Merge branch 'main' into fix-doc-123

Author: Jefffrey

.github/workflows/integration.yml

@@ -78,58 +78,112 @@ jobs:
       run:
         shell: bash
     steps:
+      - name: Monitor disk usage - Initial
+        run: |
+          echo "=== Initial Disk Usage ==="
+          df -h /
+          echo ""
+
+      - name: Remove unnecessary preinstalled software
+        run: |
+          echo "=== Cleaning up host disk space ==="
+          echo "Disk space before cleanup:"
+          df -h /
+
+          # Clean apt cache
+          apt-get clean || true
+
+          # Remove GitHub Actions tool cache
+          rm -rf /__t/* || true
+
+          # Remove large packages from host filesystem (mounted at /host/)
+          rm -rf /host/usr/share/dotnet || true
+          rm -rf /host/usr/local/lib/android || true
+          rm -rf /host/usr/local/.ghcup || true
+          rm -rf /host/opt/hostedtoolcache/CodeQL || true
+
+          echo ""
+          echo "Disk space after cleanup:"
+          df -h /
+          echo ""
+
       # This is necessary so that actions/checkout can find git
       - name: Export conda path
         run: echo "/opt/conda/envs/arrow/bin" >> $GITHUB_PATH
       # This is necessary so that Rust can find cargo
       - name: Export cargo path
         run: echo "/root/.cargo/bin" >> $GITHUB_PATH
-      - name: Check rustup
-        run: which rustup
-      - name: Check cmake
-        run: which cmake
+
+      # Checkout repos (using shallow clones with fetch-depth: 1)
       - name: Checkout Arrow
         uses: actions/checkout@v6
         with:
           repository: apache/arrow
           submodules: true
-          fetch-depth: 0
+          fetch-depth: 1
       - name: Checkout Arrow Rust
         uses: actions/checkout@v6
         with:
           path: rust
           submodules: true
-          fetch-depth: 0
+          fetch-depth: 1
       - name: Checkout Arrow .NET
         uses: actions/checkout@v6
         with:
           repository: apache/arrow-dotnet
           path: dotnet
+          fetch-depth: 1
       - name: Checkout Arrow Go
         uses: actions/checkout@v6
         with:
           repository: apache/arrow-go
           path: go
+          fetch-depth: 1
       - name: Checkout Arrow Java
         uses: actions/checkout@v6
         with:
           repository: apache/arrow-java
           path: java
+          fetch-depth: 1
       - name: Checkout Arrow JavaScript
         uses: actions/checkout@v6
         with:
           repository: apache/arrow-js
           path: js
+          fetch-depth: 1
       - name: Checkout Arrow nanoarrow
         uses: actions/checkout@v6
         with:
           repository: apache/arrow-nanoarrow
           path: nanoarrow
+          fetch-depth: 1
+
+      - name: Monitor disk usage - After checkouts
+        run: |
+          echo "=== After Checkouts ==="
+          df -h /
+          echo ""
+
       - name: Build
         run: conda run --no-capture-output ci/scripts/integration_arrow_build.sh $PWD /build
+
+      - name: Monitor disk usage - After build
+        if: always()
+        run: |
+          echo "=== After Build ==="
+          df -h /
+          echo ""
+
       - name: Run
         run: conda run --no-capture-output ci/scripts/integration_arrow.sh $PWD /build
 
+      - name: Monitor disk usage - After tests
+        if: always()
+        run: |
+          echo "=== After Tests ==="
+          df -h /
+          echo ""
+
   # test FFI against the C-Data interface exposed by pyarrow
   pyarrow-integration-test:
     name: Pyarrow C Data Interface

CHANGELOG-old.md

@@ -68,7 +68,7 @@ exclude = [
 ]
 
 [workspace.package]
-version = "57.1.0"
+version = "57.2.0"
 homepage = "https://github.com/apache/arrow-rs"
 repository = "https://github.com/apache/arrow-rs"
 authors = ["Apache Arrow <dev@arrow.apache.org>"]
@@ -85,26 +85,26 @@ edition = "2024"
 rust-version = "1.85"
 
