JSON Schema: Checking More Than Syntax

A syntax validator can accept {"age":"twenty"} even when an API expects an integer. JSON Schema describes additional rules for values. A schema validator compares a document against those rules.

A small contract

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "type": "object",
  "properties": {
    "name": { "type": "string", "minLength": 1 },
    "age": { "type": "integer", "minimum": 0 }
  },
  "required": ["name"],
  "additionalProperties": false
}

properties defines checks for named fields. required makes name mandatory; defining a property alone does not require it. additionalProperties: false rejects names not listed here. The $schema declaration identifies the dialect used by the schema.

Compare a few documents

Document	Outcome under this schema
{"name":"Sam"}	Passes; age is optional
{"name":"Sam","age":20}	Passes
{"age":20}	Fails; name is missing
{"name":"Sam","age":"20"}	Fails; age is a string
{"name":"Sam","team":"A"}	Fails; team is not allowed

These are all syntactically valid JSON. Use syntax validation first when a document cannot even be parsed.

Using this site

Use the JSON Schema validator for Draft 7 contract checks. Paste your data and schema separately; the tool checks the schema and reports a failing instance path when the document does not match. Local references are supported, but external URLs are not fetched. The ordinary JSON validator checks syntax only; its green status does not certify a schema match.

When integrating a schema validator into your application, select a supported dialect and decide how referenced schemas are supplied. A local-only workflow should supply dependencies explicitly instead of automatically retrieving arbitrary URLs from untrusted schemas.

Continue with the official JSON Schema tutorial. Keep contract checks separate from transformations: silently changing a string into a number can hide a producer’s error.