feat: Add 'nulls_equal' parameter to is_in (#21426)

pola-rs · Feb 28, 2025 · 4e4d259 · 4e4d259
1 parent e537b3b
commit 4e4d259
Show file tree

Hide file tree

Showing 24 changed files with 536 additions and 131 deletions.
diff --git a/crates/polars-expr/src/expressions/apply.rs b/crates/polars-expr/src/expressions/apply.rs
@@ -438,7 +438,7 @@ impl PhysicalExpr for ApplyExpr {
         match function {
             FunctionExpr::Boolean(BooleanFunction::IsNull) => Some(self),
             #[cfg(feature = "is_in")]
-            FunctionExpr::Boolean(BooleanFunction::IsIn) => Some(self),
+            FunctionExpr::Boolean(BooleanFunction::IsIn { .. }) => Some(self),
             #[cfg(feature = "is_between")]
             FunctionExpr::Boolean(BooleanFunction::IsBetween { closed: _ }) => Some(self),
             FunctionExpr::Boolean(BooleanFunction::IsNotNull) => Some(self),
@@ -573,7 +573,7 @@ impl ApplyExpr {
                 }
             },
             #[cfg(feature = "is_in")]
-            FunctionExpr::Boolean(BooleanFunction::IsIn) => {
+            FunctionExpr::Boolean(BooleanFunction::IsIn { .. }) => {
                 let should_read = || -> Option<bool> {
                     let root = expr_to_leaf_column_name(&input[0]).ok()?;
 

diff --git a/crates/polars-lazy/src/tests/io.rs b/crates/polars-lazy/src/tests/io.rs
@@ -136,7 +136,7 @@ fn test_parquet_statistics() -> PolarsResult<()> {
 
     // issue: 13427
     let out = scan_foods_parquet(par)
-        .filter(col("calories").is_in(lit(Series::new("".into(), [0, 500]))))
+        .filter(col("calories").is_in(lit(Series::new("".into(), [0, 500])), false))
         .collect()?;
     assert_eq!(out.shape(), (0, 4));
 

diff --git a/crates/polars-lazy/src/tests/predicate_queries.rs b/crates/polars-lazy/src/tests/predicate_queries.rs
@@ -48,7 +48,7 @@ fn test_issue_2472() -> PolarsResult<()> {
         .extract(lit(r"(\d+-){4}(\w+)-"), 2)
         .cast(DataType::Int32)
         .alias("age");
-    let predicate = col("age").is_in(lit(Series::new("".into(), [2i32])));
+    let predicate = col("age").is_in(lit(Series::new("".into(), [2i32])), false);
 
     let out = base
         .clone()

diff --git a/crates/polars-lazy/src/tests/queries.rs b/crates/polars-lazy/src/tests/queries.rs
@@ -355,7 +355,7 @@ fn test_lazy_query_8() -> PolarsResult<()> {
     let mut selection = vec![];
 
     for &c in &["A", "B", "C", "D", "E"] {
-        let e = when(col(c).is_in(col("E")))
+        let e = when(col(c).is_in(col("E"), false))
             .then(col("A"))
             .otherwise(Null {}.lit())
             .alias(c);
@@ -1761,7 +1761,7 @@ fn test_is_in() -> PolarsResult<()> {
         .clone()
         .lazy()
         .group_by_stable([col("fruits")])
-        .agg([col("cars").is_in(col("cars").filter(col("cars").eq(lit("beetle"))))])
+        .agg([col("cars").is_in(col("cars").filter(col("cars").eq(lit("beetle"))), false)])
         .collect()?;
     let out = out.column("cars").unwrap();
     let out = out.explode()?;
@@ -1775,7 +1775,7 @@ fn test_is_in() -> PolarsResult<()> {
     let out = df
         .lazy()
         .group_by_stable([col("fruits")])
-        .agg([col("cars").is_in(lit(Series::new("a".into(), ["beetle", "vw"])))])
+        .agg([col("cars").is_in(lit(Series::new("a".into(), ["beetle", "vw"])), false)])
         .collect()?;
 
     let out = out.column("cars").unwrap();