Skip to content
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

Jump to bottom

perf: Add ArrayChunks to optimize codegen of BatchDecoder #17632

Merged
merged 1 commit into from
Jul 15, 2024
Merged

perf: Add ArrayChunks to optimize codegen of BatchDecoder #17632

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
4 changes: 2 additions & 2 deletions crates/polars-parquet/src/arrow/read/deserialize/binary/basic.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ impl<'a, O: Offset> StateTranslation<'a, BinaryDecoder<O>> for BinaryStateTransl
values,
page_values,
)?,
(T::Dictionary(page), None) => {
(T::Dictionary(page, _), None) => {
// Already done on the dict.
validate_utf8 = false;
let page_dict = &page.dict;
Expand All @@ -90,7 +90,7 @@ impl<'a, O: Offset> StateTranslation<'a, BinaryDecoder<O>> for BinaryStateTransl
}
page.values.get_result()?;
},
(T::Dictionary(page), Some(page_validity)) => {
(T::Dictionary(page, _), Some(page_validity)) => {
// Already done on the dict.
validate_utf8 = false;
let page_dict = &page.dict;
Expand Down
9 changes: 5 additions & 4 deletions crates/polars-parquet/src/arrow/read/deserialize/binary/decoders.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
use arrow::array::specification::try_check_utf8;
use arrow::array::{BinaryArray, MutableBinaryValuesArray};
use arrow::array::{BinaryArray, MutableBinaryValuesArray, View};
use polars_error::PolarsResult;

use super::super::utils;
Expand Down Expand Up @@ -140,7 +140,7 @@ impl<'a> ValuesDictionary<'a> {
#[derive(Debug)]
pub(crate) enum BinaryStateTranslation<'a> {
Unit(BinaryIter<'a>),
Dictionary(ValuesDictionary<'a>),
Dictionary(ValuesDictionary<'a>, Option<Vec<View>>),
Delta(Delta<'a>),
DeltaBytes(DeltaBytes<'a>),
}
Expand All @@ -160,6 +160,7 @@ impl<'a> BinaryStateTranslation<'a> {
}
Ok(BinaryStateTranslation::Dictionary(
ValuesDictionary::try_new(page, dict)?,
None,
))
},
(Encoding::Plain, _) => {
Expand All @@ -180,7 +181,7 @@ impl<'a> BinaryStateTranslation<'a> {
pub(crate) fn len_when_not_nullable(&self) -> usize {
match self {
Self::Unit(v) => v.len_when_not_nullable(),
Self::Dictionary(v) => v.len(),
Self::Dictionary(v, _) => v.len(),
Self::Delta(v) => v.len(),
Self::DeltaBytes(v) => v.size_hint().0,
}
Expand All @@ -193,7 +194,7 @@ impl<'a> BinaryStateTranslation<'a> {

match self {
Self::Unit(t) => _ = t.by_ref().nth(n - 1),
Self::Dictionary(t) => t.values.skip_in_place(n)?,
Self::Dictionary(t, _) => t.values.skip_in_place(n)?,
Self::Delta(t) => _ = t.by_ref().nth(n - 1),
Self::DeltaBytes(t) => _ = t.by_ref().nth(n - 1),
}
Expand Down
16 changes: 8 additions & 8 deletions crates/polars-parquet/src/arrow/read/deserialize/binview/basic.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -69,13 +69,13 @@ impl<'a> StateTranslation<'a, BinViewDecoder> for BinaryStateTranslation<'a> {
values,
page_values,
)?,
(Self::Dictionary(page), None) => {
(Self::Dictionary(page, views_dict), None) => {
// Already done on the dict.
validate_utf8 = false;

let page_dict = &page.dict;
let views_dict = binary_views_dict(values, page_dict);
let translator = DictionaryTranslator(&views_dict);
let views_dict =
views_dict.get_or_insert_with(|| binary_views_dict(values, page.dict));
let translator = DictionaryTranslator(views_dict);

