i.e. is it possible to make field required similar to ProtoBuf:
message SearchRequest {
required string query = 1;
}
All fields are required in Avro by default. As is mentioned in the official documentation, if you want to make something optional, you have to make it nullable by unioning its type with null, like this
{ "namespace": "example.avro",
"type": "record",
"name": "User",
"fields": [
{"name": "name", "type": "string"},
{"name": "favorite_number", "type": ["int", "null"]},
{"name": "favorite_color", "type": ["string", "null"]}
]
}
In this example, name is required, favorite_number and favorite_color are optional. I recommend spending some more time with the documentation.
Related
I am completely new to Avro serialization and I am trying to get my head around how complex types are defined.
I am puzzled by how Avro generates the Enums in Java.
{
"type":"record",
"namespace":"com.example",
"name": "Customer",
"doc":"Avro Schema for our Customer",
"fields":[
{"name":"first_name","type":"string","doc":"First Name of Customer"},
{"name":"last_name","type":"string","doc":"Last Name of Customer"},
{"name":"automated_email","type":"boolean","doc":"true if the user wants marketing email", "default":true},
{
"name": "customer_type",
"type": ["null",
{
"name": "Customertype",
"type": "enum",
"symbols": ["OLD","NEW"]
}
]
}
]
}
Notice the customer_type field. If I give null, then in my generated sources I get the correct Enum type which is :
private com.example.Customertype customer_type;
But the moment I remove the null value and define customer_type in the following way:
{
"name": "customer_type",
"type": [
{
"name": "Customertype",
"type": "enum",
"symbols": ["OLD","NEW"]
}
]
}
The declaration changes to :
private Object customer_type;
What does that "null" string signify ? Why is it important ?
I have tried looking through the AVRO specification but nothing has given me a clear cut answer why this is working the way it is.
I am using the AVRO Maven plugin.
Any beginner resources for AVRO will also be appreciated.
Thank you.
If you are going to remove the null, you should remove the [ and ] brackets (because it is no longer a union).
So your customer_type schema should look like this:
{
"name": "customer_type",
"type": {
"name": "Customertype",
"type": "enum",
"symbols": ["OLD","NEW"]
}
}
We have some legacy file format, which I would need to migrate to Avro storage. The tricky part is that the records basically have
some common fields,
a discriminator field and
some unique fields, specific to the type selected by the discriminator field
all of them stored in the same file, without any order, fully mixed with each other. (It's legacy...)
In Java/object-oriented programming, one could represent our records concept as the following:
abstract class RecordWithCommonFields {
private Long commonField1;
private String commonField2;
...
}
class RecordTypeA extends RecordWithCommonFields {
private Integer specificToA1;
private String specificToA1;
...
}
class RecordTypeB extends RecordWithCommonFields {
private Boolean specificToB1;
private String specificToB1;
...
}
Imagine the data being something like this:
commonField1Value;commonField2Value,TYPE_IS_A,specificToA1Value,specificToA1Value
commonField1Value;commonField2Value,TYPE_IS_B,specificToB1Value,specificToB1Value
So I would like to process an incoming file and write its content to Avro format, somehow representing the different types of the records.
Can someone give me some ideas on how to achieve this?
Nandor from the Avro users email list was kind enough to help me out with this answer, credits go to him; this answer is for the record just in case someone else hits the same issue.
His solution is simple, basically using composition rather than inheritance, by introducing a common container class and a field referencing a specific subclass.
With this approach the mapping looks like this:
{
"namespace": "com.foobar",
"name": "UnionRecords",
"type": "array",
"items": {
"type": "record",
"name": "RecordWithCommonFields",
"fields": [
{"name": "commonField1", "type": "string"},
{"name": "commonField2", "type": "string"},
{"name": "subtype", "type": [
{
"type" : "record",
"name": "RecordTypeA",
"fields" : [
{"name": "integerSpecificToA1", "type": ["null", "long"] },
{"name": "stringSpecificToA1", "type": ["null", "string"]}
]
},
{
"type" : "record",
"name": "RecordTypeB",
"fields" : [
{"name": "booleanSpecificToB1", "type": ["null", "boolean"]},
{"name": "stringSpecificToB1", "type": ["null", "string"]}
]
}
]}
]
}
}
I tried different things with the following web ui
https://schema-registry-ui.landoop.com
I couldn't seem to put the following into the registry:
{
"namespace": "test.avro",
"type": "record",
"name": "test",
"fields": [
{
"name": "field1",
"type": "string"
},
{
"name": "field2",
"type": "record",
"fields":[
{"name": "field1", "type": "string" },
{"name": "field2", "type": "string"},
{"name": "intField", "type": "int"}
]
}
]
}
Also, is there a way to refer to another schema from inside the current one to create a compound/nested schema?
