Skip to content

[format] Introduce byte stream split for writing to parquet #9499

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
JingsongLi merged 1 commit into from
Sep 2, 2026
Merged

[format] Introduce byte stream split for writing to parquet #9499

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
6 changes: 6 additions & 0 deletions ...n-format/src/main/java/org/apache/paimon/format/parquet/writer/RowDataParquetBuilder.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,8 @@
/** A {@link ParquetBuilder} for {@link InternalRow}. */
public class RowDataParquetBuilder implements MetadataParquetBuilder<InternalRow> {

private static final String ENABLE_BYTE_STREAM_SPLIT = "parquet.enable.bytestreamsplit";

private final RowType rowType;
private final Configuration conf;

Expand Down Expand Up @@ -96,6 +98,10 @@ public ParquetWriter<InternalRow> createWriter(
conf.getBoolean(
ParquetOutputFormat.ENABLE_DICTIONARY,
ParquetProperties.DEFAULT_IS_DICTIONARY_ENABLED))
.withByteStreamSplitEncoding(
conf.getBoolean(
ENABLE_BYTE_STREAM_SPLIT,
ParquetProperties.DEFAULT_IS_BYTE_STREAM_SPLIT_ENABLED))
.withValidation(conf.getBoolean(ParquetOutputFormat.VALIDATION, false))
.withWriterVersion(
ParquetProperties.WriterVersion.fromString(
Expand Down
41 changes: 41 additions & 0 deletions paimon-format/src/test/java/org/apache/paimon/format/parquet/ParquetFormatReadWriteTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,6 +40,7 @@
import org.apache.paimon.types.DataTypes;
import org.apache.paimon.types.RowType;

import org.apache.parquet.column.Encoding;
import org.apache.parquet.column.values.bloomfilter.BloomFilter;
import org.apache.parquet.hadoop.ParquetFileReader;
import org.apache.parquet.hadoop.metadata.BlockMetaData;
Expand All @@ -52,6 +53,7 @@
import org.junit.jupiter.params.provider.ValueSource;

import java.nio.charset.StandardCharsets;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.LinkedHashMap;
import java.util.List;
Expand Down Expand Up @@ -292,4 +294,43 @@ public void testColumnCompressionCodec() throws Exception {
.containsEntry("name", CompressionCodecName.UNCOMPRESSED);
}
}

@Test
public void testWriteByteStreamSplit() throws Exception {
Options options = new Options();
options.set("parquet.enable.dictionary", "false");
options.set("parquet.enable.bytestreamsplit", "true");
ParquetFileFormat format =
new ParquetFileFormat(new FileFormatFactory.FormatContext(options, 1024, 1024));
RowType rowType =
DataTypes.ROW(
DataTypes.FIELD(0, "float_value", DataTypes.FLOAT()),
DataTypes.FIELD(1, "double_value", DataTypes.DOUBLE()));

write(
format.createWriterFactory(rowType),
file,
GenericRow.of(1.25f, 2.5d),
GenericRow.of(3.75f, 5.0d));

try (ParquetFileReader reader =
ParquetUtil.getParquetReader(
fileIO, file, fileIO.getFileSize(file), new Options())) {
for (ColumnChunkMetaData column : reader.getFooter().getBlocks().get(0).getColumns()) {
Assertions.assertThat(column.getEncodings()).contains(Encoding.BYTE_STREAM_SPLIT);
}
}

try (RecordReader<InternalRow> reader =
format.createReaderFactory(rowType, rowType, java.util.Collections.emptyList())
.createReader(
new FormatReaderContext(
fileIO, file, fileIO.getFileSize(file), null, null))) {
InternalRowSerializer serializer = new InternalRowSerializer(rowType);
List<InternalRow> rows = new ArrayList<>();
reader.forEachRemaining(row -> rows.add(serializer.copy(row)));
Assertions.assertThat(rows)
.containsExactly(GenericRow.of(1.25f, 2.5d), GenericRow.of(3.75f, 5.0d));
}
}
}
Loading