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

Fix parquet definition levels #511

Merged
merged 2 commits into from
Jul 3, 2021
Merged

Fix parquet definition levels #511

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4885c32
Fix parquet definition levels
nevi-me Jun 30, 2021
9f7d2f9
roundtrip tests
nevi-me Jul 3, 2021
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
152 changes: 70 additions & 82 deletions parquet/src/arrow/arrow_writer.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -563,8 +563,8 @@ fn get_fsb_array_slice(
mod tests {
use super::*;

use std::fs::File;
use std::sync::Arc;
use std::{fs::File, io::Seek};

use arrow::datatypes::ToByteSlice;
use arrow::datatypes::{DataType, Field, Schema, UInt32Type, UInt8Type};
Expand Down Expand Up @@ -592,16 +592,11 @@ mod tests {
let b = Int32Array::from(vec![Some(1), None, None, Some(4), Some(5)]);

// build a record batch
let batch = RecordBatch::try_new(
Arc::new(schema.clone()),
vec![Arc::new(a), Arc::new(b)],
)
.unwrap();
let batch =
RecordBatch::try_new(Arc::new(schema), vec![Arc::new(a), Arc::new(b)])
.unwrap();

let file = get_temp_file("test_arrow_writer.parquet", &[]);
let mut writer = ArrowWriter::try_new(file, Arc::new(schema), None).unwrap();
writer.write(&batch).unwrap();
writer.close().unwrap();
roundtrip("test_arrow_write.parquet", batch, Some(SMALL_SIZE / 2));
}

#[test]
Expand Down Expand Up @@ -660,22 +655,22 @@ mod tests {
let a = Int32Array::from(vec![1, 2, 3, 4, 5]);

// build a record batch
let batch =
RecordBatch::try_new(Arc::new(schema.clone()), vec![Arc::new(a)]).unwrap();
let batch = RecordBatch::try_new(Arc::new(schema), vec![Arc::new(a)]).unwrap();

let file = get_temp_file("test_arrow_writer_non_null.parquet", &[]);
let mut writer = ArrowWriter::try_new(file, Arc::new(schema), None).unwrap();
writer.write(&batch).unwrap();
writer.close().unwrap();
roundtrip(
"test_arrow_writer_non_null.parquet",
batch,
Some(SMALL_SIZE / 2),
);
}

#[test]
fn arrow_writer_list() {
// define schema
let schema = Schema::new(vec![Field::new(
"a",
DataType::List(Box::new(Field::new("item", DataType::Int32, true))),
false,
DataType::List(Box::new(Field::new("item", DataType::Int32, false))),
true,
)]);

// create some data
Expand All @@ -690,7 +685,7 @@ mod tests {
let a_list_data = ArrayData::builder(DataType::List(Box::new(Field::new(
"item",
DataType::Int32,
true,
false,
))))
.len(5)
.add_buffer(a_value_offsets)
Expand All @@ -700,16 +695,52 @@ mod tests {
let a = ListArray::from(a_list_data);

// build a record batch
let batch =
RecordBatch::try_new(Arc::new(schema.clone()), vec![Arc::new(a)]).unwrap();
let batch = RecordBatch::try_new(Arc::new(schema), vec![Arc::new(a)]).unwrap();

// I think this setup is incorrect because this should pass
assert_eq!(batch.column(0).data().null_count(), 1);

let file = get_temp_file("test_arrow_writer_list.parquet", &[]);
let mut writer = ArrowWriter::try_new(file, Arc::new(schema), None).unwrap();
writer.write(&batch).unwrap();
writer.close().unwrap();
// This test fails if the max row group size is less than the batch's length
// see https://github.com/apache/arrow-rs/issues/518
roundtrip("test_arrow_writer_list.parquet", batch, None);
}

#[test]
fn arrow_writer_list_non_null() {
// define schema
let schema = Schema::new(vec![Field::new(
"a",
DataType::List(Box::new(Field::new("item", DataType::Int32, false))),
false,
)]);

// create some data
let a_values = Int32Array::from(vec![1, 2, 3, 4, 5, 6, 7, 8, 9, 10]);

