Flask-JsonSchema
This repository is forked from the [original](https://github.com/mattupstate/flask-jsonschema) in order to support JSON Schema References (additionally because the original has not been updated in nearly 5 years). See documentation below.
Differences from the original
- Support for JSON Schema References
- Any changes to this repository will officially only support Python 3. The library is so small it might just end up supporting Python 2 anyway though.
- Support for auto-reloading schemas
Installation
With pip:
pip install git+https://github.com/b-ryan/flask-jsonschema.git@master
If this gains any amount of popularity I can make a PyPi package for it.
Usage
JSON request validation for Flask applications.
Place schemas in the specified JSONSCHEMA_DIR.
import os
from flask import Flask, request
from flask_jsonschema import JsonSchema, ValidationError
app = Flask(__name__)
app.config['JSONSCHEMA_DIR'] = os.path.join(app.root_path, 'schemas')
jsonschema = JsonSchema(app)
@app.errorhandler(ValidationError)
def on_validation_error(e):
return "error"
@app.route('/books', methods=['POST'])
@jsonschema.validate('books', 'create')
def create_book():
# create the book
return 'success'
The schema for the example above should be named books.json and should
reside in the configured folder. It should look like so:
{
"create": {
"type": "object",
"properties": {
"title": {},
"author": {}
},
"required": ["title", "author"]
},
"update": {
"type": "object",
"properties": {
"title": {},
"author": {}
}
}
}
Notice the top level action names. Flask-JsonSchema supports one "path" level so
that you can organize related schemas in one file. If you do not wish to use this
feature you can simply use one schema per file and remove the second parameter
to the @jsonschema.validate call.
JSON Schema References
Sometimes you may want to DRY up your schemas and put commonly-used schemas in
a central location. You can do this by creating a definitions directory
within your schemas directory. Each JSON file in this directory may contain
a single schema that will then be made available to each of your top-level
schemas. For example, you may have a file named definitions/author.json
with contents:
{
"type": "string"
}
And then modify the above example schema file to look like this:
{
"create": {
"type": "object",
"properties": {
"title": {},
"author": {"$ref": "#/definitions/author"}
},
"required": ["title", "author"]
},
"update": {
"type": "object",
"properties": {
"title": {},
"author": {"$ref": "#/definitions/author"}
}
}
}