Generating validation code from JSON schema

Question

Is there an existing python package that can help me generate code from a json schema?

For example if I have a JSON object like

{       
    "attribute": "obj.value",      
    "operator":  "greater than",      
    "value" : 235 
}

I want to be able to take this JSON and apply it as a rule over different objects to see which ones pass the rule.

So ideally I want to have something like

is_valid(obj,schema)

where

is_valid({"value":300},{"attribute":"value","operator":"greater than","value":235})

returns True

A quick google shows pypi.org/project/jsonschema, which does exactly what you want. — Martijn Pieters
– Martijn Pieters, Commented Aug 31, 2018 at 12:02
You can also use github.com/keleshev/schema to the same goal — Evgeny
– Evgeny, Commented Aug 31, 2018 at 12:09
@EPo: that's not a JSON Schema library. It has the same goal, but schemas are defined in Python, not in JSON. — Martijn Pieters
– Martijn Pieters, Commented Aug 31, 2018 at 12:12
However, your schema doesn't appear to be following the JSON schema format for numbers. Don't invent your own schema format, stick to a standard. — Martijn Pieters
– Martijn Pieters, Commented Aug 31, 2018 at 12:13

Martijn Pieters · Accepted Answer · 2018-08-31 12:21:22Z

3

The jsonschema project does exactly that, validate Python datastructures against a valid JSON schema:

from jsonschema import validate

validate(obj, schema)

This returns None when the schema is valid, and raises an exception when it is not. If you must have a boolean, use:

import jsonschema

def is_valid(obj, schema):
    try:
        jsonschema.validate(obj, schema)
    except jsonschema.ValidationError:
        return False
    else:
        return True

You do need to use valid JSON schema constraints. For integer values, limit the range if your value needs to adhere to boundaries, for example.

The dictionary {"value": 300} is a JSON object with a single key, where that single key is an integer with a lower boundary, so define that as a JSON schema:

schema = {
    "type": "object",
    "properties": {
        "value": {
            "type": "integer",
            "minimum": 235,
            "exclusiveMinimum": True
        }
    }
}

This schema validates your sample value:

>>> import jsonschema
>>> def is_valid(obj, schema):
...     try:
...         jsonschema.validate(obj, schema)
...     except jsonschema.ValidationError:
...         return False
...     else:
...         return True
...
>>> schema = {
...     "type": "object",
...     "properties": {
...         "value": {
...             "type": "integer",
...             "minimum": 235,
...             "exclusiveMinimum": True
...         }
...     }
... }
>>> is_valid({'value': 300}, schema)
True
>>> is_valid({'value': 1}, schema)
False

Read Understanding JSON Schema for a great tutorial on how to write such schemas.

edited Aug 31, 2018 at 12:21

answered Aug 31, 2018 at 12:03

Martijn Pieters

1.1m326 gold badges4.2k silver badges3.4k bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

Akshay Over a year ago

from what I understand jsonschema is only looking at types whereas I am looking at something more complicated like {"attribute":"value","operator":"not in","value":['a','b','c']} Correct me if I am wrong

Martijn Pieters Over a year ago

@Akshay: if your schema is a valid JSON schema, then jsonschema.validate() works. The documentation includes examples like using "items" : {"enum" : [1, 2, 3]}, to restrict values, for example.

Martijn Pieters Over a year ago

@Akshay: I added a sample schema for your sample object.

Martijn Pieters Over a year ago

@Akshay: for your specific example of excluding a subset (not in a given list), you can use "not": {"enum": ["a", "b", "c"]}, so JSON schema almost certainly covers your use cases already.

Collectives™ on Stack Overflow

Generating validation code from JSON schema

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related