Skip to content

make_decoder accepts borrowed DataType instead of owned #9270

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

Merged
alamb merged 4 commits into from
Jan 27, 2026
Merged

make_decoder accepts borrowed DataType instead of owned #9270

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
03ae73b
make_decoder accepts borrowed DataType instead of owned
scovich Jan 26, 2026
b558aae
fmt
scovich Jan 26, 2026
6e2340d
cleanup
scovich Jan 26, 2026
0d13528
Merge remote-tracking branch 'oss/main' into json-decoder-datatype-refs
scovich Jan 26, 2026
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
8 changes: 4 additions & 4 deletions arrow-json/src/reader/list_array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,18 +34,18 @@ pub struct ListArrayDecoder<O> {
impl<O: OffsetSizeTrait> ListArrayDecoder<O> {
pub fn new(
ctx: &DecoderContext,
data_type: DataType,
data_type: &DataType,
is_nullable: bool,
) -> Result<Self, ArrowError> {
let field = match &data_type {
let field = match data_type {
DataType::List(f) if !O::IS_LARGE => f,
DataType::LargeList(f) if O::IS_LARGE => f,
_ => unreachable!(),
};
let decoder = ctx.make_decoder(field.data_type().clone(), field.is_nullable())?;
let decoder = ctx.make_decoder(field.data_type(), field.is_nullable())?;

Ok(Self {
data_type,
data_type: data_type.clone(),
decoder,
phantom: Default::default(),
is_nullable,
Expand Down
10 changes: 5 additions & 5 deletions arrow-json/src/reader/map_array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,10 +33,10 @@ pub struct MapArrayDecoder {
impl MapArrayDecoder {
pub fn new(
ctx: &DecoderContext,
data_type: DataType,
data_type: &DataType,
is_nullable: bool,
) -> Result<Self, ArrowError> {
let fields = match &data_type {
let fields = match data_type {
DataType::Map(_, true) => {
return Err(ArrowError::NotYetImplemented(
"Decoding MapArray with sorted fields".to_string(),
Expand All @@ -53,11 +53,11 @@ impl MapArrayDecoder {
_ => unreachable!(),
};

let keys = ctx.make_decoder(fields[0].data_type().clone(), fields[0].is_nullable())?;
let values = ctx.make_decoder(fields[1].data_type().clone(), fields[1].is_nullable())?;
let keys = ctx.make_decoder(fields[0].data_type(), fields[0].is_nullable())?;
let values = ctx.make_decoder(fields[1].data_type(), fields[1].is_nullable())?;

Ok(Self {
data_type,
data_type: data_type.clone(),
keys,
values,
is_nullable,
Expand Down
24 changes: 13 additions & 11 deletions arrow-json/src/reader/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -137,6 +137,7 @@ use crate::StructMode;
use crate::reader::binary_array::{
BinaryArrayDecoder, BinaryViewDecoder, FixedSizeBinaryArrayDecoder,
};
use std::borrow::Cow;
use std::io::BufRead;
use std::sync::Arc;

Expand Down Expand Up @@ -295,20 +296,21 @@ impl ReaderBuilder {

/// Create a [`Decoder`]
pub fn build_decoder(self) -> Result<Decoder, ArrowError> {
let (data_type, nullable) = match self.is_field {
false => (DataType::Struct(self.schema.fields.clone()), false),
true => {
let field = &self.schema.fields[0];
(field.data_type().clone(), field.is_nullable())
}
let (data_type, nullable) = if self.is_field {
let field = &self.schema.fields[0];
let data_type = Cow::Borrowed(field.data_type());
(data_type, field.is_nullable())
} else {
let data_type = Cow::Owned(DataType::Struct(self.schema.fields.clone()));
(data_type, false)
};

let ctx = DecoderContext {
coerce_primitive: self.coerce_primitive,
strict_mode: self.strict_mode,
struct_mode: self.struct_mode,
};
let decoder = ctx.make_decoder(data_type, nullable)?;
let decoder = ctx.make_decoder(data_type.as_ref(), nullable)?;

let num_fields = self.schema.flattened_fields().len();

