) -> Vec<Arc<dyn ExprPlanner>> {
vec![]
}
+
+ fn udafs(&self) -> HashSet<String> {
+ self.aggregate_functions.keys().cloned().collect()
+ }
+
+ fn udwfs(&self) -> HashSet<String> {
+ self.window_functions.keys().cloned().collect()
+ }
}
#[cfg(test)]
diff --git a/datafusion/expr/src/registry.rs b/datafusion/expr/src/registry.rs
index 4eb49710bc..8ea9e34dac 100644
--- a/datafusion/expr/src/registry.rs
+++ b/datafusion/expr/src/registry.rs
@@ -27,9 +27,25 @@ use std::sync::Arc;
/// A registry knows how to build logical expressions out of user-defined
function' names
pub trait FunctionRegistry {
- /// Set of all available udfs.
+ /// Returns names of all available scalar user defined functions.
fn udfs(&self) -> HashSet<String>;
+ /// Returns names of all available aggregate user defined functions.
+ fn udafs(&self) -> HashSet<String> {
+ // This default implementation is provided temporarily
+ // to maintain backward compatibility for the 50.1 release.
+ // It will be reverted to a required method in future versions.
+ HashSet::default()
+ }
+
+ /// Returns names of all available window user defined functions.
+ fn udwfs(&self) -> HashSet<String> {
+ // This default implementation is provided temporarily
+ // to maintain backward compatibility for the 50.1 release.
+ // It will be reverted to a required method in future versions.
+ HashSet::default()
+ }
+
/// Returns a reference to the user defined scalar function (udf) named
/// `name`.
fn udf(&self, name: &str) -> Result<Arc<ScalarUDF>>;
@@ -200,4 +216,12 @@ impl FunctionRegistry for MemoryFunctionRegistry {
fn expr_planners(&self) -> Vec<Arc<dyn ExprPlanner>> {
vec![]
}
+
+ fn udafs(&self) -> HashSet<String> {
+ self.udafs.keys().cloned().collect()
+ }
+
+ fn udwfs(&self) -> HashSet<String> {
+ self.udwfs.keys().cloned().collect()
+ }
}
diff --git a/datafusion/proto/src/bytes/mod.rs
b/datafusion/proto/src/bytes/mod.rs
index da01d89c0c..16d65c419a 100644
--- a/datafusion/proto/src/bytes/mod.rs
+++ b/datafusion/proto/src/bytes/mod.rs
@@ -170,6 +170,14 @@ impl Serializeable for Expr {
fn expr_planners(&self) -> Vec<Arc<dyn ExprPlanner>> {
vec![]
}
+
+ fn udafs(&self) -> std::collections::HashSet<String> {
+ std::collections::HashSet::default()
+ }
+
+ fn udwfs(&self) -> std::collections::HashSet<String> {
+ std::collections::HashSet::default()
+ }
}
Expr::from_bytes_with_registry(&bytes, &PlaceHolderRegistry)?;
diff --git a/datafusion/proto/src/bytes/registry.rs
b/datafusion/proto/src/bytes/registry.rs
index eae2425f8a..5d46d41f79 100644
--- a/datafusion/proto/src/bytes/registry.rs
+++ b/datafusion/proto/src/bytes/registry.rs
@@ -59,4 +59,12 @@ impl FunctionRegistry for NoRegistry {
fn expr_planners(&self) -> Vec<Arc<dyn ExprPlanner>> {
vec![]
}
+
+ fn udafs(&self) -> HashSet<String> {
+ HashSet::new()
+ }
+
+ fn udwfs(&self) -> HashSet<String> {
+ HashSet::new()
+ }
}
diff --git a/datafusion/spark/src/lib.rs b/datafusion/spark/src/lib.rs
index 531883a6c4..bec7d90062 100644
--- a/datafusion/spark/src/lib.rs
+++ b/datafusion/spark/src/lib.rs
@@ -53,6 +53,8 @@
//! # impl FunctionRegistry for SessionContext {
//! # fn register_udf(&mut self, _udf: Arc<ScalarUDF>) ->
Result<Option<Arc<ScalarUDF>>> { Ok (None) }
//! # fn udfs(&self) -> HashSet<String> { unimplemented!() }
+//! # fn udafs(&self) -> HashSet<String> { unimplemented!() }
+//! # fn udwfs(&self) -> HashSet<String> { unimplemented!() }
//! # fn udf(&self, _name: &str) -> Result<Arc<ScalarUDF>> {
unimplemented!() }
//! # fn udaf(&self, name: &str) -> Result<Arc<AggregateUDF>>
{unimplemented!() }
//! # fn udwf(&self, name: &str) -> Result<Arc<WindowUDF>> {
unimplemented!() }
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