page.values.translate_and_collect_n_into(
values.views_mut(),
Expand All @@ -86,13 +86,13 @@ impl<'a> StateTranslation<'a, BinViewDecoder> for BinaryStateTranslation<'a> {
validity.extend_constant(additional, true);
}
},
(Self::Dictionary(page), Some(page_validity)) => {
(Self::Dictionary(page, views_dict), Some(page_validity)) => {
// Already done on the dict.
validate_utf8 = false;

let page_dict = &page.dict;
let views_dict = binary_views_dict(values, page_dict);
let translator = DictionaryTranslator(&views_dict);
let views_dict =
views_dict.get_or_insert_with(|| binary_views_dict(values, page.dict));
let translator = DictionaryTranslator(views_dict);
let collector = TranslatedHybridRle::new(&mut page.values, &translator);

extend_from_decoder(validity, page_validity, Some(additional), values, collector)?;
Expand Down
16 changes: 16 additions & 0 deletions crates/polars-parquet/src/arrow/read/deserialize/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,20 @@
//! APIs to read from Parquet format.

macro_rules! decoder_fn {
(($x:ident $(, $field:ident:$ty:ty)* $(,)?) => <$p:ty, $t:ty> => $expr:expr) => {{
#[derive(Clone, Copy)]
struct DecoderFn($($ty),*);
impl crate::arrow::read::deserialize::primitive::DecoderFunction<$p, $t> for DecoderFn {
#[inline(always)]
fn decode(self, $x: $p) -> $t {
let Self($($field),*) = self;
$expr
}
}
DecoderFn($($field),*)
}};
}

mod binary;
mod binview;
mod boolean;
Expand Down
49 changes: 25 additions & 24 deletions crates/polars-parquet/src/arrow/read/deserialize/nested.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ use arrow::match_integer_type;
use ethnum::I256;
use polars_error::polars_bail;

use self::primitive::{AsDecoderFunction, IntoDecoderFunction, UnitDecoderFunction};
use super::*;

