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.

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

Always validate the array data (except the Decimal) when creating array in IPC reader #2547

Merged
merged 2 commits into from
Aug 23, 2022
Merged
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
60 changes: 31 additions & 29 deletions arrow/src/ipc/reader.rs
Original file line number Diff line number Diff line change
Expand Up @@ -411,12 +411,12 @@ fn create_primitive_array(
}
FixedSizeBinary(_) => {
// read 2 buffers: null buffer (optional) and data buffer
let builder = ArrayData::builder(data_type.clone())
ArrayData::builder(data_type.clone())
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

unsafe { builder.build_unchecked() }
.null_bit_buffer(null_buffer)
.build()
.unwrap()
}
Int8
| Int16
Expand All @@ -429,45 +429,45 @@ fn create_primitive_array(
| Interval(IntervalUnit::YearMonth) => {
if buffers[1].len() / 8 == length && length != 1 {
// interpret as a signed i64, and cast appropriately
let builder = ArrayData::builder(DataType::Int64)
let data = ArrayData::builder(DataType::Int64)
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

let data = unsafe { builder.build_unchecked() };
.null_bit_buffer(null_buffer)
.build()
.unwrap();
let values = Arc::new(Int64Array::from(data)) as ArrayRef;
// this cast is infallible, the unwrap is safe
let casted = cast(&values, data_type).unwrap();
casted.into_data()
} else {
let builder = ArrayData::builder(data_type.clone())
ArrayData::builder(data_type.clone())
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

unsafe { builder.build_unchecked() }
.null_bit_buffer(null_buffer)
.build()
.unwrap()
}
}
Float32 => {
if buffers[1].len() / 8 == length && length != 1 {
// interpret as a f64, and cast appropriately
let builder = ArrayData::builder(DataType::Float64)
let data = ArrayData::builder(DataType::Float64)
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

let data = unsafe { builder.build_unchecked() };
.null_bit_buffer(null_buffer)
.build()
.unwrap();
let values = Arc::new(Float64Array::from(data)) as ArrayRef;
// this cast is infallible, the unwrap is safe
let casted = cast(&values, data_type).unwrap();
casted.into_data()
} else {
let builder = ArrayData::builder(data_type.clone())
ArrayData::builder(data_type.clone())
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

unsafe { builder.build_unchecked() }
.null_bit_buffer(null_buffer)
.build()
.unwrap()
}
}
Boolean
Expand All @@ -479,21 +479,23 @@ fn create_primitive_array(
| Date64
| Duration(_)
| Interval(IntervalUnit::DayTime)
| Interval(IntervalUnit::MonthDayNano) => {
let builder = ArrayData::builder(data_type.clone())
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

unsafe { builder.build_unchecked() }
}
| Interval(IntervalUnit::MonthDayNano) => ArrayData::builder(data_type.clone())
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer)
.build()
.unwrap(),
Decimal128(_, _) | Decimal256(_, _) => {
// read 2 buffers: null buffer (optional) and data buffer
let builder = ArrayData::builder(data_type.clone())
.len(length)
.add_buffer(buffers[1].clone())
.null_bit_buffer(null_buffer);

// Don't validate the decimal array so far,
// becasue validating decimal is some what complicated
// and there is no conclusion on whether we should do it.
// For more infomation, please look at https://github.com/apache/arrow-rs/issues/2387
unsafe { builder.build_unchecked() }
}
t => unreachable!("Data type {:?} either unsupported or not primitive", t),
Expand Down Expand Up @@ -527,7 +529,7 @@ fn create_list_array(

_ => unreachable!("Cannot create list or map array from {:?}", data_type),
};
make_array(unsafe { builder.build_unchecked() })
make_array(builder.build().unwrap())
}

/// Reads the correct number of buffers based on list type and null_count, and creates a
Expand Down