-
Notifications
You must be signed in to change notification settings - Fork 2.4k
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
Avro: Support default values for generic data #11786
Changes from all commits
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,187 @@ | ||
/* | ||
* Licensed to the Apache Software Foundation (ASF) under one | ||
* or more contributor license agreements. See the NOTICE file | ||
* distributed with this work for additional information | ||
* regarding copyright ownership. The ASF licenses this file | ||
* to you under the Apache License, Version 2.0 (the | ||
* "License"); you may not use this file except in compliance | ||
* with the License. You may obtain a copy of the License at | ||
* | ||
* http://www.apache.org/licenses/LICENSE-2.0 | ||
* | ||
* Unless required by applicable law or agreed to in writing, | ||
* software distributed under the License is distributed on an | ||
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY | ||
* KIND, either express or implied. See the License for the | ||
* specific language governing permissions and limitations | ||
* under the License. | ||
*/ | ||
package org.apache.iceberg.data.avro; | ||
|
||
import java.io.IOException; | ||
import java.util.List; | ||
import java.util.Map; | ||
import java.util.function.Supplier; | ||
import org.apache.avro.LogicalType; | ||
import org.apache.avro.LogicalTypes; | ||
import org.apache.avro.Schema; | ||
import org.apache.avro.io.DatumReader; | ||
import org.apache.avro.io.Decoder; | ||
import org.apache.iceberg.avro.AvroSchemaUtil; | ||
import org.apache.iceberg.avro.AvroWithPartnerVisitor; | ||
import org.apache.iceberg.avro.SupportsRowPosition; | ||
import org.apache.iceberg.avro.ValueReader; | ||
import org.apache.iceberg.avro.ValueReaders; | ||
import org.apache.iceberg.relocated.com.google.common.collect.ImmutableMap; | ||
import org.apache.iceberg.types.Type; | ||
import org.apache.iceberg.types.Types; | ||
import org.apache.iceberg.util.Pair; | ||
|
||
public class PlannedDataReader<T> implements DatumReader<T>, SupportsRowPosition { | ||
|
||
public static <D> PlannedDataReader<D> create(org.apache.iceberg.Schema expectedSchema) { | ||
return create(expectedSchema, ImmutableMap.of()); | ||
} | ||
|
||
public static <D> PlannedDataReader<D> create( | ||
org.apache.iceberg.Schema expectedSchema, Map<Integer, ?> idToConstant) { | ||
return new PlannedDataReader<>(expectedSchema, idToConstant); | ||
} | ||
|
||
private final org.apache.iceberg.Schema expectedSchema; | ||
private final Map<Integer, ?> idToConstant; | ||
private ValueReader<T> reader; | ||
|
||
protected PlannedDataReader( | ||
org.apache.iceberg.Schema expectedSchema, Map<Integer, ?> idToConstant) { | ||
this.expectedSchema = expectedSchema; | ||
this.idToConstant = idToConstant; | ||
} | ||
|
||
@Override | ||
@SuppressWarnings("unchecked") | ||
public void setSchema(Schema fileSchema) { | ||
this.reader = | ||
(ValueReader<T>) | ||
AvroWithPartnerVisitor.visit( | ||
expectedSchema.asStruct(), | ||
fileSchema, | ||
new ReadBuilder(idToConstant), | ||
AvroWithPartnerVisitor.FieldIDAccessors.get()); | ||
} | ||
|
||
@Override | ||
public T read(T reuse, Decoder decoder) throws IOException { | ||
return reader.read(decoder, reuse); | ||
} | ||
|
||
@Override | ||
public void setRowPositionSupplier(Supplier<Long> posSupplier) { | ||
if (reader instanceof SupportsRowPosition) { | ||
((SupportsRowPosition) reader).setRowPositionSupplier(posSupplier); | ||
} | ||
} | ||
|
||
private static class ReadBuilder extends AvroWithPartnerVisitor<Type, ValueReader<?>> { | ||
private final Map<Integer, ?> idToConstant; | ||
|
||
private ReadBuilder(Map<Integer, ?> idToConstant) { | ||
this.idToConstant = idToConstant; | ||
} | ||
|
||
@Override | ||
public ValueReader<?> record(Type partner, Schema record, List<ValueReader<?>> fieldReaders) { | ||
if (partner == null) { | ||
return ValueReaders.skipStruct(fieldReaders); | ||
} | ||
|
||
Types.StructType expected = partner.asStructType(); | ||
List<Pair<Integer, ValueReader<?>>> readPlan = | ||
ValueReaders.buildReadPlan(expected, record, fieldReaders, idToConstant); | ||
|
||
return GenericReaders.struct(readPlan, expected); | ||
} | ||
|
||
@Override | ||
public ValueReader<?> union(Type partner, Schema union, List<ValueReader<?>> options) { | ||
return ValueReaders.union(options); | ||
} | ||
|
||
@Override | ||
public ValueReader<?> array(Type ignored, Schema array, ValueReader<?> elementReader) { | ||
return ValueReaders.array(elementReader); | ||
} | ||
|
||
@Override | ||
public ValueReader<?> arrayMap( | ||
Type ignored, Schema map, ValueReader<?> keyReader, ValueReader<?> valueReader) { | ||
return ValueReaders.arrayMap(keyReader, valueReader); | ||
} | ||
|
||
@Override | ||
public ValueReader<?> map(Type ignored, Schema map, ValueReader<?> valueReader) { | ||
return ValueReaders.map(ValueReaders.strings(), valueReader); | ||
} | ||
|
||
@Override | ||
public ValueReader<?> primitive(Type partner, Schema primitive) { | ||
LogicalType logicalType = primitive.getLogicalType(); | ||
if (logicalType != null) { | ||
switch (logicalType.getName()) { | ||
