You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@iceberg.apache.org by "jbonofre (via GitHub)" <gi...@apache.org> on 2024/10/24 08:12:10 UTC
[PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
jbonofre opened a new pull request, #11386:
URL: https://github.com/apache/iceberg/pull/11386
(no comment)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "RussellSpitzer (via GitHub)" <gi...@apache.org>.
RussellSpitzer commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2442236629
@pvary + @jbonofre how important is this for 1.7? If it is important we need to wrap this up ASAP
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2439933515
I'm fixing a new `ClassCastException`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "RussellSpitzer (via GitHub)" <gi...@apache.org>.
RussellSpitzer commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2444803082
Alright! Chatted with folks and we will do the old-path tests in a follow up.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "pvary (via GitHub)" <gi...@apache.org>.
pvary commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2454428555
I think we should backport this change to Flink 1.19, and Flink 1.18 as well.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2435037863
It seems the upsert test is stale. I'm investigating (maybe the schema type mapping).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2439382344
I fixed the issue on `ValueReaders` about strings.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2434604147
@pvary @aokolnychyi @rdblue this PR mimics what has been done in Spark to use Avro planned reads.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2440687833
I fixed all types (primitves and arrays) reads. Tests should be happy now 😄
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "RussellSpitzer (via GitHub)" <gi...@apache.org>.
RussellSpitzer commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2444840468
Thanks @jbonofre for the PR and @Fokko and @pvary for Review. Let's add some additional tests for Spark and Flink in a followup for the old path.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2443246368
@RussellSpitzer the planned Avro reads has been added to Spark (for Iceberg 1.7.x). This one is not a blocker for 1.7.0 but a good to have to benefit the same performance boost as Spark.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "pvary (via GitHub)" <gi...@apache.org>.
pvary commented on code in PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#discussion_r1817698482
##########
flink/v1.20/flink/src/test/java/org/apache/iceberg/flink/data/TestFlinkAvroReaderWriter.java:
##########
@@ -91,7 +91,7 @@ private void writeAndValidate(Schema schema, List<Record> expectedRecords, int n
try (CloseableIterable<RowData> reader =
Avro.read(Files.localInput(recordsFile))
.project(schema)
- .createReaderFunc(FlinkAvroReader::new)
+ .createResolvingReader(FlinkPlannedAvroReader::create)
Review Comment:
QQ: Do we have remaining tests for the old reader? I usually try to keep at least a few tests for the deprecated features as well, so they are not broken unintentionally by future changes (and mark them as deprecated, so we don't forget to remove them when the feature is removed).
If there are other tests? Do we test the same functions for the new reader?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on code in PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#discussion_r1818458279
##########
flink/v1.20/flink/src/test/java/org/apache/iceberg/flink/data/TestFlinkAvroReaderWriter.java:
##########
@@ -91,7 +91,7 @@ private void writeAndValidate(Schema schema, List<Record> expectedRecords, int n
try (CloseableIterable<RowData> reader =
Avro.read(Files.localInput(recordsFile))
.project(schema)
- .createReaderFunc(FlinkAvroReader::new)
+ .createResolvingReader(FlinkPlannedAvroReader::create)
Review Comment:
That's a good point. I can add tests specific to the "old" `FlinkAvroReader`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "RussellSpitzer (via GitHub)" <gi...@apache.org>.
RussellSpitzer merged PR #11386:
URL: https://github.com/apache/iceberg/pull/11386
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on code in PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#discussion_r1816086840
##########
flink/v1.20/flink/src/main/java/org/apache/iceberg/flink/data/FlinkPlannedAvroReader.java:
##########
@@ -0,0 +1,191 @@
+/*
+ * 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.flink.data;
+
+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.flink.table.data.RowData;
+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 FlinkPlannedAvroReader implements DatumReader<RowData>, SupportsRowPosition {
+
+ private final Types.StructType expectedType;
+ private final Map<Integer, ?> idToConstant;
+ private ValueReader<RowData> reader;
+
+ public static FlinkPlannedAvroReader create(org.apache.iceberg.Schema schema) {
+ return create(schema, ImmutableMap.of());
+ }
+
+ public static FlinkPlannedAvroReader create(
+ org.apache.iceberg.Schema schema, Map<Integer, ?> constants) {
+ return new FlinkPlannedAvroReader(schema, constants);
+ }
+
+ private FlinkPlannedAvroReader(
+ org.apache.iceberg.Schema expectedSchema, Map<Integer, ?> constants) {
+ this.expectedType = expectedSchema.asStruct();
+ this.idToConstant = constants;
+ }
+
+ @Override
+ @SuppressWarnings("unchecked")
+ public void setSchema(Schema fileSchema) {
+ this.reader =
+ (ValueReader<RowData>)
+ AvroWithPartnerVisitor.visit(
+ expectedType,
+ fileSchema,
+ new ReadBuilder(idToConstant),
+ AvroWithPartnerVisitor.FieldIDAccessors.get());
+ }
+
+ @Override
+ public RowData read(RowData 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);
+
+ // TODO: should this pass expected so that struct.get can reuse containers?
