JSON has like XSL for XML a schema to validate de content of a JSON-file.
Schema files are name regularly named *.schema.json.
The json-schema.org
site has the detailed information.
Actually validation is needed while:
For each a solution given including installation or configuration instructions.
This example uses references to other schema files for types or objects.
For the example the files are structured accordingly.
<project-root>
├─ json (*.schema.json)
└─ schema (test.*.json)
Simple json-file named test.customer.json of customer to ship products.
Name of the file is important for the IDE (JetBrains) file mappings.
{
"name": {
"first": "Prins",
"middle": "van",
"last": "Oranje"
},
"billing_address": {
"street_address": "Noordeinde 68",
"postal_code": "2514GL",
"city": "'s-Gravenhage",
"state": "Zuid-Holland"
},
"shipping_address": {
"street_address": "Molenstraat 27",
"postal_code": "2513BJ",
"city": "'s-Gravenhage",
"state": "Zuid-Hoilland"
},
"parcel_size": {
"height": 200,
"width": 80,
"depth": 30
}
}
Multiple schema files are used to demonstrate a way to organise and reuse schemas in a structured way.
This file contains specific individual types and are referenced by multiple schema files.
It is also subject to a schema itself and in this case http://json-schema.org/draft-07/schema#.
{
"$id": "defs",
"$schema": "http://json-schema.org/draft-07/schema#",
"definitions": {
"firstName": {
"type": "string",
"minLength": 1,
"maxLength": 20
},
"middleName": {
"type": "string",
"minLength": 0,
"maxLength": 15
},
"lastName": {
"type": "string",
"minLength": 1,
"maxLength": 30
},
"cityName": {
"type": "string",
"pattern": "^.{1,40}$"
},
"stateName": {
"type": "string",
"pattern": "^.{1,40}$"
},
"parcelSizeHeight": {
"type": "number",
"maximum": 210,
"minimum": 30
},
"parcelSizeWidth": {
"type": "number",
"maximum": 85,
"minimum": 30
},
"parcelSizeDepth": {
"type": "number",
"maximum": 250,
"minimum": 30
},
"postalCode": {
"type": "string",
"pattern": "^[1-9][0-9]{3}[A-Z]{2}$"
}
}
}
This file describes an object and is referenced in full.
The file also references defs.schema.json for individual types.
{
"$id": "address",
"type": "object",
"additionalProperties": false,
"properties": {
"street_address": {
"type": "string"
},
"postal_code": {
"$ref": "defs.schema.json#/definitions/postalCode"
},
"city": {
"$ref": "defs.schema.json#/definitions/cityName"
},
"state": {
"$ref": "defs.schema.json#/definitions/stateName"
}
},
"required": [
"street_address",
"postal_code",
"city",
"state"
]
}
This file uses a mix of objects like name which is locally defined.
Objects bgilling_address and shipping_address are externally defined.
In order for the validator (python3-jsonschema) to find the referenced schemas
locally a full path file-url is needed.
{
"$id": "file://<full-dir-from-root>/schema/customer.schema.json",
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"additionalProperties": false,
"properties": {
"name": {
"type": "object",
"additionalProperties": false,
"properties": {
"first": {
"$ref": "defs.schema.json#/definitions/firstName"
},
"middle": {
"$ref": "defs.schema.json#/definitions/middleName"
},
"last": {
"$ref": "defs.schema.json#/definitions/lastName"
}
},
"required": [
"first",
"middle",
"last"
]
},
"shipping_address": {
"$ref": "address.schema.json"
},
"billing_address": {
"$ref": "address.schema.json"
},
"parcel_size": {
"type": "object",
"additionalProperties": false,
"properties": {
"height": {
"$ref": "defs.schema.json#/definitions/parcelSizeHeight"
},
"width": {
"$ref": "defs.schema.json#/definitions/parcelSizeWidth"
},
"depth": {
"$ref": "defs.schema.json#/definitions/parcelSizeDepth"
}
}
}
},
"required": [
"name",
"shipping_address",
"billing_address",
"parcel_size"
]
}
Python has an application which can be installed by installing package python3-jsonschema.
For debian with the apt package manager this is.
sudo apt install python3-jsonschema`
When done the command jsonschema is available and jsonschema --help tells that for our purpose
jsonschema -i <project-root>/json/test.customer.json <project-root>/schema/customer.schema.json
In JetBrains products json-schemas can be configured using file mappings allowing wildcards.
Since the schema is also a JSON-file it is also subject to a schema itself and this is configured too.
In the JetBrains product like CLion go to "Settings | Languages & Frameworks | Schemas & DTDs | JSON Schema Mappings".
Add an external schema called "JSON Schema version 7" using URL http://json-schema.org/draft-07/schema and add file pattern *.schema.json.
Add another one called "Customer Data" using file <project-root>/schema/customer.schema.json and file pattern *.customer.json.
After this the file `test.customer.json in the editor is automatically validate.
In the editor itself and in "Problems" tab the validation errors are reported.
At the time of writing a working version is at GitHub pboettch/json-schema-validator
Add the file cmake/nlohmann_jsonConfig.cmake in the project with the following content.
# FetchContent added in CMake 3.11, downloads during the configure step.
include(FetchContent)
# Import Json library.
FetchContent_Declare(
json
GIT_REPOSITORY https://github.com/nlohmann/json
GIT_TAG v3.8.0
)
# Adds nlohmann_json::nlohmann_json
FetchContent_MakeAvailable(json)
Insert the following into the main CMakeLists.txt file to make it find the cmake/nlohmann_jsonConfig.cmake file.
# Make it so our own packages are found.
list(APPEND CMAKE_PREFIX_PATH "${CMAKE_CURRENT_LIST_DIR}/cmake")
As of writing the GitHub nlohmann_json version is at v3.11.2.
It seems the library does not handle file URLs (file:///home/user/schema) but only HTTP of HTTPS protocols.