case "date": | ||
return GenericReaders.dates(); | ||
|
||
case "time-micros": | ||
return GenericReaders.times(); | ||
|
||
case "timestamp-micros": | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. does there need to be case for nanos as well? There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. There does, but adding the timestamp nanos read path is a separate PR. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I noticed that we have this string all over the place, probably a good idea to make a constant for it: https://github.com/search?q=repo%3Aapache%2Ficeberg+timestamp-micros+language%3AJava&type=code&l=Java We can also do that in a separate PR. |
||
if (AvroSchemaUtil.isTimestamptz(primitive)) { | ||
return GenericReaders.timestamptz(); | ||
} | ||
return GenericReaders.timestamps(); | ||
|
||
case "decimal": | ||
return ValueReaders.decimal( | ||
ValueReaders.decimalBytesReader(primitive), | ||
((LogicalTypes.Decimal) logicalType).getScale()); | ||
|
||
case "uuid": | ||
return ValueReaders.uuids(); | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Separate PR, but I think we also need to take the physical type into account: https://github.com/apache/avro/blob/main/doc/content/en/docs/%2B%2Bversion%2B%2B/Specification/_index.md#uuid First Avro only supported UUIDs stored as strings, but now also as There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. The Iceberg spec only supports |
||
|
||
default: | ||
throw new IllegalArgumentException("Unknown logical type: " + logicalType); | ||
} | ||
} | ||
|
||
switch (primitive.getType()) { | ||
case NULL: | ||
return ValueReaders.nulls(); | ||
case BOOLEAN: | ||
return ValueReaders.booleans(); | ||
case INT: | ||
if (partner != null && partner.typeId() == Type.TypeID.LONG) { | ||
return ValueReaders.intsAsLongs(); | ||
} | ||
return ValueReaders.ints(); | ||
case LONG: | ||
return ValueReaders.longs(); | ||
case FLOAT: | ||
if (partner != null && partner.typeId() == Type.TypeID.DOUBLE) { | ||
return ValueReaders.floatsAsDoubles(); | ||
} | ||
return ValueReaders.floats(); | ||
case DOUBLE: | ||
return ValueReaders.doubles(); | ||
case STRING: | ||
// might want to use a binary-backed container like Utf8 | ||
return ValueReaders.strings(); | ||
case FIXED: | ||
return ValueReaders.fixed(primitive.getFixedSize()); | ||
case BYTES: | ||
return ValueReaders.byteBuffers(); | ||
default: | ||
throw new IllegalArgumentException("Unsupported type: " + primitive); | ||
} | ||
} | ||
} | ||
} |
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -33,6 +33,27 @@ | |
public class RawDecoder<D> extends MessageDecoder.BaseDecoder<D> { | ||
private static final ThreadLocal<BinaryDecoder> DECODER = new ThreadLocal<>(); | ||
|
||
/** | ||
* Creates a new {@link MessageDecoder} that constructs datum instances described by the {@link | ||
* Schema readSchema}. | ||
* | ||
* <p>The {@code readSchema} is used for the expected schema and the {@code writeSchema} is the | ||
* schema used to decode buffers. The {@code writeSchema} must be the schema that was used to | ||
* encode all buffers decoded by this class. | ||
* | ||
* @param readSchema an Iceberg schema to produce when reading | ||
* @param readerFunction a function that produces a DatumReader from the read schema | ||
* @param writeSchema an Avro schema that describes serialized data to be read | ||
*/ | ||
public static <D> RawDecoder<D> create( | ||
org.apache.iceberg.Schema readSchema, | ||
Function<org.apache.iceberg.Schema, DatumReader<D>> readerFunction, | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Nice! |
||
Schema writeSchema) { | ||
DatumReader<D> reader = readerFunction.apply(readSchema); | ||
reader.setSchema(writeSchema); | ||
return new RawDecoder<>(reader); | ||
} | ||
|
||
private final DatumReader<D> reader; | ||
|
||
/** | ||
|
@@ -42,7 +63,11 @@ public class RawDecoder<D> extends MessageDecoder.BaseDecoder<D> { | |
* <p>The {@code readSchema} is used for the expected schema and the {@code writeSchema} is the | ||
* schema used to decode buffers. The {@code writeSchema} must be the schema that was used to | ||
* encode all buffers decoded by this class. | ||
* | ||
* @deprecated will be removed in 2.0.0; use {@link #create(org.apache.iceberg.Schema, Function, | ||
* Schema)} instead | ||
*/ | ||
@Deprecated | ||
public RawDecoder( | ||
org.apache.iceberg.Schema readSchema, | ||
Function<Schema, DatumReader<?>> readerFunction, | ||
|
@@ -51,6 +76,13 @@ public RawDecoder( | |
this.reader.setSchema(writeSchema); | ||
} | ||
|
||
/** | ||
* Creates a new {@link MessageDecoder} that constructs datum instances using the {@code reader}. | ||
*/ | ||
private RawDecoder(DatumReader<D> reader) { | ||
this.reader = reader; | ||
} | ||
|
||
@Override | ||
public D decode(InputStream stream, D reuse) { | ||
BinaryDecoder decoder = DecoderFactory.get().directBinaryDecoder(stream, DECODER.get()); | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Most of the changed files in this PR are to avoid using this class now that it is deprecated.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Checked, and didn't find any usage anymore 👍