Struct UnionArray

pub struct UnionArray {
    data_type: DataType,
    type_ids: ScalarBuffer<i8>,
    offsets: Option<ScalarBuffer<i32>>,
    fields: Vec<Option<Arc<dyn Array>>>,
}

An array of values of varying types

Each slot in a UnionArray can have a value chosen from a number of types. Each of the possible types are named like the fields of a StructArray. A UnionArray can have two possible memory layouts, “dense” or “sparse”. For more information on please see the specification.

UnionBuilder can be used to create UnionArray’s of primitive types. UnionArray’s of nested types are also supported but not via UnionBuilder, see the tests for examples.

Examples

Create a dense UnionArray [1, 3.2, 34]

use arrow_buffer::ScalarBuffer;
use arrow_schema::*;
use std::sync::Arc;
use arrow_array::{Array, Int32Array, Float64Array, UnionArray};

let int_array = Int32Array::from(vec![1, 34]);
let float_array = Float64Array::from(vec![3.2]);
let type_ids = [0, 1, 0].into_iter().collect::<ScalarBuffer<i8>>();
let offsets = [0, 0, 1].into_iter().collect::<ScalarBuffer<i32>>();

let union_fields = [
    (0, Arc::new(Field::new("A", DataType::Int32, false))),
    (1, Arc::new(Field::new("B", DataType::Float64, false))),
].into_iter().collect::<UnionFields>();

let children = vec![
    Arc::new(int_array) as Arc<dyn Array>,
    Arc::new(float_array),
];

let array = UnionArray::try_new(
    union_fields,
    type_ids,
    Some(offsets),
    children,
).unwrap();

let value = array.value(0).as_any().downcast_ref::<Int32Array>().unwrap().value(0);
assert_eq!(1, value);

let value = array.value(1).as_any().downcast_ref::<Float64Array>().unwrap().value(0);
assert!(3.2 - value < f64::EPSILON);

let value = array.value(2).as_any().downcast_ref::<Int32Array>().unwrap().value(0);
assert_eq!(34, value);

Create a sparse UnionArray [1, 3.2, 34]

use arrow_buffer::ScalarBuffer;
use arrow_schema::*;
use std::sync::Arc;
use arrow_array::{Array, Int32Array, Float64Array, UnionArray};

let int_array = Int32Array::from(vec![Some(1), None, Some(34)]);
let float_array = Float64Array::from(vec![None, Some(3.2), None]);
let type_ids = [0_i8, 1, 0].into_iter().collect::<ScalarBuffer<i8>>();

let union_fields = [
    (0, Arc::new(Field::new("A", DataType::Int32, false))),
    (1, Arc::new(Field::new("B", DataType::Float64, false))),
].into_iter().collect::<UnionFields>();

let children = vec![
    Arc::new(int_array) as Arc<dyn Array>,
    Arc::new(float_array),
];

let array = UnionArray::try_new(
    union_fields,
    type_ids,
    None,
    children,
).unwrap();

let value = array.value(0).as_any().downcast_ref::<Int32Array>().unwrap().value(0);
assert_eq!(1, value);

let value = array.value(1).as_any().downcast_ref::<Float64Array>().unwrap().value(0);
assert!(3.2 - value < f64::EPSILON);

let value = array.value(2).as_any().downcast_ref::<Int32Array>().unwrap().value(0);
assert_eq!(34, value);

Fields

data_type: DataType

type_ids: ScalarBuffer<i8>

offsets: Option<ScalarBuffer<i32>>

fields: Vec<Option<Arc<dyn Array>>>

Implementations

impl UnionArray

pub unsafe fn new_unchecked( fields: UnionFields, type_ids: ScalarBuffer<i8>, offsets: Option<ScalarBuffer<i32>>, children: Vec<Arc<dyn Array>>, ) -> UnionArray

Creates a new UnionArray.

Accepts type ids, child arrays and optionally offsets (for dense unions) to create a new UnionArray. This method makes no attempt to validate the data provided by the caller and assumes that each of the components are correct and consistent with each other. See try_new for an alternative that validates the data provided.

Safety

The type_ids values should be positive and must match one of the type ids of the fields provided in fields. These values are used to index into the children arrays.

The offsets is provided in the case of a dense union, sparse unions should use None. If provided the offsets values should be positive and must be less than the length of the corresponding array.

