Validating schema against empty json

[aij0]

Hi,

I was wondering why doesn't the jsonschema.validate() raise an exception if I have empty json.
Am I missing some option or have I misunderstood the what the validation should check?

serverSchema = {
        "type": "object",
        "properties": {
            "server": {"type": "string"},
            "address": {"type": "string", "format": "ip-address"},
            "poll_period": {"type": "number"},
            "regex": {"type": "string"}
        }
    }

If I test this against server = {} it passes without exceptions.
Also if I give it a nested json it also passes without any exceptions:

        nested_json = {
            "servers": [
                {"server": 1234,
                    "address": "hello",
                    "poll_period": "there",
                    "regex": "Weight"
                 },
                {"server": "fourth",
                    "address": "127.0.0.1:5000",
                    "poll_period": 120,
                    "regex": ""
                 }
            ]
        }

Finally if I have the following json which is like the schema but with wrong types it will raise an Exception as expected.

        invalid_json = {
                {"server": 1234,
                    "address": "hello",
                    "poll_period": "there",
                    "regex": "Weight"
                 }

[Julian]

In the JSON Schema specification (all current versions of it), properties are not required by default until you say so, either via the required validator or by saying additionalProperties: false.

Without doing so, what your schema means is simply that those subschemas apply only if the corresponding property is present.

[aij0]

Thank you for quick response!