feat: array-empty (#7313)

* feat: array-empty

* add definition in pbjson

* add definition

* remove useless tests

* add the function in proto

* add doc

* refactor

* format

* remove useless code

* fix format

* support NULL

* remove redundant code

* support NULL

* fix clippy

Author: Weijun-H

datafusion/expr/src/built_in_function.rs

@@ -138,6 +138,8 @@ pub enum BuiltinScalarFunction {
     ArrayDims,
     /// array_element
     ArrayElement,
+    /// array_empty
+    ArrayEmpty,
     /// array_length
     ArrayLength,
     /// array_ndims
@@ -360,6 +362,7 @@ impl BuiltinScalarFunction {
             BuiltinScalarFunction::Trunc => Volatility::Immutable,
             BuiltinScalarFunction::ArrayAppend => Volatility::Immutable,
             BuiltinScalarFunction::ArrayConcat => Volatility::Immutable,
+            BuiltinScalarFunction::ArrayEmpty => Volatility::Immutable,
             BuiltinScalarFunction::ArrayHasAll => Volatility::Immutable,
             BuiltinScalarFunction::ArrayHasAny => Volatility::Immutable,
             BuiltinScalarFunction::ArrayHas => Volatility::Immutable,
@@ -536,7 +539,8 @@ impl BuiltinScalarFunction {
             }
             BuiltinScalarFunction::ArrayHasAll
             | BuiltinScalarFunction::ArrayHasAny
-            | BuiltinScalarFunction::ArrayHas => Ok(Boolean),
+            | BuiltinScalarFunction::ArrayHas
+            | BuiltinScalarFunction::ArrayEmpty => Ok(Boolean),
             BuiltinScalarFunction::ArrayDims => {
                 Ok(List(Arc::new(Field::new("item", UInt64, true))))
             }
@@ -829,6 +833,7 @@ impl BuiltinScalarFunction {
                 Signature::variadic_any(self.volatility())
             }
             BuiltinScalarFunction::ArrayDims => Signature::any(1, self.volatility()),
+            BuiltinScalarFunction::ArrayEmpty => Signature::any(1, self.volatility()),
             BuiltinScalarFunction::ArrayElement => Signature::any(2, self.volatility()),
             BuiltinScalarFunction::Flatten => Signature::any(1, self.volatility()),
             BuiltinScalarFunction::ArrayHasAll
@@ -1319,6 +1324,7 @@ fn aliases(func: &BuiltinScalarFunction) -> &'static [&'static str] {
             &["array_concat", "array_cat", "list_concat", "list_cat"]
         }
         BuiltinScalarFunction::ArrayDims => &["array_dims", "list_dims"],
+        BuiltinScalarFunction::ArrayEmpty => &["empty"],
         BuiltinScalarFunction::ArrayElement => &[
             "array_element",
             "array_extract",

datafusion/expr/src/expr_fn.rs

@@ -552,6 +552,12 @@ scalar_expr!(
     first_array second_array,
     "returns true, if the element appears in the first array, otherwise false."
 );
+scalar_expr!(
+    ArrayEmpty,
+    array_empty,
+    array,
+    "returns 1 for an empty array or 0 for a non-empty array."
+);
 scalar_expr!(
     ArrayHasAll,
     array_has_all,

datafusion/physical-expr/src/array_expressions.rs

@@ -982,6 +982,21 @@ macro_rules! general_repeat_list {
     }};
 }
 
+/// Array_empty SQL function
+pub fn array_empty(args: &[ArrayRef]) -> Result<ArrayRef> {
+    println!("args[0]: {:?}", &args[0]);
+    if args[0].as_any().downcast_ref::<NullArray>().is_some() {
+        return Ok(args[0].clone());
+    }
+
+    let array = as_list_array(&args[0])?;
+    let builder = array
+        .iter()
+        .map(|arr| arr.map(|arr| arr.len() == arr.null_count()))
+        .collect::<BooleanArray>();
+    Ok(Arc::new(builder))
+}
+
 /// Array_repeat SQL function
 pub fn array_repeat(args: &[ArrayRef]) -> Result<ArrayRef> {
     let element = &args[0];

datafusion/physical-expr/src/functions.rs

@@ -425,6 +425,9 @@ pub fn create_physical_fun(
         BuiltinScalarFunction::ArrayConcat => {
             Arc::new(|args| make_scalar_function(array_expressions::array_concat)(args))
         }
+        BuiltinScalarFunction::ArrayEmpty => {
+            Arc::new(|args| make_scalar_function(array_expressions::array_empty)(args))
+        }
         BuiltinScalarFunction::ArrayHasAll => {
             Arc::new(|args| make_scalar_function(array_expressions::array_has_all)(args))
         }

datafusion/proto/proto/datafusion.proto

@@ -597,6 +597,7 @@ enum ScalarFunction {
   Flatten = 112;
   Isnan = 113;
   Iszero = 114;
+  ArrayEmpty = 115;
 }
 
