Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
37 changes: 37 additions & 0 deletions rust/arrow/src/array/array.rs
Original file line number Diff line number Diff line change
Expand Up @@ -323,6 +323,12 @@ pub fn make_array(data: ArrayDataRef) -> ArrayRef {
}
}

/// Creates a new empty array
pub fn new_empty_array(data_type: &DataType) -> ArrayRef {
let data = ArrayData::new_empty(data_type);
make_array(Arc::new(data))
}

/// Creates a new array from two FFI pointers. Used to import arrays from the C Data Interface
/// # Safety
/// Assumes that these pointers represent valid C Data Interfaces, both in memory
Expand Down Expand Up @@ -375,3 +381,34 @@ where
}
Ok(())
}

#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_empty_primitive() {
let array = new_empty_array(&DataType::Int32);
let a = array.as_any().downcast_ref::<Int32Array>().unwrap();
assert_eq!(a.len(), 0);
let expected: &[i32] = &[];
assert_eq!(a.values(), expected);
}

#[test]
fn test_empty_variable_sized() {
let array = new_empty_array(&DataType::Utf8);
let a = array.as_any().downcast_ref::<StringArray>().unwrap();
assert_eq!(a.len(), 0);
assert_eq!(a.value_offset(0), 0i32);
}

#[test]
fn test_empty_list_primitive() {
let data_type =
DataType::List(Box::new(Field::new("item", DataType::Int32, false)));
let array = new_empty_array(&data_type);
let a = array.as_any().downcast_ref::<ListArray>().unwrap();
assert_eq!(a.len(), 0);
assert_eq!(a.value_offset(0), 0i32);
}
}
297 changes: 1 addition & 296 deletions rust/arrow/src/array/array_list.rs
Original file line number Diff line number Diff line change
Expand Up @@ -19,19 +19,15 @@ use std::any::Any;
use std::convert::From;
use std::fmt;
use std::mem;
use std::sync::Arc;

use num::Num;

use super::{
array::print_long_array, make_array, raw_pointer::RawPtrBox, Array, ArrayDataRef,
ArrayRef, BinaryBuilder, BooleanBuilder, FixedSizeListBuilder, PrimitiveBuilder,
StringBuilder,
ArrayRef,
};
use crate::array::builder::GenericListBuilder;
use crate::datatypes::ArrowNativeType;
use crate::datatypes::*;
use crate::error::{ArrowError, Result};

/// trait declaring an offset size, relevant for i32 vs i64 array types.
pub trait OffsetSizeTrait: ArrowNativeType + Num + Ord + std::ops::AddAssign {
Expand Down Expand Up @@ -302,264 +298,6 @@ impl fmt::Debug for FixedSizeListArray {
}
}

macro_rules! build_empty_list_array_with_primitive_items {
($item_type:ident, $offset_type:ident) => {{
let values_builder = PrimitiveBuilder::<$item_type>::new(0);
let mut builder =
GenericListBuilder::<$offset_type, PrimitiveBuilder<$item_type>>::new(
values_builder,
);
let empty_list_array = builder.finish();
Ok(Arc::new(empty_list_array))
}};
}

macro_rules! build_empty_list_array_with_non_primitive_items {
($type_builder:ident, $offset_type:ident) => {{
let values_builder = $type_builder::new(0);
let mut builder =
GenericListBuilder::<$offset_type, $type_builder>::new(values_builder);
let empty_list_array = builder.finish();
Ok(Arc::new(empty_list_array))
}};
}