// Construct a buffer for value offsets, for the nested array:
// [[1], [2, 3], [], [4, 5, 6], [7, 8, 9, 10]]
let a_value_offsets =
arrow::buffer::Buffer::from(&[0, 1, 3, 3, 6, 10].to_byte_slice());

// Construct a list array from the above two
let a_list_data = ArrayData::builder(DataType::List(Box::new(Field::new(
"item",
DataType::Int32,
false,
))))
.len(5)
.add_buffer(a_value_offsets)
.add_child_data(a_values.data().clone())
.build();
let a = ListArray::from(a_list_data);

// build a record batch
let batch = RecordBatch::try_new(Arc::new(schema), vec![Arc::new(a)]).unwrap();

// This test fails if the max row group size is less than the batch's length
// see https://github.com/apache/arrow-rs/issues/518
assert_eq!(batch.column(0).data().null_count(), 0);

roundtrip("test_arrow_writer_list_non_null.parquet", batch, None);
}

#[test]
Expand All @@ -733,39 +764,16 @@ mod tests {
let string_values = StringArray::from(raw_string_values.clone());
let binary_values = BinaryArray::from(raw_binary_value_refs);
let batch = RecordBatch::try_new(
Arc::new(schema.clone()),
Arc::new(schema),
vec![Arc::new(string_values), Arc::new(binary_values)],
)
.unwrap();

let mut file = get_temp_file("test_arrow_writer_binary.parquet", &[]);
let mut writer =
ArrowWriter::try_new(file.try_clone().unwrap(), Arc::new(schema), None)
.unwrap();
writer.write(&batch).unwrap();
writer.close().unwrap();

file.seek(std::io::SeekFrom::Start(0)).unwrap();
let file_reader = SerializedFileReader::new(file).unwrap();
let mut arrow_reader = ParquetFileArrowReader::new(Arc::new(file_reader));
let mut record_batch_reader = arrow_reader.get_record_reader(1024).unwrap();

let batch = record_batch_reader.next().unwrap().unwrap();
let string_col = batch
.column(0)
.as_any()
.downcast_ref::<StringArray>()
.unwrap();
let binary_col = batch
.column(1)
.as_any()
.downcast_ref::<BinaryArray>()
.unwrap();

for i in 0..batch.num_rows() {
assert_eq!(string_col.value(i), raw_string_values[i]);
assert_eq!(binary_col.value(i), raw_binary_values[i].as_slice());
}
roundtrip(
"test_arrow_writer_binary.parquet",
batch,
Some(SMALL_SIZE / 2),
);
}

#[test]
Expand All @@ -779,36 +787,16 @@ mod tests {
dec_builder.append_value(0).unwrap();
dec_builder.append_value(-100).unwrap();

let raw_decimal_i128_values: Vec<i128> = vec![10_000, 50_000, 0, -100];
let decimal_values = dec_builder.finish();
let batch = RecordBatch::try_new(
Arc::new(schema.clone()),
vec![Arc::new(decimal_values)],
)
.unwrap();

let mut file = get_temp_file("test_arrow_writer_decimal.parquet", &[]);
let mut writer =
ArrowWriter::try_new(file.try_clone().unwrap(), Arc::new(schema), None)
let batch =
RecordBatch::try_new(Arc::new(schema), vec![Arc::new(decimal_values)])
.unwrap();
writer.write(&batch).unwrap();
writer.close().unwrap();

file.seek(std::io::SeekFrom::Start(0)).unwrap();
let file_reader = SerializedFileReader::new(file).unwrap();
let mut arrow_reader = ParquetFileArrowReader::new(Arc::new(file_reader));
let mut record_batch_reader = arrow_reader.get_record_reader(1024).unwrap();

let batch = record_batch_reader.next().unwrap().unwrap();
let decimal_col = batch
.column(0)
.as_any()
.downcast_ref::<DecimalArray>()
.unwrap();

for i in 0..batch.num_rows() {
assert_eq!(decimal_col.value(i), raw_decimal_i128_values[i]);
}
roundtrip(
"test_arrow_writer_decimal.parquet",
batch,
Some(SMALL_SIZE / 2),
);
}

#[test]
Expand Down
Loading