/// Converts an iterator of arrays to a trait object returning trait objects
Expand Down Expand Up @@ -80,7 +81,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i32| x as i8,
AsDecoderFunction::<i32, i8>::default(),
))
},
Primitive(Int16) => {
Expand All @@ -92,7 +93,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i32| x as i16,
AsDecoderFunction::<i32, i16>::default(),
))
},
Primitive(Int32) => {
Expand All @@ -104,7 +105,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i32| x,
UnitDecoderFunction::<i32>::default(),
))
},
Primitive(Int64) => {
Expand All @@ -116,7 +117,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i64| x,
UnitDecoderFunction::<i64>::default(),
))
},
Primitive(UInt8) => {
Expand All @@ -128,7 +129,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i32| x as u8,
AsDecoderFunction::<i32, u8>::default(),
))
},
Primitive(UInt16) => {
Expand All @@ -140,7 +141,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i32| x as u16,
AsDecoderFunction::<i32, u16>::default(),
))
},
Primitive(UInt32) => {
Expand All @@ -153,7 +154,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i32| x as u32,
AsDecoderFunction::<i32, u32>::default(),
)),
// some implementations of parquet write arrow's u32 into i64.
PhysicalType::Int64 => primitive(primitive::NestedIter::new(
Expand All @@ -162,7 +163,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i64| x as u32,
AsDecoderFunction::<i64, u32>::default(),
)),
other => {
polars_bail!(ComputeError:
Expand All @@ -180,7 +181,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: i64| x as u64,
AsDecoderFunction::<i64, u64>::default(),
))
},
Primitive(Float32) => {
Expand All @@ -192,7 +193,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: f32| x,
UnitDecoderFunction::<f32>::default(),
))
},
Primitive(Float64) => {
Expand All @@ -204,7 +205,7 @@ where
field.data_type().clone(),
num_rows,
chunk_size,
|x: f64| x,
UnitDecoderFunction::<f64>::default(),
))
},
BinaryView | Utf8View => {
Expand Down Expand Up @@ -283,15 +284,15 @@ where
field.data_type.clone(),
num_rows,
chunk_size,
|x: i32| x as i128,
IntoDecoderFunction::<i32, i128>::default(),
)),
PhysicalType::Int64 => primitive(primitive::NestedIter::new(
columns.pop().unwrap(),
init,
field.data_type.clone(),
num_rows,
chunk_size,
|x: i64| x as i128,
IntoDecoderFunction::<i64, i128>::default(),
)),
PhysicalType::FixedLenByteArray(n) if n > 16 => {
polars_bail!(
Expand Down Expand Up @@ -346,15 +347,15 @@ where
field.data_type.clone(),
num_rows,
chunk_size,
|x: i32| i256(I256::new(x as i128)),
decoder_fn!((x) => <i32, i256> => i256(I256::new(x as i128))),
)),
PhysicalType::Int64 => primitive(primitive::NestedIter::new(
columns.pop().unwrap(),
init,
field.data_type.clone(),
num_rows,
chunk_size,
|x: i64| i256(I256::new(x as i128)),
decoder_fn!((x) => <i64, i256> => i256(I256::new(x as i128))),
)),
PhysicalType::FixedLenByteArray(n) if n <= 16 => {
let iter = fixed_size_binary::NestedIter::new(
Expand Down Expand Up @@ -501,39 +502,39 @@ fn dict_read<'a, K: DictionaryKey, I: 'a + PagesIter>(
data_type,
num_rows,
chunk_size,
|x: i32| x as u8,
AsDecoderFunction::<i32, u8>::default(),
)),
UInt16 => primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
iter,
init,
data_type,
num_rows,
chunk_size,
|x: i32| x as u16,
AsDecoderFunction::<i32, u16>::default(),
)),
UInt32 => primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
iter,
init,
data_type,
num_rows,
chunk_size,
|x: i32| x as u32,
AsDecoderFunction::<i32, u32>::default(),
)),
Int8 => primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
iter,
init,
data_type,
num_rows,
chunk_size,
|x: i32| x as i8,
AsDecoderFunction::<i32, i8>::default(),
)),
Int16 => primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
iter,
init,
data_type,
num_rows,
chunk_size,
|x: i32| x as i16,
AsDecoderFunction::<i32, i16>::default(),
)),
Int32 | Date32 | Time32(_) | Interval(IntervalUnit::YearMonth) => {
primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
Expand All @@ -542,7 +543,7 @@ fn dict_read<'a, K: DictionaryKey, I: 'a + PagesIter>(
data_type,
num_rows,
chunk_size,
|x: i32| x,
UnitDecoderFunction::<i32>::default(),
))
},
Int64 | Date64 | Time64(_) | Duration(_) => {
Expand All @@ -552,7 +553,7 @@ fn dict_read<'a, K: DictionaryKey, I: 'a + PagesIter>(
data_type,
num_rows,
chunk_size,
|x: i64| x as i32,
AsDecoderFunction::<i64, i32>::default(),
))
},
Float32 => primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
Expand All @@ -561,15 +562,15 @@ fn dict_read<'a, K: DictionaryKey, I: 'a + PagesIter>(
data_type,
num_rows,
chunk_size,
|x: f32| x,
UnitDecoderFunction::<f32>::default(),
)),
Float64 => primitive(primitive::NestedDictIter::<K, _, _, _, _>::new(
iter,
init,
data_type,
num_rows,
chunk_size,
|x: f64| x,
UnitDecoderFunction::<f64>::default(),
)),
LargeUtf8 | LargeBinary => primitive(binary::NestedDictIter::<K, i64, _>::new(
iter, init, data_type, num_rows, chunk_size,
Expand Down
Loading