Merge remote-tracking branch 'apache/main' into nth_value_func_ref

.github/workflows/docs.yaml

@@ -26,7 +26,7 @@ jobs:
       - name: Setup Python
         uses: actions/setup-python@v5
         with:
-          python-version: "3.10"
+          python-version: "3.12"
 
       - name: Install dependencies
         run: |
@@ -61,4 +61,4 @@ jobs:
             git add --all
             git commit -m 'Publish built docs triggered by ${{ github.sha }}'
             git push || git push --force
-          fi
+          fi

Cargo.toml

@@ -141,6 +141,7 @@ pbjson = { version = "0.7.0" }
 prost = "0.13.1"
 prost-derive = "0.13.1"
 rand = "0.8"
+recursive = "0.1.1"
 regex = "1.8"
 rstest = "0.23.0"
 serde_json = "1"

datafusion-examples/examples/advanced_parquet_index.rs

@@ -229,9 +229,9 @@ async fn main() -> Result<()> {
 /// `file1.parquet` contains values `0..1000`
 #[derive(Debug)]
 pub struct IndexTableProvider {
-    /// Where the file is stored (cleanup on drop)
-    #[allow(dead_code)]
-    tmpdir: TempDir,
+    /// Pointer to temporary file storage. Keeping it in scope to prevent temporary folder
+    /// to be deleted prematurely
+    _tmpdir: TempDir,
     /// The file that is being read.
     indexed_file: IndexedFile,
     /// The underlying object store
@@ -250,7 +250,7 @@ impl IndexTableProvider {
 
         Ok(Self {
             indexed_file,
-            tmpdir,
+            _tmpdir: tmpdir,
             object_store,
             use_row_selections: AtomicBool::new(false),
         })

datafusion/common/Cargo.toml

@@ -63,11 +63,12 @@ object_store = { workspace = true, optional = true }
 parquet = { workspace = true, optional = true, default-features = true }
 paste = "1.0.15"
 pyo3 = { version = "0.22.0", optional = true }
+recursive = { workspace = true }
 sqlparser = { workspace = true }
 tokio = { workspace = true }
 
 [target.'cfg(target_family = "wasm")'.dependencies]
-instant = { version = "0.1", features = ["wasm-bindgen"] }
+web-time = "1.1.0"
 
 [dev-dependencies]
 rand = { workspace = true }

datafusion/common/src/instant.rs

@@ -18,9 +18,9 @@
 //! WASM-compatible `Instant` wrapper.
 
 #[cfg(target_family = "wasm")]
-/// DataFusion wrapper around [`std::time::Instant`]. Uses [`instant::Instant`]
+/// DataFusion wrapper around [`std::time::Instant`]. Uses [`web_time::Instant`]
 /// under `wasm` feature gate. It provides the same API as [`std::time::Instant`].
-pub type Instant = instant::Instant;
+pub type Instant = web_time::Instant;
 
 #[allow(clippy::disallowed_types)]
 #[cfg(not(target_family = "wasm"))]

datafusion/common/src/tree_node.rs

@@ -17,6 +17,7 @@
 
 //! [`TreeNode`] for visiting and rewriting expression and plan trees
 
+use recursive::recursive;
 use std::sync::Arc;
 
 use crate::Result;
@@ -123,6 +124,7 @@ pub trait TreeNode: Sized {
     /// TreeNodeVisitor::f_up(ChildNode2)
     /// TreeNodeVisitor::f_up(ParentNode)
     /// ```
+    #[recursive]
     fn visit<'n, V: TreeNodeVisitor<'n, Node = Self>>(
         &'n self,
         visitor: &mut V,
@@ -172,6 +174,7 @@ pub trait TreeNode: Sized {
     /// TreeNodeRewriter::f_up(ChildNode2)
     /// TreeNodeRewriter::f_up(ParentNode)
     /// ```
+    #[recursive]
     fn rewrite<R: TreeNodeRewriter<Node = Self>>(
         self,
         rewriter: &mut R,
@@ -194,6 +197,7 @@ pub trait TreeNode: Sized {
         &'n self,
         mut f: F,
     ) -> Result<TreeNodeRecursion> {
+        #[recursive]
         fn apply_impl<'n, N: TreeNode, F: FnMut(&'n N) -> Result<TreeNodeRecursion>>(
             node: &'n N,
             f: &mut F,
@@ -228,6 +232,7 @@ pub trait TreeNode: Sized {
         self,
         mut f: F,
     ) -> Result<Transformed<Self>> {
+        #[recursive]
         fn transform_down_impl<N: TreeNode, F: FnMut(N) -> Result<Transformed<N>>>(
             node: N,
             f: &mut F,
@@ -251,6 +256,7 @@ pub trait TreeNode: Sized {
         self,
         mut f: F,
     ) -> Result<Transformed<Self>> {
+        #[recursive]
         fn transform_up_impl<N: TreeNode, F: FnMut(N) -> Result<Transformed<N>>>(
             node: N,
             f: &mut F,
@@ -365,6 +371,7 @@ pub trait TreeNode: Sized {
         mut f_down: FD,
         mut f_up: FU,
     ) -> Result<Transformed<Self>> {
+        #[recursive]
         fn transform_down_up_impl<
             N: TreeNode,
             FD: FnMut(N) -> Result<Transformed<N>>,
@@ -2079,4 +2086,17 @@ pub(crate) mod tests {
 
         Ok(())
     }
+
+    #[test]
+    fn test_large_tree() {
+        let mut item = TestTreeNode::new_leaf("initial".to_string());
+        for i in 0..3000 {
+            item = TestTreeNode::new(vec![item], format!("parent-{}", i));
+        }
+
+        let mut visitor =
+            TestVisitor::new(Box::new(visit_continue), Box::new(visit_continue));
+
+        item.visit(&mut visitor).unwrap();
+    }
 }