Added deserializeWithObjectId to SuperSonicBeanDeserializer.

[Janekdererste]

This enables the deserializing cyclic object graphs which were created with @JsonIdentityInfo

[Janekdererste]

This fixes #54

[Janekdererste]

Please have a look at the unit tests. I didn't know how to force the usage of SuperSonicBeanDeserializer so I just created a Bean with many properties. Maybe there is a more elegant way to do this?

[Janekdererste]

And I think some of your unit tests are broken. They also failed on my machine right after I checked out the project.

[headw01]

Overriding the deserializeWithObjectId of BeanDeserializerBase still leaves the SuperSonicBeanDeserializer.deserializeFromObject exposed with the bug.

I think that you should just put this code:

        /* See BeanDeserializer.deserializeFromObject.  [databind#622] 
         * ...Allow Object Id References to come in as JSON Objects as well...
         */
        if ((_objectIdReader != null) && _objectIdReader.maySerializeAsObject()) {
            if (p.hasTokenId(JsonTokenId.ID_FIELD_NAME)
                    && _objectIdReader.isValidReferencePropertyName(p.getCurrentName(), p)) {
                return deserializeFromObjectId(p, ctxt);
            }
        }

to the top of the SuperSonicBeanDeserializer.deserializeFromObject method itself.

[Janekdererste]

Sorry for the late response, since I was on vacation. I moved the test for object Ids as recommended.

[cowtowncoder]

@Janekdererste Thank you for contributing this! I will see if I can merge that in 2.9 (assuming this is problem with 2.x too, not just 3.0), and will also want to verify I understand the issue.

[Janekdererste]

If you have any questions while trying to understand the issue, feel free to ask. I've thought a little bit about this now :-)

[cowtowncoder]

@Janekdererste Trying to reproduce the issue with 2.9, but for some reason tests actually pass for me.
So it may be that problem only occurs for master, which is for 3.0.0. That would also explain test failures as 3.x codebase has changed a lot and sometimes changes in databind are not fully reflected for extension modules (that is, downstream changes sometimes lag).

However... #54 suggests 2.9.x does have issues. So the first thing is to make sure there is reproduction. I'll see if I can port databind test 622 over and go from there.

[cowtowncoder]

Was able to reproduce with a test from databind; manually merged fix as suggested (thank you again for your work). Will be include din 2.9.7 release.