fix(rust): fix bug in unneeded projection pruning (#5071)

polars/polars-lazy/polars-plan/src/logical_plan/optimizer/projection_pushdown.rs

@@ -288,8 +288,11 @@ impl ProjectionPushDown {
                 // So we:
                 //      - add the root of the projections to accumulation,
                 //      - also do the complete projection locally to keep the schema (column order) and the alias.
+
+                // set this flag outside the loop as we modify within the loop
+                let has_pushed_down = !acc_projections.is_empty();
                 for e in &expr {
-                    if !acc_projections.is_empty() {
+                    if has_pushed_down {
                         // remove projections that are not used upstream
                         if projections_seen > 0 {
                             let input_schema = lp_arena.get(input).schema(lp_arena);
@@ -313,7 +316,7 @@ impl ProjectionPushDown {
                         // In this query, bar cannot pass this projection, as it would not exist in DF.
                         // THE ORDER IS IMPORTANT HERE!
                         // this removes projection names, so any checks ot upstream names should
-                        // be done befor this branch.
+                        // be done before this branch.
                         for (_, ae) in (&*expr_arena).iter(*e) {
                             if let AExpr::Alias(_, name) = ae {
                                 if projected_names.remove(name) {

polars/tests/it/lazy/projection_queries.rs

@@ -92,3 +92,29 @@ fn test_err_no_found() {
         Err(PolarsError::NotFound(_))
     ));
 }
+
+#[test]
+fn test_many_aliasing_projections_5070() -> PolarsResult<()> {
+    let df = df! {
+        "date" => [1, 2, 3],
+        "val" => [1, 2, 3],
+    }?;
+
+    let out = df
+        .lazy()
+        .filter(col("date").gt(lit(1)))
+        .select([col("*")])
+        .with_columns([col("val").max().alias("max")])
+        .with_column(col("max").alias("diff"))
+        .with_column((col("val") / col("diff")).alias("output"))
+        .select([all().exclude(&["max", "diff"])])
+        .collect()?;
+    let expected = df![
+        "date" => [2, 3],
+        "val" => [2, 3],
+        "output" => [0, 1],
+    ]?;
+    assert!(out.frame_equal(&expected));
+
+    Ok(())
+}