 message ScalarFunctionNode {

datafusion/proto/src/generated/pbjson.rs

@@ -34,7 +34,6 @@ use datafusion_common::{
     internal_err, Column, DFField, DFSchema, DFSchemaRef, DataFusionError,
     OwnedTableReference, Result, ScalarValue,
 };
-use datafusion_expr::expr::{Alias, Placeholder};
 use datafusion_expr::{
     abs, acos, acosh, array, array_append, array_concat, array_dims, array_element,
     array_has, array_has_all, array_has_any, array_length, array_ndims, array_position,
@@ -59,6 +58,10 @@ use datafusion_expr::{
     JoinConstraint, JoinType, Like, Operator, TryCast, WindowFrame, WindowFrameBound,
     WindowFrameUnits,
 };
+use datafusion_expr::{
+    array_empty,
+    expr::{Alias, Placeholder},
+};
 use std::sync::Arc;
 
 #[derive(Debug)]
@@ -452,6 +455,7 @@ impl From<&protobuf::ScalarFunction> for BuiltinScalarFunction {
             ScalarFunction::ToTimestamp => Self::ToTimestamp,
             ScalarFunction::ArrayAppend => Self::ArrayAppend,
             ScalarFunction::ArrayConcat => Self::ArrayConcat,
+            ScalarFunction::ArrayEmpty => Self::ArrayEmpty,
             ScalarFunction::ArrayHasAll => Self::ArrayHasAll,
             ScalarFunction::ArrayHasAny => Self::ArrayHasAny,
             ScalarFunction::ArrayHas => Self::ArrayHas,
@@ -1355,6 +1359,9 @@ pub fn parse_expr(
                     parse_expr(&args[0], registry)?,
                     parse_expr(&args[1], registry)?,
                 )),
+                ScalarFunction::ArrayEmpty => {
+                    Ok(array_empty(parse_expr(&args[0], registry)?))
+                }
                 ScalarFunction::ArrayNdims => {
                     Ok(array_ndims(parse_expr(&args[0], registry)?))
                 }

datafusion/proto/src/generated/prost.rs

@@ -1451,6 +1451,7 @@ impl TryFrom<&BuiltinScalarFunction> for protobuf::ScalarFunction {
             BuiltinScalarFunction::ToTimestamp => Self::ToTimestamp,
             BuiltinScalarFunction::ArrayAppend => Self::ArrayAppend,
             BuiltinScalarFunction::ArrayConcat => Self::ArrayConcat,
+            BuiltinScalarFunction::ArrayEmpty => Self::ArrayEmpty,
             BuiltinScalarFunction::ArrayHasAll => Self::ArrayHasAll,
             BuiltinScalarFunction::ArrayHasAny => Self::ArrayHasAny,
             BuiltinScalarFunction::ArrayHas => Self::ArrayHas,

datafusion/proto/src/logical_plan/from_proto.rs

@@ -2363,6 +2363,42 @@ from flatten_table;
 [1, 2, 3] [1, 2, 3, 4, 5, 6] [1, 2, 3] [1.0, 2.1, 2.2, 3.2, 3.3, 3.4]
 [1, 2, 3, 4, 5, 6] [8] [1, 2, 3] [1.0, 2.0, 3.0, 4.0, 5.0, 6.0]
 
+# empty scalar function #1
+query B
+select empty(make_array(1));
+----
+false
+
+# empty scalar function #2
+query B
+select empty(make_array());
+----
+true
+
+# empty scalar function #3
+query B
+select empty(make_array(NULL));
+----
+true
+
+# empty scalar function #4
+query B
+select empty(NULL);
+----
+NULL
+
+# empty scalar function #5
+query B
+select empty(column1) from arrays;
+----
+false
+false
+false
+false
+NULL
+false
+false
+
 ### Delete tables
 
 statement ok

datafusion/proto/src/logical_plan/to_proto.rs

@@ -1495,6 +1495,7 @@ from_unixtime(expression)
 - [array_slice](#array_slice)
 - [array_to_string](#array_to_string)
 - [cardinality](#cardinality)
+- [empty](#empty)
 - [list_append](#list_append)
 - [list_cat](#list_cat)
 - [list_concat](#list_concat)
@@ -1693,6 +1694,8 @@ array_element(array, index)
 - list_element
 - list_extract
 
+### `array_empty`
+
 ### `array_extract`
 
 _Alias of [array_element](#array_element)._
@@ -2188,6 +2191,30 @@ cardinality(array)
 +--------------------------------------+
 ```
 
+### `empty`
+
+Returns 1 for an empty array or 0 for a non-empty array.
+
+```
+empty(array)
+```
+
+#### Arguments
+
+- **array**: Array expression.
+  Can be a constant, column, or function, and any combination of array operators.
+
+#### Example
+
+```
+❯ select empty([1]);
++------------------+
+| empty(List([1])) |
++------------------+
+| 0                |
++------------------+
+```
+
 ### `list_append`
 
 _Alias of [array_append](#array_append)._