+ return FlinkValueReaders.struct(readPlan, expected.fields().size());
+ }
+
+ @Override
+ public ValueReader<?> union(Type partner, Schema union, List<ValueReader<?>> options) {
+ return ValueReaders.union(options);
+ }
+
+ @Override
+ public ValueReader<?> array(Type partner, Schema array, ValueReader<?> elementReader) {
+ return ValueReaders.array(elementReader);
+ }
+
+ @Override
+ public ValueReader<?> arrayMap(
+ Type partner, Schema map, ValueReader<?> keyReader, ValueReader<?> valueReader) {
+ return FlinkValueReaders.arrayMap(keyReader, valueReader);
+ }
+
+ @Override
+ public ValueReader<?> map(Type partner, Schema map, ValueReader<?> valueReader) {
+ return FlinkValueReaders.map(FlinkValueReaders.strings(), valueReader);
+ }
+
+ @Override
+ public ValueReader<?> primitive(Type partner, Schema primitive) {
+ LogicalType logicalType = primitive.getLogicalType();
+ if (logicalType != null) {
+ switch (logicalType.getName()) {
+ case "date":
+ // Flink uses the same representation
+ return ValueReaders.ints();
+
+ case "timestamp-millis":
+ // adjust to microseconds
+ ValueReader<Long> longs = ValueReaders.longs();
+ return (ValueReader<Long>) (decoder, ignore) -> longs.read(decoder, null) * 1000L;
+
+ case "timestamp-micros":
+ // Flink uses the same representation
+ return ValueReaders.longs();
+
+ case "decimal":
+ return FlinkValueReaders.decimal(
+ ValueReaders.decimalBytesReader(primitive),
+ ((LogicalTypes.Decimal) logicalType).getPrecision(),
+ ((LogicalTypes.Decimal) logicalType).getScale());
+
+ case "uuid":
+ return FlinkValueReaders.uuids();
+
+ 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:
+ return ValueReaders.strings();
Review Comment:
Yes that's my bad. I'm fixing it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "pvary (via GitHub)" <gi...@apache.org>.
pvary commented on code in PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#discussion_r1816085856
##########
flink/v1.20/flink/src/main/java/org/apache/iceberg/flink/data/FlinkPlannedAvroReader.java:
##########
@@ -0,0 +1,191 @@
+/*
+ * 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.flink.data;
+
+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.flink.table.data.RowData;
+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 FlinkPlannedAvroReader implements DatumReader<RowData>, SupportsRowPosition {
+
+ private final Types.StructType expectedType;
+ private final Map<Integer, ?> idToConstant;
+ private ValueReader<RowData> reader;
+
+ public static FlinkPlannedAvroReader create(org.apache.iceberg.Schema schema) {
+ return create(schema, ImmutableMap.of());
+ }
+
+ public static FlinkPlannedAvroReader create(
+ org.apache.iceberg.Schema schema, Map<Integer, ?> constants) {
+ return new FlinkPlannedAvroReader(schema, constants);
+ }
+
+ private FlinkPlannedAvroReader(
+ org.apache.iceberg.Schema expectedSchema, Map<Integer, ?> constants) {
+ this.expectedType = expectedSchema.asStruct();
+ this.idToConstant = constants;
+ }
+
+ @Override
+ @SuppressWarnings("unchecked")
+ public void setSchema(Schema fileSchema) {
+ this.reader =
+ (ValueReader<RowData>)
+ AvroWithPartnerVisitor.visit(
+ expectedType,
+ fileSchema,
+ new ReadBuilder(idToConstant),
+ AvroWithPartnerVisitor.FieldIDAccessors.get());
+ }
+
+ @Override
+ public RowData read(RowData 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);
+
+ // TODO: should this pass expected so that struct.get can reuse containers?