Have a look at the example at
https://github.com/Landoop/schema-registry-ui/issues/43
You need to define schema as an array - with the 1st element the nested record
and as a 2nd element the main avro record
I have a simple schema as follows:
{
"name": "owner",
"type": "record",
"doc": "todo",
"fields": [
{ "name": "version", "type": "int", "default": 1},
{ "name": "name", "type": "string" },
{ "name": "age", "type": "int" },
]
}
However when I use the avro-maven-plugin to generate a java object from this specificaion, it does not set the default value of version field to 1.
How do I make that happen?
Nevermind, It works fine as is.
I was looking at the generated Java class, and could not figure out where it was setting the default value to 1. But when I serialize it using json, I see the default value in the output. Also, the getter returns the default value as well.
I have two questions:
Is it possible to use the same reader and parse records that were written with two schemas that are compatible, e.g. Schema V2 only has an additional optional field compared to Schema V1 and I want the reader to understand both? I think the answer here is no, but if yes, how do I do that?
I have tried writing a record with Schema V1 and reading it with Schema V2, but I get the following error:
org.apache.avro.AvroTypeException: Found foo, expecting foo
I used avro-1.7.3 and:
writer = new GenericDatumWriter<GenericData.Record>(SchemaV1);
reader = new GenericDatumReader<GenericData.Record>(SchemaV2, SchemaV1);
Here are examples of the two schemas (I have tried adding a namespace as well, but no luck).
Schema V1:
{
"name": "foo",
"type": "record",
"fields": [{
"name": "products",
"type": {
"type": "array",
"items": {
"name": "product",
"type": "record",
"fields": [{
"name": "a1",
"type": "string"
}, {
"name": "a2",
"type": {"type": "fixed", "name": "a3", "size": 1}
}, {
"name": "a4",
"type": "int"
}, {
"name": "a5",
"type": "int"
}]
}
}
}]
}
Schema V2:
{
"name": "foo",
"type": "record",
"fields": [{
"name": "products",
"type": {
"type": "array",
"items": {
"name": "product",
"type": "record",
"fields": [{
"name": "a1",
"type": "string"
}, {
"name": "a2",
"type": {"type": "fixed", "name": "a3", "size": 1}
}, {
"name": "a4",
"type": "int"
}, {
"name": "a5",
"type": "int"
}]
}
}
},
{
"name": "purchases",
"type": ["null",{
"type": "array",
"items": {
"name": "purchase",
"type": "record",
"fields": [{
"name": "a1",
"type": "int"
}, {
"name": "a2",
"type": "int"
}]
}
}]
}]
}
Thanks in advance.
I encountered the same issue. That might be a bug of avro, but you probably can work around by adding "default": null to the field of "purchase".
Check my blog for details: http://ben-tech.blogspot.com/2013/05/avro-schema-evolution.html
You can do opposite of it . Mean you can parse data schem 1 and write data from schema 2 . Beacause at write time it write data into file and if we don't provide any field at reading time than it will be ok. But if we write less field than read than it will not recognize extra field at reading time so , it will give error .
Best way is to have a schema mapping to maintain the schema like Confluent Avro schema registry.
Key Take Aways:
1. Unlike Thrift, avro serialized objects do not hold any schema.
2. As there is no schema stored in the serialized byte array, one has to provide the schema with which it was written.
3. Confluent Schema Registry provides a service to maintain schema versions.
4. Confluent provides Cached Schema Client, which checks in cache first before sending the request over the network.
5. Json Schema present in “avsc” file is different from the schema present in Avro Object.
6. All Avro objects extends from Generic Record
7. During Serialization : based on schema of the Avro Object a schema Id is requested from the Confluent Schema Registry.
8. The schemaId which is a INTEGER is converted to Bytes and prepend to serialized AvroObject.
9. During Deserialization : First 4 bytes are removed from the ByteArray. 4 bytes are converted back to INTEGER(SchemaId)
10. Schema is requested from the Confluent Schema Registry and using this schema the byteArray is deserialized.
http://bytepadding.com/big-data/spark/avro/avro-serialization-de-serialization-using-confluent-schema-registry/