Avro Schema Parser
An AsyncAPI schema parser for Avro 1.x schemas.
Note
Version >= 2.0.0 of this package is only supported by the @asyncapi/parser version >= 2.0.0.
Installation
npm install @asyncapi/avro-schema-parser
// OR
yarn add @asyncapi/avro-schema-parser
Usage
import { Parser } from '@asyncapi/parser';
import { AvroSchemaParser } from '@asyncapi/avro-schema-parser';
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload: # The following is an Avro schema in YAML format (JSON format is also supported)
type: record
name: User
namespace: com.company
doc: User information
fields:
- name: displayName
type: string
- name: email
type: string
- name: age
type: int
`
const { document } = await parser.parse(asyncapiWithAvro);
const { Parser } = require('@asyncapi/parser');
const { AvroSchemaParser } = require('@asyncapi/raml-dt-schema-parser');
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload: # The following is an Avro schema in YAML format (JSON format is also supported)
type: record
name: User
namespace: com.company
doc: User information
fields:
- name: displayName
type: string
- name: email
type: string
- name: age
type: int
Usage with remote references
import { Parser } from '@asyncapi/parser';
import { AvroSchemaParser } from '@asyncapi/avro-schema-parser';
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
$ref: 'https://schemas.example.com/user'
`
const { document } = await parser.parse(asyncapiWithAvro);
const { Parser } = require('@asyncapi/parser');
const { AvroSchemaParser } = require('@asyncapi/raml-dt-schema-parser');
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
$ref: 'https://schemas.example.com/user'
`
const { document } = await parser.parse(asyncapiWithAvro);
Usage with local references
import { Parser } from '@asyncapi/parser';
import { AvroSchemaParser } from '@asyncapi/avro-schema-parser';
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
$ref: 'local/path/to/file/user'
`
const { document } = await parser.parse(asyncapiWithAvro);
const { Parser } = require('@asyncapi/parser');
const { AvroSchemaParser } = require('@asyncapi/raml-dt-schema-parser');
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
$ref: 'local/path/to/file/user'
`
const { document } = await parser.parse(asyncapiWithAvro);
Usage with Confluent Schema Registry
Create an API key

Copy the key and the secret

Use them on your AsyncAPI document
import { Parser } from '@asyncapi/parser';
import { AvroSchemaParser } from '@asyncapi/avro-schema-parser';
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
$ref: 'https://LY422RBU2HN6JQ5T:+f8wz9a0iM06AX7xfwbzSM9YPw/JIkr22Cvl5EKT5Hb1d/nz5nOpbXV/vZC+Iz5c@example.europe-west3.gcp.confluent.cloud/subjects/test/versions/1/schema'
`
const { document } = await parser.parse(asyncapiWithAvro);
const { Parser } = require('@asyncapi/parser');
const { AvroSchemaParser } = require('@asyncapi/raml-dt-schema-parser');
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
$ref: 'https://LY422RBU2HN6JQ5T:+f8wz9a0iM06AX7xfwbzSM9YPw/JIkr22Cvl5EKT5Hb1d/nz5nOpbXV/vZC+Iz5c@example.europe-west3.gcp.confluent.cloud/subjects/test/versions/1/schema'
`
const { document } = await parser.parse(asyncapiWithAvro);
Features
Validation of Avro schemas
Avro schemas included in parsed AsyncAPI documents are validated using avsc. The parser.validate(...) function returns all validation issues.
import { Parser } from '@asyncapi/parser';
import { AvroSchemaParser } from '@asyncapi/avro-schema-parser';
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithInvalidAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
type: notAValidAvroType
`;
const diagnostics = await parser.validate(doc);
const avroDiagnostics = diagnostics.filter(d => d.code === 'asyncapi-schemas-v2');
console.log(avroDiagnostics);
const { Parser } = require('@asyncapi/parser');
const { AvroSchemaParser } = require('@asyncapi/raml-dt-schema-parser');
const parser = new Parser();
parser.registerSchemaParser(AvroSchemaParser());
const asyncapiWithInvalidAvro = `
asyncapi: 2.0.0
info:
title: Example with Avro
version: 0.1.0
channels:
example:
publish:
message:
schemaFormat: 'application/vnd.apache.avro;version=1.9.0'
payload:
type: notAValidAvroType
`;
const diagnostics = await parser.validate(doc);
const avroDiagnostics = diagnostics.filter(d => d.code === 'asyncapi-schemas-v2');
console.log(avroDiagnostics);
Note
parser.parse(...) function also returns diagnostics data with all the validation issues.
Support of required attributes
Required fields are fields with no default value and without the "null" union element.
Additional attributes not defined in the Avro Specification are permitted and are treated as a metadata by the specification. To improve human readability of generated AsyncAPI documentation and to leverage more features from the JSON schema we included support for the extra attributes that can be added into Avro document.
example - Can be used to define the example value from the business domain of given field. Value will be propagated into examples attribute of JSON schema and therefore will be picked for the generated "Example of payload" when using some AsyncAPI documentation generator.
For Number instances:
For String instances:
For Array instances:
Support for names and namespaces
If, at the top level of the Avro schema, the 'name' attribute is defined, it will be copied to the corresponding JSON schema's 'x-parser-schema-id' attribute. If the Avro schema also has the 'namespace' attribute defined, then that schema's fully qualified name will be put into that attribute. The fully qualified name is defined by the namespace, followed by a dot, followed by the name.
If there are two schemas that resolve to the same fully qualified name, only the last one will be returned by the parser. Make sure names of your schemas are unique.
If no name attribute is present, the 'x-parser-schema-id' will have a generated unique id with a name like 'anonymous-schema-1' generated by the main parser. 'x-parser-schema-id' is one of the custom extensions supported by the parser.
Limitations
Float and double-precision numbers
JSON numbers (RFC 4627, section 2.4) don't define any limit to the scale and/or precision of numbers. That said, we can enforce limits on int and long but we can't enforce them on float and double because they can't accurately be represented on JSON Schema.
Since we support extra attributes on field, you can set minimum and maximum attribute on float and double types to display number limits.
Hardcoded key and secret
This is not a limitation of this package per se but of the JSON Reference RFC. So far, you can only hardcode the values for key and secret on the $ref URL.
Contributors
Thanks goes to these wonderful people (emoji key):
This project follows the all-contributors specification. Contributions of any kind welcome!