JDK Serializing and using Deserialized ObjectMapper loses linkage back from JsonParser.getCodec()

[243826]

I am attaching a test project which is linked against 2.9.5 version. The test errors out. But if I change the pom to link against 2.2.0, it passes. It seems that this is broken since 2.2.1.

java.lang.IllegalStateException: No ObjectCodec defined for parser, needed for deserialization
	at com.fasterxml.jackson.core.JsonParser._codec(JsonParser.java:1788)
	at com.fasterxml.jackson.core.JsonParser.readValueAsTree(JsonParser.java:1782)
	at com.celeral.serializedobjectmapper.NewClassTest$JSONWrapperDeserializer.deserialize(NewClassTest.java:87)
	at com.celeral.serializedobjectmapper.NewClassTest$JSONWrapperDeserializer.deserialize(NewClassTest.java:72)
	at com.fasterxml.jackson.databind.ObjectMapper._readMapAndClose(ObjectMapper.java:4001)
	at com.fasterxml.jackson.databind.ObjectMapper.readValue(ObjectMapper.java:2992)
	at com.celeral.serializedobjectmapper.NewClassTest.testSerializationOfCodec(NewClassTest.java:164)
	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$2.evaluate(ParentRunner.java:268)
	at org.junit.runners.ParentRunner.run(ParentRunner.java:363)
	at org.apache.maven.surefire.junit4.JUnit4Provider.execute(JUnit4Provider.java:252)
	at org.apache.maven.surefire.junit4.JUnit4Provider.executeTestSet(JUnit4Provider.java:141)
	at org.apache.maven.surefire.junit4.JUnit4Provider.invoke(JUnit4Provider.java:112)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.lang.reflect.Method.invoke(Method.java:498)
	at org.apache.maven.surefire.util.ReflectionUtils.invokeMethodWithArray(ReflectionUtils.java:189)
	at org.apache.maven.surefire.booter.ProviderFactory$ProviderProxy.invoke(ProviderFactory.java:165)
	at org.apache.maven.surefire.booter.ProviderFactory.invokeProvider(ProviderFactory.java:85)
	at org.apache.maven.surefire.booter.ForkedBooter.runSuitesInProcess(ForkedBooter.java:115)
	at org.apache.maven.surefire.booter.ForkedBooter.main(ForkedBooter.java:75)


Results :

Tests in error:
  testSerializationOfCodec(com.celeral.serializedobjectmapper.NewClassTest): No ObjectCodec defined for parser, needed for deserialization

Tests run: 2, Failures: 0, Errors: 1, Skipped: 0

SerializedObjectMapper.zip

[cowtowncoder]

Could you include relevant class(es) from the zip here? Exception suggests that linkage between JsonParser back to ObjectMapper is not working for some reason.

[243826]

There is only one file... NewClassTest.java

[cowtowncoder]

Ok good luck. I hope someone has time to look into this.

[243826]

Do you know who can help? I found out what change in jackson-core between 2.2.0 and 2.2.1 broke this feature. But then I see recently (since late 2017) there has been a lot of refactoring to the code done by you; I can no longer do the mapping between the old code and the new code.

[cowtowncoder]

That is a very, very old version. If you had added that one class in there, I would have looked at this earlier.

Looking at package now code looks ok, and I guess its JDK serialization that does something odd.

But instead of requiring use of getCodec() from parser, you can try achieving the same only through DeserializationContext, and creating JsonParser out of JsonNode.
That way you do not need mapper to use "treeToValue", but can instead use readValue(JsonParser, Class) in DeserializationContext. So something like

JsonNode n = context.readTree(jp);
JsonParser p2 = n.asParser();
MyType value = context.readValue(p2, MyType.class);
p2.close();

Hope this helps.

[cowtowncoder]

Also: I think what you found is a bug, hope to find time to fix it if possible at some point.

[243826]

I tried to follow the comment with the code above with 2.9.5 version, but I am stuck at the first line itself.

JsonNode n = context.readTree(jp);

I did not find readTree method on DeserializationContext.

[243826]

btw - it seems this fixes the issue on top of jackson-core-2.9.5 :

diff --git a/src/main/java/com/fasterxml/jackson/core/JsonFactory.java b/src/main/java/com/fasterxml/jackson/core/JsonFactory.java
index 3abfbbb3..97152379 100644
--- a/src/main/java/com/fasterxml/jackson/core/JsonFactory.java
+++ b/src/main/java/com/fasterxml/jackson/core/JsonFactory.java
@@ -286,7 +286,7 @@ public class JsonFactory
      */
     protected JsonFactory(JsonFactory src, ObjectCodec codec)
     {
-        _objectCodec = null;
+        _objectCodec = codec;
         _factoryFeatures = src._factoryFeatures;
         _parserFeatures = src._parserFeatures;
         _generatorFeatures = src._generatorFeatures;

[cowtowncoder]

@243826 Whoa. That seems like an ... interesting oversight :)

I'll want to add a regression test for this, since serializability of ObjectMapper is somewhat of a goal (although not often needed, is useful for some distributed systems). Thank you for reporting this!

As to readTree(), I think I must have added that in 3.0 for convenience, then. With 2.4 and above there is just readValue(...), for which you would give JsonNode.class as type and that should work similarly.

[cowtowncoder]

Ok. So. Will be fixed in 2.9.6, to be released within a week or two (before end of May).
It may still make sense to avoid the problem for this particular case. Thank you again for reporting and tracking down the problem. Interesting bug survived this long :)

[243826]

Hi, kindly make a release ASAP. I have been waiting eagerly. :-)

…

On Wednesday, May 16, 2018, Tatu Saloranta ***@***.***> wrote: Ok. So. Will be fixed in 2.9.6, to be released within a week or two (before end of May). It may still make sense to avoid the problem for this particular case. Thank you again for reporting and tracking down the problem. Interesting bug survived this long :) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2038 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACk9w8oqZypiPJDWMQPMnNqDO7ZuPL2Yks5tzF-ggaJpZM4UADK8> .

-- Chetan

[mserdur]

I could reproduce this with version 2.9.7

[cowtowncoder]

@mserdur A reproduction would be needed to show how things do not work.

And if so, please file a new issue: release notes unfortunately do not work well with re-opened issues.