ISSUE-4264: simd selection

common/datavalues/src/columns/primitive/mod.rs

@@ -189,17 +189,32 @@ impl<T: PrimitiveType> Column for PrimitiveColumn<T> {
         if length == self.len() {
             return Arc::new(self.clone());
         }
-        let iter = self
-            .values()
-            .iter()
-            .zip(filter.values().iter())
-            .filter(|(_, f)| *f)
-            .map(|(v, _)| *v);
 
-        let values: Vec<T> = iter.collect();
-        let col = PrimitiveColumn {
-            values: values.into(),
-        };
+        let mut res = Vec::<T>::with_capacity(length);
+        let mut offset = 0;
+        let values = self.values();
+
+        const MASK_BITS: usize = 64;
+        for mut mask in filter.values().chunks::<u64>() {
+            if mask == u64::MAX {
+                res.extend(&values[offset..offset + MASK_BITS]);
+            } else {
+                while mask != 0 {
+                    let n = std::intrinsics::cttz(mask) as usize;
+                    res.push(values[offset + n]);
+                    mask = mask & (mask - 1);
+                }
+            }
+            offset += MASK_BITS;
+        }
+
+        for i in offset..self.len() {
+            if filter.get_data(i) {
+                res.push(values[i]);
+            }
+        }
+
+        let col = PrimitiveColumn { values: res.into() };
 
         Arc::new(col)
     }

common/datavalues/src/lib.rs

@@ -18,6 +18,7 @@
 
 #![feature(generic_associated_types)]
 #![feature(trusted_len)]
+#![feature(core_intrinsics)]
 
 #[macro_use]
 mod macros;

common/datavalues/tests/it/columns/primitive.rs

@@ -54,3 +54,48 @@ fn test_const_column() {
     let c = ConstColumn::new(Series::from_data(vec![PI]), 24).arc();
     println!("{:?}", c);
 }
+
+#[test]
+fn test_filter_column() {
+    const N: usize = 1000;
+    let it = (0..N).map(|i| i as i32);
+    let data_column: PrimitiveColumn<i32> = Int32Column::from_iterator(it);
+
+    struct Test {
+        filter: BooleanColumn,
+        expect: Vec<i32>,
+    }
+
+    let tests: Vec<Test> = vec![
+        Test {
+            filter: BooleanColumn::from_iterator((0..N).map(|_| true)),
+            expect: (0..N).map(|i| i as i32).collect(),
+        },
+        Test {
+            filter: BooleanColumn::from_iterator((0..N).map(|_| false)),
+            expect: vec![],
+        },
+        Test {
+            filter: BooleanColumn::from_iterator((0..N).map(|i| i % 10 == 0)),
+            expect: (0..N).map(|i| i as i32).filter(|i| i % 10 == 0).collect(),
+        },
+        Test {
+            filter: BooleanColumn::from_iterator((0..N).map(|i| i < 100 || i > 800)),
+            expect: (0..N)
+                .map(|i| i as i32)
+                .filter(|&i| i < 100 || i > 800)
+                .collect(),
+        },
+    ];
+
+    for test in tests {
+        let res = data_column.filter(&test.filter);
+        assert_eq!(
+            res.as_any()
+                .downcast_ref::<PrimitiveColumn<i32>>()
+                .unwrap()
+                .values(),
+            test.expect
+        );
+    }
+}