Skip to content

chore: refactor scalarvalue/encoding using available upstream arrow-rs methods #19797

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

Open
Jefffrey wants to merge 1 commit into
base: main
Choose a base branch
from
+8 −101
Open

chore: refactor scalarvalue/encoding using available upstream arrow-rs methods #19797

Show file tree
Hide file tree
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
16 changes: 4 additions & 12 deletions datafusion/common/src/scalar/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2989,13 +2989,8 @@ impl ScalarValue {
},
ScalarValue::Utf8View(e) => match e {
Some(value) => {
let mut builder =
StringViewBuilder::with_capacity(size).with_deduplicate_strings();
// Replace with upstream arrow-rs code when available:
// https://github.com/apache/arrow-rs/issues/9034
for _ in 0..size {
builder.append_value(value);
}
let mut builder = StringViewBuilder::with_capacity(size);
builder.try_append_value_n(value, size)?;
let array = builder.finish();
Arc::new(array)
}
Expand All @@ -3013,11 +3008,8 @@ impl ScalarValue {
},
ScalarValue::BinaryView(e) => match e {
Some(value) => {
let mut builder =
BinaryViewBuilder::with_capacity(size).with_deduplicate_strings();
for _ in 0..size {
builder.append_value(value);
}
let mut builder = BinaryViewBuilder::with_capacity(size);
builder.try_append_value_n(value, size)?;
let array = builder.finish();
Arc::new(array)
}
Expand Down
93 changes: 4 additions & 89 deletions datafusion/functions/src/encoding/inner.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,8 +19,8 @@

use arrow::{
array::{
Array, ArrayRef, AsArray, BinaryArrayType, FixedSizeBinaryArray,
GenericBinaryArray, GenericStringArray, OffsetSizeTrait,
Array, ArrayRef, AsArray, BinaryArrayType, GenericBinaryArray,
GenericStringArray, OffsetSizeTrait,
},
datatypes::DataType,
};
Expand Down Expand Up @@ -239,7 +239,7 @@ fn encode_array(array: &ArrayRef, encoding: Encoding) -> Result<ColumnarValue> {
encoding.encode_array::<_, i64>(&array.as_binary::<i64>())
}
DataType::FixedSizeBinary(_) => {
encoding.encode_fsb_array(array.as_fixed_size_binary())
encoding.encode_array::<_, i32>(&array.as_fixed_size_binary())
}
dt => {
internal_err!("Unexpected data type for encode: {dt}")
Expand Down Expand Up @@ -307,7 +307,7 @@ fn decode_array(array: &ArrayRef, encoding: Encoding) -> Result<ColumnarValue> {
let array = array.as_fixed_size_binary();
// TODO: could we be more conservative by accounting for nulls?
let estimate = array.len().saturating_mul(*size as usize);
encoding.decode_fsb_array(array, estimate)
encoding.decode_array::<_, i32>(&array, estimate)
}
dt => {
internal_err!("Unexpected data type for decode: {dt}")
Expand Down Expand Up @@ -404,24 +404,6 @@ impl Encoding {
}
}

// TODO: refactor this away once https://github.com/apache/arrow-rs/pull/8993 lands
fn encode_fsb_array(self, array: &FixedSizeBinaryArray) -> Result<ArrayRef> {
match self {
Self::Base64 => {
let array: GenericStringArray<i32> = array
.iter()
.map(|x| x.map(|x| BASE64_ENGINE.encode(x)))
.collect();
Ok(Arc::new(array))
}
Self::Hex => {
let array: GenericStringArray<i32> =
array.iter().map(|x| x.map(hex::encode)).collect();
Ok(Arc::new(array))
}
}
}

// OutputOffset important to ensure Large types output Large arrays
fn decode_array<'a, InputBinaryArray, OutputOffset>(
self,
Expand Down Expand Up @@ -461,73 +443,6 @@ impl Encoding {
}
}
}

// TODO: refactor this away once https://github.com/apache/arrow-rs/pull/8993 lands
fn decode_fsb_array(
self,
value: &FixedSizeBinaryArray,
approx_data_size: usize,
) -> Result<ArrayRef> {
fn hex_decode(input: &[u8], buf: &mut [u8]) -> Result<usize> {
// only write input / 2 bytes to buf
let out_len = input.len() / 2;
let buf = &mut buf[..out_len];
hex::decode_to_slice(input, buf)
.map_err(|e| exec_datafusion_err!("Failed to decode from hex: {e}"))?;
Ok(out_len)
}

fn base64_decode(input: &[u8], buf: &mut [u8]) -> Result<usize> {
BASE64_ENGINE
.decode_slice(input, buf)
.map_err(|e| exec_datafusion_err!("Failed to decode from base64: {e}"))
}

fn delegated_decode<DecodeFunction>(
decode: DecodeFunction,
input: &FixedSizeBinaryArray,
conservative_upper_bound_size: usize,
) -> Result<ArrayRef>
where
DecodeFunction: Fn(&[u8], &mut [u8]) -> Result<usize>,
{
let mut values = vec![0; conservative_upper_bound_size];
let mut offsets = OffsetBufferBuilder::new(input.len());
let mut total_bytes_decoded = 0;
for v in input.iter() {
if let Some(v) = v {
let cursor = &mut values[total_bytes_decoded..];
let decoded = decode(v, cursor)?;
total_bytes_decoded += decoded;
offsets.push_length(decoded);
} else {
offsets.push_length(0);
}
}
// We reserved an upper bound size for the values buffer, but we only use the actual size
values.truncate(total_bytes_decoded);
let binary_array = GenericBinaryArray::<i32>::try_new(
offsets.finish(),
Buffer::from_vec(values),
input.nulls().cloned(),
)?;
Ok(Arc::new(binary_array))
}

match self {
Self::Base64 => {
let upper_bound = base64::decoded_len_estimate(approx_data_size);
delegated_decode(base64_decode, value, upper_bound)
}
Self::Hex => {
// Calculate the upper bound for decoded byte size
// For hex encoding, each pair of hex characters (2 bytes) represents 1 byte when decoded
// So the upper bound is half the length of the input values.
let upper_bound = approx_data_size / 2;
delegated_decode(hex_decode, value, upper_bound)
}
}
}
}

fn delegated_decode<'a, DecodeFunction, InputBinaryArray, OutputOffset>(
Expand Down