In both cases above we use signed integer types to maintain compatibility with other Arrow implementations.

pub fn try_new( fields: UnionFields, type_ids: ScalarBuffer<i8>, offsets: Option<ScalarBuffer<i32>>, children: Vec<Arc<dyn Array>>, ) -> Result<UnionArray, ArrowError>

Attempts to create a new UnionArray, validating the inputs provided.

The order of child arrays child array order must match the fields order

pub fn child(&self, type_id: i8) -> &Arc<dyn Array>

Accesses the child array for type_id.

Panics

Panics if the type_id provided is not present in the array’s DataType in the Union.

pub fn type_id(&self, index: usize) -> i8

Returns the type_id for the array slot at index.

Panics

Panics if index is greater than or equal to the number of child arrays

pub fn type_ids(&self) -> &ScalarBuffer<i8>

Returns the type_ids buffer for this array

pub fn offsets(&self) -> Option<&ScalarBuffer<i32>>

Returns the offsets buffer if this is a dense array

pub fn value_offset(&self, index: usize) -> usize

Returns the offset into the underlying values array for the array slot at index.

Panics

Panics if index is greater than or equal the length of the array.

pub fn value(&self, i: usize) -> Arc<dyn Array>

Returns the array’s value at index i.

Panics

Panics if index i is out of bounds

pub fn type_names(&self) -> Vec<&str>

Returns the names of the types in the union.

pub fn slice(&self, offset: usize, length: usize) -> UnionArray

Returns a zero-copy slice of this array with the indicated offset and length.

pub fn into_parts( self, ) -> (UnionFields, ScalarBuffer<i8>, Option<ScalarBuffer<i32>>, Vec<Arc<dyn Array>>)

Deconstruct this array into its constituent parts

Example

let mut builder = UnionBuilder::new_dense();
builder.append::<Int32Type>("a", 1).unwrap();
let union_array = builder.build()?;

// Deconstruct into parts
let (union_fields, type_ids, offsets, children) = union_array.into_parts();

// Reconstruct from parts
let union_array = UnionArray::try_new(
    union_fields,
    type_ids,
    offsets,
    children,
);

Trait Implementations

impl Array for UnionArray

fn as_any(&self) -> &(dyn Any + 'static)

Returns the array as Any so that it can be downcasted to a specific implementation.

fn to_data(&self) -> ArrayData

Returns the underlying data of this array

fn into_data(self) -> ArrayData

Returns the underlying data of this array

fn data_type(&self) -> &DataType

Returns a reference to the DataType of this array.

fn slice(&self, offset: usize, length: usize) -> Arc<dyn Array>

Returns a zero-copy slice of this array with the indicated offset and length.

fn len(&self) -> usize

Returns the length (i.e., number of elements) of this array.

fn is_empty(&self) -> bool

Returns whether this array is empty.

fn offset(&self) -> usize

Returns the offset into the underlying data used by this array(-slice). Note that the underlying data can be shared by many arrays. This defaults to 0.

fn nulls(&self) -> Option<&NullBuffer>

Returns the null buffer of this array if any.

fn logical_nulls(&self) -> Option<NullBuffer>

Returns a potentially computed NullBuffer that represents the logical null values of this array, if any.

fn is_nullable(&self) -> bool

Returns false if the array is guaranteed to not contain any logical nulls

fn get_buffer_memory_size(&self) -> usize

Returns the total number of bytes of memory pointed to by this array. The buffers store bytes in the Arrow memory format, and include the data as well as the validity map. Note that this does not always correspond to the exact memory usage of an array, since multiple arrays can share the same buffers or slices thereof.

fn get_array_memory_size(&self) -> usize

Returns the total number of bytes of memory occupied physically by this array. This value will always be greater than returned by get_buffer_memory_size() and includes the overhead of the data structures that contain the pointers to the various buffers.

fn is_null(&self, index: usize) -> bool

Returns whether the element at index is null according to Array::nulls

fn is_valid(&self, index: usize) -> bool

Returns whether the element at index is not null, the opposite of Self::is_null.

fn null_count(&self) -> usize

Returns the total number of physical null values in this array.

fn logical_null_count(&self) -> usize

Returns the total number of logical null values in this array.

impl Clone for UnionArray

fn clone(&self) -> UnionArray

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for UnionArray

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl From<ArrayData> for UnionArray

fn from(data: ArrayData) -> UnionArray

Converts to this type from the input type.

impl From<UnionArray> for ArrayData

fn from(array: UnionArray) -> ArrayData

Converts to this type from the input type.