how to validate json schema of the given json structure

Question

Hi im having json structure as follows:

{
"datasetType": "monolingual-corpus",
"languages": {
    "sourceLanguage": "hi"
},
"collectionSource": [
    "http://pib.gov.in/"
],
"domain":[
    "news"
],
"license": "cc-by-4.0",
"submitter": {
    "name": "Project aroad",
    "aboutMe": "Open source project run by aroad foundation",
    "team": [
        {
            "name": "Navneet Kumar hegde",
            "aboutMe": "NLP team lead at Project aroad"
        },
        {
            "name": "Aswini Pradeep",
            "aboutMe": "Backend team lead at Project aroad"
        }
    ]
}

I can validate datasetType only using json schema. How can i validate other values such as "languages", "collectoinsource", "submitter". In "Submitter" all fields should be mandatory and how to validate "Team" inside "submitter"

Code i have written in python is only validating "datasetType", not able to validate remaining fields. please help me regarding this, Thanks in advance

Answer 1

try marshmallow . It's great for validating schemas.

from marshmallow import Schema, fields


class LanguageSchema(Schema):
    sourceLanguage = fields.String(required=True)


class UserSchema(Schema):
    name = fields.String(required=True)
    aboutMe = fields.String(required=True)


class SubmitterSchema(UserSchema):
    team = fields.List(fields.Nested(UserSchema()))


class ExampleSchema(Schema):
    datasetType = fields.String(required=True)
    languages = fields.Nested(LanguageSchema(), required=True)
    collectionSource = fields.List(fields.URL, required=True)
    domain = fields.List(fields.String(), required=True)
    license = fields.String(required=True)
    submitter = fields.Nested(SubmitterSchema(), required=True)


data = {
    "datasetType": "monolingual-corpus",
    "languages": {
        "sourceLanguage": "hi"
    },
    "collectionSource": [
        "http://pib.gov.in/"
    ],
    "domain": [
        "news"
    ],
    "license": "cc-by-4.0",
    "submitter": {
        "name": "Project aroad",
        "aboutMe": "Open source project run by aroad foundation",
        "team": [
            {
                "name": "Navneet Kumar hegde",
                "aboutMe": "NLP team lead at Project aroad"
            },
            {
                "name": "Aswini Pradeep",
                "aboutMe": "Backend team lead at Project aroad"
            }
        ]
    }
}

# initialize schema
schema = ExampleSchema()
# validate data, will throw error if data does not fit schema
validated_data = schema.load(data)

Answer 2

You can use a standard called JSON Schema and a python lib jsonschema to validate json data with your own schema.https://python-jsonschema.readthedocs.io/en/stable/validate/

As for writing the schema and validating what you want, you have a couple of ways to go about it.

1. To enforce the data structure you just use basic object and array specs. https://json-schema.org/learn/miscellaneous-examples.html
2. To enforce allowed values (ie in languages) you can use enum property. https://stackoverflow.com/a/30931659/4739483