pub fn build_empty_list_array<OffsetSize: OffsetSizeTrait>(
item_type: DataType,
) -> Result<ArrayRef> {
match item_type {
DataType::UInt8 => {
build_empty_list_array_with_primitive_items!(UInt8Type, OffsetSize)
}
DataType::UInt16 => {
build_empty_list_array_with_primitive_items!(UInt16Type, OffsetSize)
}
DataType::UInt32 => {
build_empty_list_array_with_primitive_items!(UInt32Type, OffsetSize)
}
DataType::UInt64 => {
build_empty_list_array_with_primitive_items!(UInt64Type, OffsetSize)
}
DataType::Int8 => {
build_empty_list_array_with_primitive_items!(Int8Type, OffsetSize)
}
DataType::Int16 => {
build_empty_list_array_with_primitive_items!(Int16Type, OffsetSize)
}
DataType::Int32 => {
build_empty_list_array_with_primitive_items!(Int32Type, OffsetSize)
}
DataType::Int64 => {
build_empty_list_array_with_primitive_items!(Int64Type, OffsetSize)
}
DataType::Float32 => {
build_empty_list_array_with_primitive_items!(Float32Type, OffsetSize)
}
DataType::Float64 => {
build_empty_list_array_with_primitive_items!(Float64Type, OffsetSize)
}
DataType::Boolean => {
build_empty_list_array_with_non_primitive_items!(BooleanBuilder, OffsetSize)
}
DataType::Date32(_) => {
build_empty_list_array_with_primitive_items!(Date32Type, OffsetSize)
}
DataType::Date64(_) => {
build_empty_list_array_with_primitive_items!(Date64Type, OffsetSize)
}
DataType::Time32(TimeUnit::Second) => {
build_empty_list_array_with_primitive_items!(Time32SecondType, OffsetSize)
}
DataType::Time32(TimeUnit::Millisecond) => {
build_empty_list_array_with_primitive_items!(
Time32MillisecondType,
OffsetSize
)
}
DataType::Time64(TimeUnit::Microsecond) => {
build_empty_list_array_with_primitive_items!(
Time64MicrosecondType,
OffsetSize
)
}
DataType::Time64(TimeUnit::Nanosecond) => {
build_empty_list_array_with_primitive_items!(Time64NanosecondType, OffsetSize)
}
DataType::Duration(TimeUnit::Second) => {
build_empty_list_array_with_primitive_items!(DurationSecondType, OffsetSize)
}
DataType::Duration(TimeUnit::Millisecond) => {
build_empty_list_array_with_primitive_items!(
DurationMillisecondType,
OffsetSize
)
}
DataType::Duration(TimeUnit::Microsecond) => {
build_empty_list_array_with_primitive_items!(
DurationMicrosecondType,
OffsetSize
)
}
DataType::Duration(TimeUnit::Nanosecond) => {
build_empty_list_array_with_primitive_items!(
DurationNanosecondType,
OffsetSize
)
}
DataType::Timestamp(TimeUnit::Second, _) => {
build_empty_list_array_with_primitive_items!(TimestampSecondType, OffsetSize)
}
DataType::Timestamp(TimeUnit::Millisecond, _) => {
build_empty_list_array_with_primitive_items!(
TimestampMillisecondType,
OffsetSize
)
}
DataType::Timestamp(TimeUnit::Microsecond, _) => {
build_empty_list_array_with_primitive_items!(
TimestampMicrosecondType,
OffsetSize
)
}
DataType::Timestamp(TimeUnit::Nanosecond, _) => {
build_empty_list_array_with_primitive_items!(
TimestampNanosecondType,
OffsetSize
)
}
DataType::Utf8 => {
build_empty_list_array_with_non_primitive_items!(StringBuilder, OffsetSize)
}
DataType::Binary => {
build_empty_list_array_with_non_primitive_items!(BinaryBuilder, OffsetSize)
}
_ => Err(ArrowError::NotYetImplemented(format!(
"GenericListBuilder of type List({:?}) is not supported",
item_type
))),
}
}

macro_rules! build_empty_fixed_size_list_array_with_primitive_items {
($item_type:ident) => {{
let values_builder = PrimitiveBuilder::<$item_type>::new(0);
let mut builder = FixedSizeListBuilder::new(values_builder, 0);
let empty_list_array = builder.finish();
Ok(Arc::new(empty_list_array))
}};
}

macro_rules! build_empty_fixed_size_list_array_with_non_primitive_items {
($type_builder:ident) => {{
let values_builder = $type_builder::new(0);
let mut builder = FixedSizeListBuilder::new(values_builder, 0);
let empty_list_array = builder.finish();
Ok(Arc::new(empty_list_array))
}};
}

