Writing Domain Specific Languages with JSON Schema

In this talk, learn how you can use JSON Schema to write a specification language. You will learn about what Domain Specific Languages are, how the WAML web automation language was developed using JSON Schema, and the far-reaching applications of JSON Schema in other domains such as Headless CMSes.

Transcript

1. View Slide

2. Writing Domain Specific Languages
   with JSON Schema
   Yos Riady
   yos.io
   bit.ly/2JbMTrn
   

   View Slide

3. Weak types in JSON
   What is JSON Schema?
   Introduction Schema Uses DSL Conclusion
   JSON Schema use
   cases
   Summary and further
   learning
   Writing a DSL with
   JSON Schema
   

   View Slide

4. View Slide

5. JSON is weakly typed
   {
   “userName”: “Yos”,
   “favouriteNumber”: 42,
   “interests”: [“programming”, “swimming”]
   }
   

   View Slide

6. View Slide

7. message Person {
   required string user_name = 1
   optional int64 favourite_number = 2
   repeated string interests = 3
   }
   Protocol Buffers are strongly typed
   

   View Slide

8. View Slide

9. How can we validate
   JSON data with types?
   

   View Slide

10. Enter JSON Schema
    {
    “userName”: { “type”: “string” },
    “favouriteNumber”: { “type”: “integer” },
    “interests”: { “type”: “array”, “items”: { “type”: “string”} }
    }
    

    View Slide

11. JSON Schema helps you validate
    JSON data.
    

    View Slide

12. jsonschemalint.com
    

    View Slide

13. JSON Schema 101
    {
    "type": "string"
    }
    Valid Invalid
    “foo” 123
    false
    

    View Slide

14. JSON Schema Basic Types
    ● string
    ● Numeric types
    ● object
    ● array
    ● boolean
    ● null
    

    View Slide

15. {
    "type": "object",
    "properties": {
    "userName": {
    "type": "string"
    }
    },
    "required": ["userName"],
    "additionalProperties": false,
    }
    JSON Schema 101 Valid Invalid
    {
    “userName”:
    “foo”
    }
    {
    “userName”: 1
    }
    {
    “name”: “foo”
    }
    {}
    

    View Slide

16. JSON Schema 101
    Invalid Validation Message
    {
    “userName”: 1
    }
    Field .userName should be a string
    {
    “name”: “foo”
    }
    Should NOT have additional properties
    {} Should have required property ‘userName’
    

    View Slide

17. Additional JSON Schema Validations
    {
    "type": "number",
    "minimum": 0,
    "maximum": 100,
    "exclusiveMaximum": true
    }
    {
    "type": "string",
    "pattern": "^(\\([0-9]{3}\\))?[0-9]{3}-[0-9]{4}$"
    }
    

    View Slide

18. Combining JSON Schemas
    ● AnyOf
    ● AllOf
    ● OneOf
    ● not
    {
    "anyOf": [
    { "type": "string" },
    { "type": "number" }
    ]
    }
    

    View Slide

19. Structuring Complex Schemas
    {
    "definitions": {
    "person": { ... }
    },
    "type": "object",
    "properties": {
    "user": { "$ref": "#/definitions/person" }
    }
    }
    

    View Slide

20. View Slide

21. jsonschemalint.com
    

    View Slide

22. Use Cases for JSON Schema
    ● Message / data validation
    ○ Object Relational Mappers (ORMs)
    ● Schemaless objects in Web Applications
    ● Content Management in Headless CMSes
    ● Writing Specification Languages
    ○ OpenAPI / Swagger
    ○ AWS States Language
    ○ WAML
    

    View Slide

23. Hapi.js object validation
    

    View Slide

24. Dynamic forms
    

    View Slide

25. Content Modeling in Headless CMSes
    

    View Slide

26. Content Modeling in Headless CMSes
    

    View Slide

27. Writing Specification Languages
    

    View Slide

28. Weak types in JSON
    What is JSON Schema?
    Introduction Schema Uses DSL Conclusion
    JSON Schema use
    cases
    Summary and further
    learning
    Writing a DSL with
    JSON Schema
    

    View Slide

29. Writing a DSL with JSON Schema
    

    View Slide

30. Domain Specific Languages
    ● Computer language specialized to a particular application domain
    ○ In contrast to General Purpose Languages (GPLs)
    

    View Slide

31. OpenAPI / Swagger
    

    View Slide

32. Web Automation Markup Language (waml.io)
    

    View Slide

33. waml.io/editor
    

    View Slide

34. Writing a DSL with JSON Schema
    "waml": {
    "description": "WAML Specification semantic version number.",
    "type": "string",
    "pattern": "^([0-9]{1,}.[0-9]{1,}.[0-9]{1,})$",
    "enum": ["0.1.0"]
    }
    

    View Slide

35. Writing a DSL with JSON Schema
    "steps": {
    "type": "array",
    "minItems": 1,
    "items": {
    "$ref": "#/definitions/Step"
    }
    }
    

    View Slide

36. Writing a DSL with JSON Schema
    "Step": {
    "type": "object",
    "description": "A step represents a single user interaction.",
    "oneOf": [
    {
    "$ref": "#/definitions/VisitStep"
    },
    {
    "$ref": "#/definitions/ClickStep"
    }
    ]
    }
    

    View Slide

37. "ClickStep": {
    "type": "object",
    "properties": {
    "click": {
    "oneOf": [
    {
    "type": "string"
    },
    {
    "type": "object",
    "properties": {
    "selector": { "type": "string" },
    "index": { "type": "integer", "minimum": 0 }
    }
    }
    ]
    },
    },
    }
    

    View Slide

38. github.com/waml-lang/waml
    

    View Slide

39. Weak types in JSON
    What is JSON Schema?
    Introduction Schema Uses DSL Conclusion
    JSON Schema use
    cases
    Summary and further
    learning
    Writing a DSL with
    JSON Schema
    

    View Slide

40. ● JSON is weakly typed
    ● JSON Schema
    ○ What is JSON Schema?
    ○ JSON validation with JSON Schema
    ○ Use cases
    ● Domain Specific Languages
    ○ What are DSLs?
    ○ Writing a DSL with JSON Schema
    In Closing
    

    View Slide

41. View Slide

42. Writing Domain Specific Languages
    with JSON Schema
    Yos Riady
    yos.io
    bit.ly/2JbMTrn
    

    View Slide

43. Q&A
    

    View Slide