The problem is that avro has it's own representation of union encoding so
your experience would encode to {"int": 50}.
In a recent project we ended up writing a slightly modded json parser to be
able to use avro schemas on existing json rest calls.
2016-02-29 9:20 GMT+01:00 Chris Miller <[email protected]>:
> Did you ever figure this out? I was having the same problem.
>
>
> --
> Chris Miller
>
> On Fri, Feb 19, 2016 at 2:53 AM, Siva <[email protected]> wrote:
>
>> Can someone help on this? Is anyone faced similar issue?
>>
>> Thanks,
>> Sivakumar Bhavanari.
>>
>> On Wed, Feb 17, 2016 at 4:21 PM, Siva <[email protected]> wrote:
>>
>>> Hi Everyone,
>>>
>>> I m new to avro, running into issues if a type is combined with "null",
>>> like ["null","int"] or ["null", "string"].
>>>
>>> I have a schema like below
>>>
>>> {
>>> "type":"record","namespace":"tutorialspoint",
>>> "name":"empdetails","fields":[ {
>>> "name":"experience",
>>> "type":["null","string"],"default":null
>>> },
>>> { "name":"age", "type":"int" }
>>> ]
>>> }
>>>
>>> Below is the json dataset.
>>>
>>> {"experience" : "da", "age": 50}
>>>
>>> java -jar avro-tools-1.7.7.jar fromjson --schema-file test.avsc
>>> test.json > test.avro
>>>
>>> If I have "null" value in "experience" column it goes though, but if it
>>> has some string it is giving below error. Similar error with int types as
>>> well (VALUE_NUMBER_INT).
>>>
>>> Exception in thread "main" org.apache.avro.AvroTypeException: Expected
>>> start-union. Got VALUE_STRING
>>> at org.apache.avro.io.JsonDecoder.error(JsonDecoder.java:697)
>>> at org.apache.avro.io.JsonDecoder.readIndex(JsonDecoder.java:441)
>>> at
>>> org.apache.avro.io.ResolvingDecoder.doAction(ResolvingDecoder.java:290)
>>> at org.apache.avro.io.parsing.Parser.advance(Parser.java:88)
>>>
>>> I have columns with strings or nulls in json, is there a work around to
>>> this error without changing the json data?
>>>
>>> Thanks in advance.
>>>
>>> Thanks,
>>> Sivakumar Bhavanari.
>>>
>>
>>
>