 [workspace.dependencies]
-arrow = { version = "57.1.0", path = "./arrow", default-features = false }
-arrow-arith = { version = "57.1.0", path = "./arrow-arith" }
-arrow-array = { version = "57.1.0", path = "./arrow-array" }
-arrow-buffer = { version = "57.1.0", path = "./arrow-buffer" }
-arrow-cast = { version = "57.1.0", path = "./arrow-cast" }
-arrow-csv = { version = "57.1.0", path = "./arrow-csv" }
-arrow-data = { version = "57.1.0", path = "./arrow-data" }
-arrow-ipc = { version = "57.1.0", path = "./arrow-ipc" }
-arrow-json = { version = "57.1.0", path = "./arrow-json" }
-arrow-ord = { version = "57.1.0", path = "./arrow-ord" }
-arrow-pyarrow = { version = "57.1.0", path = "./arrow-pyarrow" }
-arrow-row = { version = "57.1.0", path = "./arrow-row" }
-arrow-schema = { version = "57.1.0", path = "./arrow-schema" }
-arrow-select = { version = "57.1.0", path = "./arrow-select" }
-arrow-string = { version = "57.1.0", path = "./arrow-string" }
-parquet = { version = "57.1.0", path = "./parquet", default-features = false }
-parquet-geospatial = { version = "57.1.0", path = "./parquet-geospatial" }
-parquet-variant = { version = "57.1.0", path = "./parquet-variant" }
-parquet-variant-json = { version = "57.1.0", path = "./parquet-variant-json" }
-parquet-variant-compute = { version = "57.1.0", path = "./parquet-variant-compute" }
+arrow = { version = "57.2.0", path = "./arrow", default-features = false }
+arrow-arith = { version = "57.2.0", path = "./arrow-arith" }
+arrow-array = { version = "57.2.0", path = "./arrow-array" }
+arrow-buffer = { version = "57.2.0", path = "./arrow-buffer" }
+arrow-cast = { version = "57.2.0", path = "./arrow-cast" }
+arrow-csv = { version = "57.2.0", path = "./arrow-csv" }
+arrow-data = { version = "57.2.0", path = "./arrow-data" }
+arrow-ipc = { version = "57.2.0", path = "./arrow-ipc" }
+arrow-json = { version = "57.2.0", path = "./arrow-json" }
+arrow-ord = { version = "57.2.0", path = "./arrow-ord" }
+arrow-pyarrow = { version = "57.2.0", path = "./arrow-pyarrow" }
+arrow-row = { version = "57.2.0", path = "./arrow-row" }
+arrow-schema = { version = "57.2.0", path = "./arrow-schema" }
+arrow-select = { version = "57.2.0", path = "./arrow-select" }
+arrow-string = { version = "57.2.0", path = "./arrow-string" }
+parquet = { version = "57.2.0", path = "./parquet", default-features = false }
+parquet-geospatial = { version = "57.2.0", path = "./parquet-geospatial" }
+parquet-variant = { version = "57.2.0", path = "./parquet-variant" }
+parquet-variant-json = { version = "57.2.0", path = "./parquet-variant-json" }
+parquet-variant-compute = { version = "57.2.0", path = "./parquet-variant-compute" }
 
 chrono = { version = "0.4.40", default-features = false, features = ["clock"] }
 

CHANGELOG.md

@@ -332,10 +332,10 @@ fn aggregate<T: ArrowNativeTypeOp, P: ArrowPrimitiveType<Native = T>, A: Numeric
 
 /// Returns the minimum value in the boolean array.
 ///
+/// # Example
 /// ```
 /// # use arrow_array::BooleanArray;
 /// # use arrow_arith::aggregate::min_boolean;
-///
 /// let a = BooleanArray::from(vec![Some(true), None, Some(false)]);
 /// assert_eq!(min_boolean(&a), Some(false))
 /// ```
@@ -390,10 +390,10 @@ pub fn min_boolean(array: &BooleanArray) -> Option<bool> {
 