Expand Down Expand Up @@ -713,7 +715,7 @@ impl DecoderContext {
/// implementation.
fn make_decoder(
&self,
data_type: DataType,
data_type: &DataType,
is_nullable: bool,
) -> Result<Box<dyn ArrayDecoder>, ArrowError> {
make_decoder(self, data_type, is_nullable)
Expand All @@ -728,12 +730,12 @@ macro_rules! primitive_decoder {

fn make_decoder(
ctx: &DecoderContext,
data_type: DataType,
data_type: &DataType,
is_nullable: bool,
) -> Result<Box<dyn ArrayDecoder>, ArrowError> {
let coerce_primitive = ctx.coerce_primitive();
downcast_integer! {
data_type => (primitive_decoder, data_type),
*data_type => (primitive_decoder, data_type),
DataType::Null => Ok(Box::<NullArrayDecoder>::default()),
DataType::Float16 => primitive_decoder!(Float16Type, data_type),
DataType::Float32 => primitive_decoder!(Float32Type, data_type),
Expand Down Expand Up @@ -792,7 +794,7 @@ fn make_decoder(
DataType::FixedSizeBinary(len) => Ok(Box::new(FixedSizeBinaryArrayDecoder::new(len))),
DataType::BinaryView => Ok(Box::new(BinaryViewDecoder::default())),
DataType::Map(_, _) => Ok(Box::new(MapArrayDecoder::new(ctx, data_type, is_nullable)?)),
d => Err(ArrowError::NotYetImplemented(format!("Support for {d} in JSON reader")))
_ => Err(ArrowError::NotYetImplemented(format!("Support for {data_type} in JSON reader")))
}
}

Expand Down
4 changes: 2 additions & 2 deletions arrow-json/src/reader/primitive_array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,9 +80,9 @@ pub struct PrimitiveArrayDecoder<P: ArrowPrimitiveType> {
}

impl<P: ArrowPrimitiveType> PrimitiveArrayDecoder<P> {
pub fn new(data_type: DataType) -> Self {
pub fn new(data_type: &DataType) -> Self {
Self {
data_type,
data_type: data_type.clone(),
phantom: Default::default(),
}
}
Expand Down
20 changes: 11 additions & 9 deletions arrow-json/src/reader/struct_array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -81,28 +81,30 @@ pub struct StructArrayDecoder {
impl StructArrayDecoder {
pub fn new(
ctx: &DecoderContext,
data_type: DataType,
data_type: &DataType,
is_nullable: bool,
) -> Result<Self, ArrowError> {
let struct_mode = ctx.struct_mode();
let fields = struct_fields(&data_type);

// If this struct nullable, need to permit nullability in child array
// StructArrayDecoder::decode verifies that if the child is not nullable
// it doesn't contain any nulls not masked by its parent
Comment on lines -90 to -92
Copy link
Contributor Author

@scovich scovich Jan 26, 2026 •
edited
Loading

Choose a reason for hiding this comment

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

#9271 merged too quickly... this comment was supposed to remain inside the .map call. So I'm restoring this code back to follow the original upstream approach, instead of opening a separate PR just for that.

Copy link
Contributor

@alamb alamb Jan 26, 2026

Choose a reason for hiding this comment

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

thank you

let fields = struct_fields(data_type);
let decoders = fields
.iter()
.map(|f| ctx.make_decoder(f.data_type().clone(), f.is_nullable() || is_nullable))
.map(|f| {
// If this struct nullable, need to permit nullability in child array
// StructArrayDecoder::decode verifies that if the child is not nullable
// it doesn't contain any nulls not masked by its parent
let nullable = f.is_nullable() || is_nullable;
ctx.make_decoder(f.data_type(), nullable)
})
.collect::<Result<Vec<_>, ArrowError>>()?;

let struct_mode = ctx.struct_mode();
let field_name_to_index = if struct_mode == StructMode::ObjectOnly {
build_field_index(fields)
} else {
None
};

Ok(Self {
data_type,
data_type: data_type.clone(),
decoders,
strict_mode: ctx.strict_mode(),
is_nullable,
Expand Down
4 changes: 2 additions & 2 deletions arrow-json/src/reader/timestamp_array.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,9 +37,9 @@ pub struct TimestampArrayDecoder<P: ArrowTimestampType, Tz: TimeZone> {
}

impl<P: ArrowTimestampType, Tz: TimeZone> TimestampArrayDecoder<P, Tz> {
pub fn new(data_type: DataType, timezone: Tz) -> Self {
pub fn new(data_type: &DataType, timezone: Tz) -> Self {
Self {
data_type,
data_type: data_type.clone(),
timezone,
phantom: Default::default(),
}
Expand Down
Loading