Skip to content

ARROW-11327: [Rust][DataFusion] Add DictionarySupport to create_batch_empty #9276

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
alamb wants to merge 1 commit into from
Closed

ARROW-11327: [Rust][DataFusion] Add DictionarySupport to create_batch_empty #9276

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
311 changes: 168 additions & 143 deletions rust/datafusion/src/physical_plan/common.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,8 +31,15 @@ use array::{
Time32MillisecondArray, Time32SecondArray, UInt16Array, UInt32Array, UInt64Array,
UInt8Array,
};
use arrow::error::Result as ArrowResult;
use arrow::record_batch::RecordBatch;
use arrow::{
array::PrimitiveBuilder,
datatypes::{
ArrowPrimitiveType, Int16Type, Int32Type, Int64Type, Int8Type, UInt16Type,
UInt32Type, UInt64Type, UInt8Type,
},
error::Result as ArrowResult,
};
use arrow::{
array::{self, ArrayRef},
datatypes::Schema,
Expand Down Expand Up @@ -121,130 +128,138 @@ pub fn build_file_list(dir: &str, filenames: &mut Vec<String>, ext: &str) -> Res
Ok(())
}

/// creates an empty record batch.
/// Creates an empty (0 row) record batch with the specified schema
pub fn create_batch_empty(schema: &Schema) -> ArrowResult<RecordBatch> {
let columns = schema
.fields()
.iter()
.map(|f| match f.data_type() {
Copy link
Contributor Author

@alamb alamb Jan 20, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this code is just hoisted into its own function so it can be called recursively

DataType::Float32 => {
Ok(Arc::new(Float32Array::from(vec![] as Vec<f32>)) as ArrayRef)
}
DataType::Float64 => {
Ok(Arc::new(Float64Array::from(vec![] as Vec<f64>)) as ArrayRef)
}
DataType::Int64 => {
Ok(Arc::new(Int64Array::from(vec![] as Vec<i64>)) as ArrayRef)
}
DataType::Int32 => {
Ok(Arc::new(Int32Array::from(vec![] as Vec<i32>)) as ArrayRef)
}
DataType::Int16 => {
Ok(Arc::new(Int16Array::from(vec![] as Vec<i16>)) as ArrayRef)
}
DataType::Int8 => {
Ok(Arc::new(Int8Array::from(vec![] as Vec<i8>)) as ArrayRef)
}
DataType::UInt64 => {
Ok(Arc::new(UInt64Array::from(vec![] as Vec<u64>)) as ArrayRef)
}
DataType::UInt32 => {
Ok(Arc::new(UInt32Array::from(vec![] as Vec<u32>)) as ArrayRef)
}
DataType::UInt16 => {
Ok(Arc::new(UInt16Array::from(vec![] as Vec<u16>)) as ArrayRef)
}
DataType::UInt8 => {
Ok(Arc::new(UInt8Array::from(vec![] as Vec<u8>)) as ArrayRef)
}
DataType::Utf8 => {
Ok(Arc::new(StringArray::from(vec![] as Vec<&str>)) as ArrayRef)
.map(|f| create_empty_array(f.data_type()))
.collect::<Result<_>>()
.map_err(DataFusionError::into_arrow_external_error)?;