pub fn build_empty_fixed_size_list_array(item_type: DataType) -> Result<ArrayRef> {
match item_type {
DataType::UInt8 => {
build_empty_fixed_size_list_array_with_primitive_items!(UInt8Type)
}
DataType::UInt16 => {
build_empty_fixed_size_list_array_with_primitive_items!(UInt16Type)
}
DataType::UInt32 => {
build_empty_fixed_size_list_array_with_primitive_items!(UInt32Type)
}
DataType::UInt64 => {
build_empty_fixed_size_list_array_with_primitive_items!(UInt64Type)
}
DataType::Int8 => {
build_empty_fixed_size_list_array_with_primitive_items!(Int8Type)
}
DataType::Int16 => {
build_empty_fixed_size_list_array_with_primitive_items!(Int16Type)
}
DataType::Int32 => {
build_empty_fixed_size_list_array_with_primitive_items!(Int32Type)
}
DataType::Int64 => {
build_empty_fixed_size_list_array_with_primitive_items!(Int64Type)
}
DataType::Float32 => {
build_empty_fixed_size_list_array_with_primitive_items!(Float32Type)
}
DataType::Float64 => {
build_empty_fixed_size_list_array_with_primitive_items!(Float64Type)
}
DataType::Boolean => {
build_empty_fixed_size_list_array_with_non_primitive_items!(BooleanBuilder)
}
DataType::Date32(_) => {
build_empty_fixed_size_list_array_with_primitive_items!(Date32Type)
}
DataType::Date64(_) => {
build_empty_fixed_size_list_array_with_primitive_items!(Date64Type)
}
DataType::Time32(TimeUnit::Second) => {
build_empty_fixed_size_list_array_with_primitive_items!(Time32SecondType)
}
DataType::Time32(TimeUnit::Millisecond) => {
build_empty_fixed_size_list_array_with_primitive_items!(Time32MillisecondType)
}
DataType::Time64(TimeUnit::Microsecond) => {
build_empty_fixed_size_list_array_with_primitive_items!(Time64MicrosecondType)
}
DataType::Time64(TimeUnit::Nanosecond) => {
build_empty_fixed_size_list_array_with_primitive_items!(Time64NanosecondType)
}
DataType::Duration(TimeUnit::Second) => {
build_empty_fixed_size_list_array_with_primitive_items!(DurationSecondType)
}
DataType::Duration(TimeUnit::Millisecond) => {
build_empty_fixed_size_list_array_with_primitive_items!(
DurationMillisecondType
)
}
DataType::Duration(TimeUnit::Microsecond) => {
build_empty_fixed_size_list_array_with_primitive_items!(
DurationMicrosecondType
)
}
DataType::Duration(TimeUnit::Nanosecond) => {
build_empty_fixed_size_list_array_with_primitive_items!(
DurationNanosecondType
)
}
DataType::Timestamp(TimeUnit::Second, _) => {
build_empty_fixed_size_list_array_with_primitive_items!(TimestampSecondType)
}
DataType::Timestamp(TimeUnit::Millisecond, _) => {
build_empty_fixed_size_list_array_with_primitive_items!(
TimestampMillisecondType
)
}
DataType::Timestamp(TimeUnit::Microsecond, _) => {
build_empty_fixed_size_list_array_with_primitive_items!(
TimestampMicrosecondType
)
}
DataType::Timestamp(TimeUnit::Nanosecond, _) => {
build_empty_fixed_size_list_array_with_primitive_items!(
TimestampNanosecondType
)
}
DataType::Utf8 => {
build_empty_fixed_size_list_array_with_non_primitive_items!(StringBuilder)
}
DataType::Binary => {
build_empty_fixed_size_list_array_with_non_primitive_items!(BinaryBuilder)
}
_ => Err(ArrowError::NotYetImplemented(format!(
"FixedSizeListBuilder of type FixedSizeList({:?}) is not supported",
item_type
))),
}
}

#[cfg(test)]
mod tests {
use crate::{
Expand Down Expand Up @@ -1066,37 +804,4 @@ mod tests {
.build();
ListArray::from(list_data);
}

macro_rules! make_test_build_empty_list_array {
($OFFSET:ident) => {
build_empty_list_array::<$OFFSET>(DataType::Boolean).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Int16).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Int32).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Int64).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Float32).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Float64).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Boolean).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Utf8).unwrap();
build_empty_list_array::<$OFFSET>(DataType::Binary).unwrap();
};
}

#[test]
fn test_build_empty_list_array() {
make_test_build_empty_list_array!(i32);
make_test_build_empty_list_array!(i64);
}

#[test]
fn test_build_empty_fixed_size_list_array() {
build_empty_fixed_size_list_array(DataType::Boolean).unwrap();
build_empty_fixed_size_list_array(DataType::Int16).unwrap();
build_empty_fixed_size_list_array(DataType::Int32).unwrap();
build_empty_fixed_size_list_array(DataType::Int64).unwrap();
build_empty_fixed_size_list_array(DataType::Float32).unwrap();
build_empty_fixed_size_list_array(DataType::Float64).unwrap();
build_empty_fixed_size_list_array(DataType::Boolean).unwrap();
build_empty_fixed_size_list_array(DataType::Utf8).unwrap();
build_empty_fixed_size_list_array(DataType::Binary).unwrap();
}
}
Loading