 /// Returns the maximum value in the boolean array
 ///
+/// # Example
 /// ```
 /// # use arrow_array::BooleanArray;
 /// # use arrow_arith::aggregate::max_boolean;
-///
 /// let a = BooleanArray::from(vec![Some(true), None, Some(false)]);
 /// assert_eq!(max_boolean(&a), Some(true))
 /// ```
@@ -809,6 +809,15 @@ where
 
 /// Returns the minimum value in the array, according to the natural order.
 /// For floating point arrays any NaN values are considered to be greater than any other non-null value
+///
+/// # Example
+/// ```rust
+/// # use arrow_array::Int32Array;
+/// # use arrow_arith::aggregate::min;
+/// let array = Int32Array::from(vec![8, 2, 4]);
+/// let result = min(&array);
+/// assert_eq!(result, Some(2));
+/// ```
 pub fn min<T: ArrowNumericType>(array: &PrimitiveArray<T>) -> Option<T::Native>
 where
     T::Native: PartialOrd,
@@ -818,6 +827,15 @@ where
 
 /// Returns the maximum value in the array, according to the natural order.
 /// For floating point arrays any NaN values are considered to be greater than any other non-null value
+///
+/// # Example
+/// ```rust
+/// # use arrow_array::Int32Array;
+/// # use arrow_arith::aggregate::max;
+/// let array = Int32Array::from(vec![4, 8, 2]);
+/// let result = max(&array);
+/// assert_eq!(result, Some(8));
+/// ```
 pub fn max<T: ArrowNumericType>(array: &PrimitiveArray<T>) -> Option<T::Native>
 where
     T::Native: PartialOrd,

Cargo.toml

@@ -286,6 +286,8 @@ impl BooleanArray {
     }
 }
 
+impl super::private::Sealed for BooleanArray {}
+
 impl Array for BooleanArray {
     fn as_any(&self) -> &dyn Any {
         self

arrow-arith/src/aggregate.rs

@@ -462,6 +462,8 @@ impl<T: ByteArrayType> std::fmt::Debug for GenericByteArray<T> {
     }
 }
 
+impl<T: ByteArrayType> super::private::Sealed for GenericByteArray<T> {}
+
 impl<T: ByteArrayType> Array for GenericByteArray<T> {
     fn as_any(&self) -> &dyn Any {
         self

arrow-array/src/array/boolean_array.rs

@@ -854,6 +854,8 @@ impl<T: ByteViewType + ?Sized> Debug for GenericByteViewArray<T> {
     }
 }
 
+impl<T: ByteViewType + ?Sized> super::private::Sealed for GenericByteViewArray<T> {}
+
 impl<T: ByteViewType + ?Sized> Array for GenericByteViewArray<T> {
     fn as_any(&self) -> &dyn Any {
         self

arrow-array/src/array/byte_array.rs

@@ -697,6 +697,8 @@ impl<'a, T: ArrowDictionaryKeyType> FromIterator<&'a str> for DictionaryArray<T>
     }
 }
 
+impl<T: ArrowDictionaryKeyType> super::private::Sealed for DictionaryArray<T> {}
+
 impl<T: ArrowDictionaryKeyType> Array for DictionaryArray<T> {
     fn as_any(&self) -> &dyn Any {
         self
@@ -856,6 +858,8 @@ impl<'a, K: ArrowDictionaryKeyType, V> TypedDictionaryArray<'a, K, V> {
     }
 }
 
+impl<K: ArrowDictionaryKeyType, V: Sync> super::private::Sealed for TypedDictionaryArray<'_, K, V> {}
+
 impl<K: ArrowDictionaryKeyType, V: Sync> Array for TypedDictionaryArray<'_, K, V> {
     fn as_any(&self) -> &dyn Any {
         self.dictionary

arrow-array/src/array/byte_view_array.rs

@@ -602,6 +602,8 @@ impl std::fmt::Debug for FixedSizeBinaryArray {
     }
 }
 
+impl super::private::Sealed for FixedSizeBinaryArray {}
+
 impl Array for FixedSizeBinaryArray {
     fn as_any(&self) -> &dyn Any {
         self