RecordBatch::try_new(Arc::new(schema.to_owned()), columns)
}

fn create_empty_array(data_type: &DataType) -> Result<ArrayRef> {
match data_type {
DataType::Float32 => {
Ok(Arc::new(Float32Array::from(vec![] as Vec<f32>)) as ArrayRef)
}
DataType::Float64 => {
Ok(Arc::new(Float64Array::from(vec![] as Vec<f64>)) as ArrayRef)
}
DataType::Int64 => Ok(Arc::new(Int64Array::from(vec![] as Vec<i64>)) as ArrayRef),
DataType::Int32 => Ok(Arc::new(Int32Array::from(vec![] as Vec<i32>)) as ArrayRef),
DataType::Int16 => Ok(Arc::new(Int16Array::from(vec![] as Vec<i16>)) as ArrayRef),
DataType::Int8 => Ok(Arc::new(Int8Array::from(vec![] as Vec<i8>)) as ArrayRef),
DataType::UInt64 => {
Ok(Arc::new(UInt64Array::from(vec![] as Vec<u64>)) as ArrayRef)
}
DataType::UInt32 => {
Ok(Arc::new(UInt32Array::from(vec![] as Vec<u32>)) as ArrayRef)
}
DataType::UInt16 => {
Ok(Arc::new(UInt16Array::from(vec![] as Vec<u16>)) as ArrayRef)
}
DataType::UInt8 => Ok(Arc::new(UInt8Array::from(vec![] as Vec<u8>)) as ArrayRef),
DataType::Utf8 => {
Ok(Arc::new(StringArray::from(vec![] as Vec<&str>)) as ArrayRef)
}
DataType::LargeUtf8 => {
Ok(Arc::new(LargeStringArray::from(vec![] as Vec<&str>)) as ArrayRef)
}
DataType::Boolean => {
Ok(Arc::new(BooleanArray::from(vec![] as Vec<bool>)) as ArrayRef)
}
DataType::Decimal(scale, precision) => {
let array_data = ArrayData::builder(DataType::Decimal(*scale, *precision))
.len(0)
.add_buffer(Buffer::from(&[]))
.build();
Ok(Arc::new(DecimalArray::from(array_data)) as ArrayRef)
}
DataType::Timestamp(TimeUnit::Nanosecond, tz) => Ok(Arc::new(
TimestampNanosecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
) as ArrayRef),
DataType::Timestamp(TimeUnit::Microsecond, tz) => Ok(Arc::new(
TimestampMicrosecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
) as ArrayRef),
DataType::Timestamp(TimeUnit::Millisecond, tz) => Ok(Arc::new(
TimestampMillisecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
) as ArrayRef),
DataType::Timestamp(TimeUnit::Second, tz) => Ok(Arc::new(
TimestampSecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
) as ArrayRef),
DataType::Date32(_) => {
Ok(Arc::new(Date32Array::from(vec![] as Vec<i32>)) as ArrayRef)
}
DataType::Date64(_) => {
Ok(Arc::new(Date64Array::from(vec![] as Vec<i64>)) as ArrayRef)
}
DataType::Time32(unit) => match unit {
TimeUnit::Second => {
Ok(Arc::new(Time32SecondArray::from(vec![] as Vec<i32>)) as ArrayRef)
}
DataType::LargeUtf8 => {
Ok(Arc::new(LargeStringArray::from(vec![] as Vec<&str>)) as ArrayRef)
TimeUnit::Millisecond => {
Ok(Arc::new(Time32MillisecondArray::from(vec![] as Vec<i32>))
as ArrayRef)
}
DataType::Boolean => {
Ok(Arc::new(BooleanArray::from(vec![] as Vec<bool>)) as ArrayRef)
TimeUnit::Microsecond | TimeUnit::Nanosecond => {
Err(DataFusionError::NotImplemented(format!(
"Cannot convert datatype {:?} to array",
data_type
)))
}
DataType::Decimal(scale, precision) => {
let array_data =
ArrayData::builder(DataType::Decimal(*scale, *precision))
.len(0)
.add_buffer(Buffer::from(&[]))
.build();