+ return FlinkValueReaders.struct(readPlan, expected.fields().size());
+ }
+
+ @Override
+ public ValueReader<?> union(Type partner, Schema union, List<ValueReader<?>> options) {
+ return ValueReaders.union(options);
+ }
+
+ @Override
+ public ValueReader<?> array(Type partner, Schema array, ValueReader<?> elementReader) {
+ return ValueReaders.array(elementReader);
+ }
+
+ @Override
+ public ValueReader<?> arrayMap(
+ Type partner, Schema map, ValueReader<?> keyReader, ValueReader<?> valueReader) {
+ return FlinkValueReaders.arrayMap(keyReader, valueReader);
+ }
+
+ @Override
+ public ValueReader<?> map(Type partner, Schema map, ValueReader<?> valueReader) {
+ return FlinkValueReaders.map(FlinkValueReaders.strings(), valueReader);
+ }
+
+ @Override
+ public ValueReader<?> primitive(Type partner, Schema primitive) {
+ LogicalType logicalType = primitive.getLogicalType();
+ if (logicalType != null) {
+ switch (logicalType.getName()) {
+ case "date":
+ // Flink uses the same representation
+ return ValueReaders.ints();
+
+ case "timestamp-millis":
+ // adjust to microseconds
+ ValueReader<Long> longs = ValueReaders.longs();
+ return (ValueReader<Long>) (decoder, ignore) -> longs.read(decoder, null) * 1000L;
+
+ case "timestamp-micros":
+ // Flink uses the same representation
+ return ValueReaders.longs();
+
+ case "decimal":
+ return FlinkValueReaders.decimal(
+ ValueReaders.decimalBytesReader(primitive),
+ ((LogicalTypes.Decimal) logicalType).getPrecision(),
+ ((LogicalTypes.Decimal) logicalType).getScale());
+
+ case "uuid":
+ return FlinkValueReaders.uuids();
+
+ 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:
+ return ValueReaders.strings();
Review Comment:
Maybe this needs to be `FlinkValueReaders.strings()`?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org
Re: [PR] Flink 1.20: Update Flink to use planned Avro reads [iceberg]
Posted by "jbonofre (via GitHub)" <gi...@apache.org>.
jbonofre commented on PR #11386:
URL: https://github.com/apache/iceberg/pull/11386#issuecomment-2435789965
The problem seems to be related to:
```
java.lang.ClassCastException: class java.lang.String cannot be cast to class org.apache.flink.table.data.StringData (java.lang.String is in module java.base of loader 'bootstrap'; org.apache.flink.table.data.StringData is in unnamed module of loader 'app')
at org.apache.flink.table.data.GenericRowData.getString(GenericRowData.java:169)
at org.apache.flink.table.data.RowData.lambda$createFieldGetter$245ca7d1$1(RowData.java:221)
at org.apache.iceberg.flink.data.RowDataProjection.getValue(RowDataProjection.java:172)
at org.apache.iceberg.flink.data.RowDataProjection.isNullAt(RowDataProjection.java:193)
at org.apache.iceberg.flink.data.RowDataUtil.clone(RowDataUtil.java:96)
at org.apache.iceberg.flink.source.reader.RowDataRecordFactory.clone(RowDataRecordFactory.java:71)
at org.apache.iceberg.flink.source.reader.RowDataRecordFactory.clone(RowDataRecordFactory.java:28)
at org.apache.iceberg.flink.source.reader.ArrayPoolDataIteratorBatcher$ArrayPoolBatchIterator.next(ArrayPoolDataIteratorBatcher.java:98)
at org.apache.iceberg.flink.source.reader.ArrayPoolDataIteratorBatcher$ArrayPoolBatchIterator.next(ArrayPoolDataIteratorBatcher.java:67)
at org.apache.iceberg.flink.source.reader.IcebergSourceSplitReader.fetch(IcebergSourceSplitReader.java:96)
at org.apache.flink.connector.base.source.reader.fetcher.FetchTask.run(FetchTask.java:58)
at org.apache.flink.connector.base.source.reader.fetcher.SplitFetcher.runOnce(SplitFetcher.java:165)
... 6 more
```
So it seems related to string read from the `GenericRowData`. I'm checking.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org