fix: SQL interface "off-by-one' indexing error with GROUP BY clause…

…s that use position ordinals (#15584)

crates/polars-sql/src/context.rs

@@ -411,7 +411,7 @@ impl SQLContext {
             })
             .collect::<PolarsResult<_>>()?;
 
-        // Check for group by (after projections since there might be numbers).
+        // Check for group by (after projections as there may be ordinal/position ints).
         let group_by_keys: Vec<Expr>;
         if let GroupByExpr::Expressions(group_by_exprs) = &select_stmt.group_by {
             group_by_keys = group_by_exprs.iter()
@@ -425,7 +425,8 @@ impl SQLContext {
                             )),
                             Ok(idx) => Ok(idx),
                         }?;
-                        Ok(projections[idx].clone())
+                        // note: sql queries are 1-indexed
+                        Ok(projections[idx - 1].clone())
                     },
                     SQLExpr::Value(_) => Err(polars_err!(
                         ComputeError:
@@ -694,7 +695,6 @@ impl SQLContext {
             ComputeError: "group_by error: can't process wildcard in group_by"
         );
         let schema_before = lf.schema()?;
-
         let group_by_keys_schema =
             expressions_to_schema(group_by_keys, &schema_before, Context::Default)?;
 
@@ -703,24 +703,21 @@ impl SQLContext {
         let mut aliases: BTreeSet<&str> = BTreeSet::new();
 
         for mut e in projections {
-            // If it is a simple expression & has alias,
-            // we must defer the aliasing until after the group_by.
+            // If simple aliased expression we defer aliasing until after the group_by.
             if e.clone().meta().is_simple_projection() {
                 if let Expr::Alias(expr, name) = e {
                     aliases.insert(name);
                     e = expr
                 }
             }
-
             let field = e.to_field(&schema_before, Context::Default)?;
             if group_by_keys_schema.get(&field.name).is_none() {
                 aggregation_projection.push(e.clone())
             }
         }
-
-        let aggregated = lf.group_by(group_by_keys).agg(&aggregation_projection);
         let projection_schema =
             expressions_to_schema(projections, &schema_before, Context::Default)?;
+
         // A final projection to get the proper order.
         let final_projection = projection_schema
             .iter_names()
@@ -734,6 +731,7 @@ impl SQLContext {
             })
             .collect::<Vec<_>>();
 
+        let aggregated = lf.group_by(group_by_keys).agg(&aggregation_projection);
         Ok(aggregated.select(&final_projection))
     }
 

crates/polars-sql/tests/functions_string.rs

@@ -83,7 +83,7 @@ fn test_string_functions() {
 }
 
 #[test]
-fn array_to_string() {
+fn test_array_to_string() {
     let df = df! {
         "a" => &["first", "first", "third"],
         "b" => &[1, 1, 42],

crates/polars-sql/tests/simple_exprs.rs

@@ -53,6 +53,7 @@ fn test_nested_expr() -> PolarsResult<()> {
     assert_eq!(df_sql, df_pl);
     Ok(())
 }
+
 #[test]
 fn test_group_by_simple() -> PolarsResult<()> {
     let df = create_sample_df()?;
@@ -76,6 +77,7 @@ fn test_group_by_simple() -> PolarsResult<()> {
             },
         )
         .collect()?;
+
     let df_pl = df
         .lazy()
         .group_by(&[col("a")])
@@ -249,7 +251,7 @@ fn test_null_exprs_in_where() {
 }
 
 #[test]
-fn binary_functions() {
+fn test_binary_functions() {
     let df = create_sample_df().unwrap();
     let mut context = SQLContext::new();
     context.register("df", df.clone().lazy());

py-polars/tests/unit/sql/test_group_by.py

@@ -5,6 +5,7 @@
 import pytest
 
 import polars as pl
+from polars.testing import assert_frame_equal
 
 
 @pytest.fixture()
@@ -62,3 +63,36 @@ def test_group_by(foods_ipc_path: Path) -> None:
         """
     )
     assert out.to_dict(as_series=False) == {"grp": ["c"], "n_dist_attr": [2]}
+
+
+def test_group_by_ordinal_position() -> None:
+    df = pl.DataFrame(
+        {
+            "a": ["xx", "yy", "xx", "yy", "xx", "zz"],
+            "b": [1, 2, 3, 4, 5, 6],
+            "c": [99, 99, 66, 66, 66, 66],
+        }
+    )
+    expected = pl.LazyFrame({"c": [66, 99], "total_b": [18, 3]})
+
+    with pl.SQLContext(frame=df) as ctx:
+        res1 = ctx.execute(
+            """
+            SELECT c, SUM(b) AS total_b
+            FROM frame
+            GROUP BY 1
+            ORDER BY c
+            """
+        )
+        assert_frame_equal(res1, expected)
+
+        res2 = ctx.execute(
+            """
+            WITH "grp" AS (
+              SELECT NULL::date as dt, c, SUM(b) AS total_b
+              FROM frame
+              GROUP BY 2, 1
+            )
+            SELECT c, total_b FROM grp ORDER BY c"""
+        )
+        assert_frame_equal(res2, expected)