[
https://issues.apache.org/jira/browse/AVRO-1891?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15460903#comment-15460903
]
Yibing Shi commented on AVRO-1891:
----------------------------------
I have another look into the code, and find that we might have to extract
{{ConversionResolver}}, because:
# {{Generic/SpecificData}} are public classes, and they are not final.
Applications may have extended from them, which makes it hard to create them
(especially their children) on the fly, which is necessary if we want to add
conversions when writing/reading children of {{SpecificRecordBase}}. On the
other hand, {{ConversionResolver}} is a new class, we can define it as final,
which make it easy to create objects.
# creating {{Generic/SpecificData}} object in specific datum reader/writer is
confusing, especially when the reader/writer already has a {{data}} member.
Using {{ConversionResolver}} makes code much clearer.
# Actually we don't have to mark those add/get conversion methods in
{{GenericData}} as deprecated. They are still useful and provide a convenient
way for applications to manage conversions.
What are you opinions on this, [~cutting] and [~tomwhite]?
> Generated Java code fails with union containing logical type
> ------------------------------------------------------------
>
> Key: AVRO-1891
> URL: https://issues.apache.org/jira/browse/AVRO-1891
> Project: Avro
> Issue Type: Bug
> Components: java, logical types
> Affects Versions: 1.8.1
> Reporter: Ross Black
> Priority: Blocker
> Fix For: 1.8.3
>
> Attachments: AVRO-1891.patch, AVRO-1891.yshi.1.patch,
> AVRO-1891.yshi.2.patch, AVRO-1891.yshi.3.patch
>
>
> Example schema:
> {code}
> {
> "type": "record",
> "name": "RecordV1",
> "namespace": "org.brasslock.event",
> "fields": [
> { "name": "first", "type": ["null", {"type": "long",
> "logicalType":"timestamp-millis"}]}
> ]
> }
> {code}
> The avro compiler generates a field using the relevant joda class:
> {code}
> public org.joda.time.DateTime first
> {code}
> Running the following code to perform encoding:
> {code}
> final RecordV1 record = new
> RecordV1(DateTime.parse("2016-07-29T10:15:30.00Z"));
> final DatumWriter<RecordV1> datumWriter = new
> SpecificDatumWriter<>(record.getSchema());
> final ByteArrayOutputStream stream = new ByteArrayOutputStream(8192);
> final BinaryEncoder encoder =
> EncoderFactory.get().directBinaryEncoder(stream, null);
> datumWriter.write(record, encoder);
> encoder.flush();
> final byte[] bytes = stream.toByteArray();
> {code}
> fails with the exception stacktrace:
> {code}
> org.apache.avro.AvroRuntimeException: Unknown datum type
> org.joda.time.DateTime: 2016-07-29T10:15:30.000Z
> at org.apache.avro.generic.GenericData.getSchemaName(GenericData.java:741)
> at
> org.apache.avro.specific.SpecificData.getSchemaName(SpecificData.java:293)
> at org.apache.avro.generic.GenericData.resolveUnion(GenericData.java:706)
> at
> org.apache.avro.generic.GenericDatumWriter.resolveUnion(GenericDatumWriter.java:192)
> at
> org.apache.avro.generic.GenericDatumWriter.writeWithoutConversion(GenericDatumWriter.java:110)
> at
> org.apache.avro.specific.SpecificDatumWriter.writeField(SpecificDatumWriter.java:87)
> at
> org.apache.avro.generic.GenericDatumWriter.writeRecord(GenericDatumWriter.java:143)
> at
> org.apache.avro.generic.GenericDatumWriter.writeWithoutConversion(GenericDatumWriter.java:105)
> at
> org.apache.avro.generic.GenericDatumWriter.write(GenericDatumWriter.java:73)
> at
> org.apache.avro.generic.GenericDatumWriter.write(GenericDatumWriter.java:60)
> at
> org.brasslock.avro.compiler.GeneratedRecordTest.shouldEncodeLogicalTypeInUnion(GeneratedRecordTest.java:82)
> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
> at
> sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:498)
> at
> org.junit.runners.model.FrameworkMethod$1.runReflectiveCall(FrameworkMethod.java:50)
> at
> org.junit.internal.runners.model.ReflectiveCallable.run(ReflectiveCallable.java:12)
> at
> org.junit.runners.model.FrameworkMethod.invokeExplosively(FrameworkMethod.java:47)
> at
> org.junit.internal.runners.statements.InvokeMethod.evaluate(InvokeMethod.java:17)
> at org.junit.runners.ParentRunner.runLeaf(ParentRunner.java:325)
> at
> org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:78)
> at
> org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:57)
> at org.junit.runners.ParentRunner$3.run(ParentRunner.java:290)
> at org.junit.runners.ParentRunner$1.schedule(ParentRunner.java:71)
> at org.junit.runners.ParentRunner.runChildren(ParentRunner.java:288)
> at org.junit.runners.ParentRunner.access$000(ParentRunner.java:58)
> at org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:268)
> at org.junit.runners.ParentRunner.run(ParentRunner.java:363)
> at org.junit.runner.JUnitCore.run(JUnitCore.java:137)
> at
> com.intellij.junit4.JUnit4IdeaTestRunner.startRunnerWithArgs(JUnit4IdeaTestRunner.java:117)
> at
> com.intellij.junit4.JUnit4IdeaTestRunner.startRunnerWithArgs(JUnit4IdeaTestRunner.java:42)
> at
> com.intellij.rt.execution.junit.JUnitStarter.prepareStreamsAndStart(JUnitStarter.java:253)
> at com.intellij.rt.execution.junit.JUnitStarter.main(JUnitStarter.java:84)
> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
> at
> sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:498)
> at com.intellij.rt.execution.application.AppMain.main(AppMain.java:147)
> {code}
> The failure can be fixed by explicitly adding the relevant conversion(s) to
> DatumWriter / SpecificData:
> {code}
> final RecordV1 record = new
> RecordV1(DateTime.parse("2007-12-03T10:15:30.00Z"));
> final SpecificData specificData = new SpecificData();
> specificData.addLogicalTypeConversion(new
> TimeConversions.TimestampConversion());
> final DatumWriter<RecordV1> datumWriter = new
> SpecificDatumWriter<>(record.getSchema(), specificData);
> final ByteArrayOutputStream stream = new
> ByteArrayOutputStream(AvroUtil.DEFAULT_BUFFER_SIZE);
> final BinaryEncoder encoder =
> EncoderFactory.get().directBinaryEncoder(stream, null);
> datumWriter.write(record, encoder);
> encoder.flush();
> final byte[] bytes = stream.toByteArray();
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
