JSON-Schema based type resolver for golang.

Type here is any standard json-schema. However standard specification do not consider any other cases except data validation via schema. This library provides realization of subtyping and type-checking concept for schemas.

Subtype B of type A here is the new schema describing document b which can be successfully valided with both A and B. Though document a described by A is not committed to be valid for schema B. This is called structural type system.

Type A

{ "type": "string" }

Type B

{ "type": "string", "maxLength": 10 }

Type C

{ "type": "integer" }

Type B is the subtype of A: A :> B. But C is neither subtype not equal type of A.

Type A

{
  "title": "Person",
  "type": "object",
  "properties": {
    "name": {
      "type": "string"
    }
  },
  "required": [
    "name"
  ]
}

Type B

{
  "title": "Student",
  "type": "object",
  "properties": {
    "name": {
      "type": "string"
    },
    "course": {
      "type": "integer"
    }
  },
  "required": [
    "name",
    "course"
  ]
}

Type C

{
  "title": "Person_not_strict",
  "type": "object",
  "properties": {
    "name": {
      "type": "string"
    }
  }
}

Type B is the subtype of A: A :> B. But C is neither subtype not equal type of A.

Install

go get -u github.com/tariel-x/polyschema

Use

package main

import (
	"encoding/json"
	"fmt"

	"github.com/tariel-x/polyschema"
)

func main() {
	typeA := polyschema.JsonSchema{}
	typeB := polyschema.JsonSchema{}
	typeC := polyschema.JsonSchema{}

	json.Unmarshal([]byte(`{ "type": "string" }`), &typeA)
	json.Unmarshal([]byte(`{ "type": "string", "maxLength": 10 }`), &typeB)
	json.Unmarshal([]byte(`{ "type": "integer" }`), &typeC)

	fmt.Println("A :> B ", polyschema.Subtype(typeA, typeB)) // 1, that means subtype
	fmt.Println("A :> C ", polyschema.Subtype(typeA, typeC)) // -1, that means not subtype
	fmt.Println("A = A ", polyschema.Subtype(typeA, typeA)) // 0, that means identity
}