Ok(Arc::new(DecimalArray::from(array_data)) as ArrayRef)
},
DataType::Time64(unit) => match unit {
TimeUnit::Second | TimeUnit::Millisecond => {
Err(DataFusionError::NotImplemented(format!(
"Cannot convert datatype {:?} to array",
data_type
)))
}
DataType::Timestamp(TimeUnit::Nanosecond, tz) => Ok(Arc::new(
TimestampNanosecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
)
as ArrayRef),
DataType::Timestamp(TimeUnit::Microsecond, tz) => Ok(Arc::new(
TimestampMicrosecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
)
as ArrayRef),
DataType::Timestamp(TimeUnit::Millisecond, tz) => Ok(Arc::new(
TimestampMillisecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
)
as ArrayRef),
DataType::Timestamp(TimeUnit::Second, tz) => Ok(Arc::new(
TimestampSecondArray::from_vec(vec![] as Vec<i64>, tz.clone()),
) as ArrayRef),
DataType::Date32(_) => {
Ok(Arc::new(Date32Array::from(vec![] as Vec<i32>)) as ArrayRef)
TimeUnit::Microsecond => {
Ok(Arc::new(Time64MicrosecondArray::from(vec![] as Vec<i64>))
as ArrayRef)
}
DataType::Date64(_) => {
Ok(Arc::new(Date64Array::from(vec![] as Vec<i64>)) as ArrayRef)
TimeUnit::Nanosecond => {
Ok(Arc::new(Time64NanosecondArray::from(vec![] as Vec<i64>)) as ArrayRef)
}
DataType::Time32(unit) => match unit {
TimeUnit::Second => {
Ok(Arc::new(Time32SecondArray::from(vec![] as Vec<i32>)) as ArrayRef)
}
TimeUnit::Millisecond => {
Ok(Arc::new(Time32MillisecondArray::from(vec![] as Vec<i32>))
as ArrayRef)
}
TimeUnit::Microsecond | TimeUnit::Nanosecond => {
Err(DataFusionError::NotImplemented(format!(
"Cannot convert datatype {:?} to array",
f.data_type()
)))
}
},
DataType::Time64(unit) => match unit {
TimeUnit::Second | TimeUnit::Millisecond => {
Err(DataFusionError::NotImplemented(format!(
"Cannot convert datatype {:?} to array",
f.data_type()
)))
}
TimeUnit::Microsecond => {
Ok(Arc::new(Time64MicrosecondArray::from(vec![] as Vec<i64>))
as ArrayRef)
}
TimeUnit::Nanosecond => {
Ok(Arc::new(Time64NanosecondArray::from(vec![] as Vec<i64>))
as ArrayRef)
}
},
DataType::List(nested_type) => Ok(build_empty_list_array::<i32>(
nested_type.data_type().clone(),
)?),
DataType::LargeList(nested_type) => Ok(build_empty_list_array::<i64>(
nested_type.data_type().clone(),
)?),
DataType::FixedSizeList(nested_type, _) => Ok(
build_empty_fixed_size_list_array(nested_type.data_type().clone())?,
),
_ => Err(DataFusionError::NotImplemented(format!(
"Cannot convert datatype {:?} to array",
f.data_type()
))),
})
.collect::<Result<_>>()
.map_err(DataFusionError::into_arrow_external_error)?;
},
DataType::List(nested_type) => Ok(build_empty_list_array::<i32>(
nested_type.data_type().clone(),
)?),
DataType::LargeList(nested_type) => Ok(build_empty_list_array::<i64>(
nested_type.data_type().clone(),
)?),
DataType::FixedSizeList(nested_type, _) => Ok(build_empty_fixed_size_list_array(
nested_type.data_type().clone(),
)?),
DataType::Dictionary(key_type, value_type) => match key_type.as_ref() {
DataType::UInt8 => build_empty_dictionary::<UInt8Type>(value_type),
DataType::UInt16 => build_empty_dictionary::<UInt16Type>(value_type),
DataType::UInt32 => build_empty_dictionary::<UInt32Type>(value_type),
DataType::UInt64 => build_empty_dictionary::<UInt64Type>(value_type),
DataType::Int8 => build_empty_dictionary::<Int8Type>(value_type),
DataType::Int16 => build_empty_dictionary::<Int16Type>(value_type),
DataType::Int32 => build_empty_dictionary::<Int32Type>(value_type),
DataType::Int64 => build_empty_dictionary::<Int64Type>(value_type),
_ => unreachable!(),
},
_ => Err(DataFusionError::NotImplemented(format!(
"Creating empty array for type {:?} is not yet implemented",
data_type
))),
}
}

