feat: Add support for abs (#2689)

Author: andygrove

.github/workflows/pr_build_linux.yml

@@ -134,6 +134,7 @@ jobs:
               org.apache.comet.CometCastSuite
               org.apache.comet.CometExpressionSuite
               org.apache.comet.CometExpressionCoverageSuite
+              org.apache.comet.CometMathExpressionSuite
               org.apache.comet.CometNativeSuite
               org.apache.comet.CometSparkSessionExtensionsSuite
               org.apache.comet.CometStringExpressionSuite

.github/workflows/pr_build_macos.yml

@@ -99,6 +99,7 @@ jobs:
               org.apache.comet.CometCastSuite
               org.apache.comet.CometExpressionSuite
               org.apache.comet.CometExpressionCoverageSuite
+              org.apache.comet.CometMathExpressionSuite
               org.apache.comet.CometNativeSuite
               org.apache.comet.CometSparkSessionExtensionsSuite
               org.apache.comet.CometStringExpressionSuite

docs/source/user-guide/latest/configs.md

@@ -164,6 +164,7 @@ These settings can be used to determine which parts of the plan are accelerated
 <!--BEGIN:CONFIG_TABLE[enable_expr]-->
 | Config | Description | Default Value |
 |--------|-------------|---------------|
+| `spark.comet.expression.Abs.enabled` | Enable Comet acceleration for `Abs` | true |
 | `spark.comet.expression.Acos.enabled` | Enable Comet acceleration for `Acos` | true |
 | `spark.comet.expression.Add.enabled` | Enable Comet acceleration for `Add` | true |
 | `spark.comet.expression.Alias.enabled` | Enable Comet acceleration for `Alias` | true |

docs/source/user-guide/latest/expressions.md

@@ -119,6 +119,7 @@ incompatible expressions.
 
 | Expression     | SQL       | Spark-Compatible? | Compatibility Notes               |
 |----------------|-----------|-------------------|-----------------------------------|
+| Abs            | `abs`     | Yes               |                                   |
 | Acos           | `acos`    | Yes               |                                   |
 | Add            | `+`       | Yes               |                                   |
 | Asin           | `asin`    | Yes               |                                   |

native/core/src/execution/planner.rs

@@ -674,19 +674,6 @@ impl PhysicalPlanner {
                 let op = DataFusionOperator::BitwiseShiftLeft;
                 Ok(Arc::new(BinaryExpr::new(left, op, right)))
             }
-            // https://github.com/apache/datafusion-comet/issues/666
-            // ExprStruct::Abs(expr) => {
-            //     let child = self.create_expr(expr.child.as_ref().unwrap(), Arc::clone(&input_schema))?;
-            //     let return_type = child.data_type(&input_schema)?;
-            //     let args = vec![child];
-            //     let eval_mode = from_protobuf_eval_mode(expr.eval_mode)?;
-            //     let comet_abs = Arc::new(ScalarUDF::new_from_impl(Abs::new(
-            //         eval_mode,
-            //         return_type.to_string(),
-            //     )?));
-            //     let expr = ScalarFunctionExpr::new("abs", comet_abs, args, return_type);
-            //     Ok(Arc::new(expr))
-            // }
             ExprStruct::CaseWhen(case_when) => {
                 let when_then_pairs = case_when
                     .when

native/proto/src/proto/expr.proto

@@ -70,7 +70,6 @@ message Expr {
     IfExpr if = 44;
     NormalizeNaNAndZero normalize_nan_and_zero = 45;
     TruncTimestamp truncTimestamp = 47;
-    Abs abs = 49;
     Subquery subquery = 50;
     UnboundReference unbound = 51;
     BloomFilterMightContain bloom_filter_might_contain = 52;
@@ -351,11 +350,6 @@ message TruncTimestamp {
   string timezone = 3;
 }
 
-message Abs {
-  Expr child = 1;
-  EvalMode eval_mode = 2;
-}
-
 message Subquery {
   int64 id = 1;
   DataType datatype = 2;

native/spark-expr/src/comet_scalar_funcs.rs

@@ -16,6 +16,7 @@
 // under the License.
 
 use crate::hash_funcs::*;
+use crate::math_funcs::abs::abs;
 use crate::math_funcs::checked_arithmetic::{checked_add, checked_div, checked_mul, checked_sub};
 use crate::math_funcs::modulo_expr::spark_modulo;
 use crate::{
@@ -180,6 +181,10 @@ pub fn create_comet_physical_fun_with_eval_mode(
             let func = Arc::new(spark_modulo);
             make_comet_scalar_udf!("spark_modulo", func, without data_type, fail_on_error)
         }
+        "abs" => {
+            let func = Arc::new(abs);
+            make_comet_scalar_udf!("abs", func, without data_type)
+        }
         _ => registry.udf(fun_name).map_err(|e| {
             DataFusionError::Execution(format!(
                 "Function {fun_name} not found in the registry: {e}",