RecordBatch::try_new(Arc::new(schema.to_owned()), columns)
fn build_empty_dictionary<T: ArrowPrimitiveType>(
value_type: &DataType,
) -> Result<ArrayRef> {
let values: ArrayRef = create_empty_array(value_type)?;
let mut keys_builder: PrimitiveBuilder<T> = PrimitiveBuilder::new(0);
let dict_array = keys_builder.finish_dict(values);
Ok(Arc::new(dict_array))
}

#[cfg(test)]
Expand All @@ -254,41 +269,51 @@ mod tests {

#[test]
fn test_create_batch_empty() {
use DataType::*;

let schema = Schema::new(vec![
Field::new("c1", DataType::Utf8, false),
Field::new("c2", DataType::UInt32, false),
Field::new("c3", DataType::Int8, false),
Field::new("c4", DataType::Int16, false),
Field::new("c5", DataType::Int32, false),
Field::new("c6", DataType::Int64, false),
Field::new("c7", DataType::UInt8, false),
Field::new("c8", DataType::UInt16, false),
Field::new("c9", DataType::UInt32, false),
Field::new("c10", DataType::UInt64, false),
Field::new("c11", DataType::Float32, false),
Field::new("c12", DataType::Float64, false),
Field::new("c13", DataType::Utf8, false),
Field::new("c14", DataType::Decimal(10, 10), false),
Field::new("c15", DataType::Timestamp(TimeUnit::Second, None), false),
Field::new("c1", Utf8, false),
Field::new("c2", UInt32, false),
Field::new("c3", Int8, false),
Field::new("c4", Int16, false),
Field::new("c5", Int32, false),
Field::new("c6", Int64, false),
Field::new("c7", UInt8, false),
Field::new("c8", UInt16, false),
Field::new("c9", UInt32, false),
Field::new("c10", UInt64, false),
Field::new("c11", Float32, false),
Field::new("c12", Float64, false),
Field::new("c13", Utf8, false),
Field::new("c14", Decimal(10, 10), false),
Field::new("c15", Timestamp(TimeUnit::Second, None), false),
Field::new("c16", Timestamp(TimeUnit::Microsecond, None), false),
Field::new("c17", Timestamp(TimeUnit::Millisecond, None), false),
Field::new("c18", Timestamp(TimeUnit::Nanosecond, None), false),
Field::new("c19", Boolean, false),
Field::new("20", Dictionary(Box::new(UInt8), Box::new(Utf8)), false),
Field::new("21", Dictionary(Box::new(UInt16), Box::new(Utf8)), false),
Field::new("22", Dictionary(Box::new(UInt32), Box::new(Utf8)), false),
Field::new("23", Dictionary(Box::new(UInt64), Box::new(Utf8)), false),
Field::new("24", Dictionary(Box::new(Int8), Box::new(Utf8)), false),
Field::new("25", Dictionary(Box::new(Int16), Box::new(Utf8)), false),
Field::new("26", Dictionary(Box::new(Int32), Box::new(Utf8)), false),
Field::new("27", Dictionary(Box::new(Int64), Box::new(Utf8)), false),
// try non string dictionary
Field::new("28", Dictionary(Box::new(UInt8), Box::new(Int64)), false),
Field::new(
"c16",
DataType::Timestamp(TimeUnit::Microsecond, None),
"29",
Dictionary(Box::new(UInt8), Box::new(LargeUtf8)),
false,
),
Field::new(
"c17",
DataType::Timestamp(TimeUnit::Millisecond, None),
false,
),
Field::new(
"c18",
DataType::Timestamp(TimeUnit::Nanosecond, None),
false,
),
Field::new("c19", DataType::Boolean, false),
]);

let batch = create_batch_empty(&schema).unwrap();
assert_eq!(batch.columns().len(), 19);
assert_eq!(batch.columns().len(), 29);
assert_eq!(batch.num_rows(), 0);

for (i, array) in batch.columns().iter().enumerate() {
assert_eq!(array.len(), 0, "Array[{}] was zero length", i);
}
}
}