This library implements several JSON specs, like JSON Schema, JSON Reference and JSON Pointer:
- It works on python 2.7, python 3.3 and above
- It is release under the BSD license
This library has only weak dependencies. You can simply use pip:
$ pip install json-spec
Regading you needs, you can install more features. For example, this command will enable colorated messages:
$ pip install json-spec[cli]
This one will enable ip format for json schema:
$ pip install json-spec[ip]
...
This module expose 2 cli commands.
json-extract will extract parts of your json document:
$ json-extract '#/foo/1' --document-json='{"foo": ["bar", "baz"]}'
$ echo '{"foo": ["bar", "baz"]}' | json-extract '#/foo/1'
$ json-extract '#/foo/1' --document-file=doc.json
$ json-extract '#/foo/1' < doc.json
json-validate will validate your document against a schema:
$ json-validate --schema-file=schema.json --document-json='{"foo": ["bar", "baz"]}'
$ echo '{"foo": ["bar", "baz"]}' | json-validate --schema-file=schema.json
$ json-validate --schema-file=schema.json --document-file=doc.json
$ json-validate --schema-file=schema.json < doc.json
Let say you want to fetch / validate JSON like objects in you python scripts.
You can extract member of an object with JSON Pointer:
from jsonspec.pointer import extract
obj = {
'foo': ['bar', 'baz', 'quux']
}
assert 'baz' == extract(obj, '/foo/1')
You can resolve member of any object with JSON Reference:
from jsonspec.reference import resolve
obj = {
'foo': ['bar', 'baz', {
'$ref': '#/sub'
}],
'sub': 'quux'
}
assert 'quux' == resolve(obj, '#/foo/2')
You can describe you data with JSON Schema:
from jsonspec.validators import load
# data will validate against this schema
validator = load({
'title': 'Example Schema',
'type': 'object',
'properties': {
'age': {
'description': 'Age in years',
'minimum': 0,
'type': 'integer'
},
'firstName': {
'type': 'string'
},
'lastName': {
'type': 'string'
}
},
'required': [
'firstName',
'lastName'
]
})
# validate this data
validator.validate({
'firstName': 'John',
'lastName': 'Noone',
'age': 33,
})
Other examples can be found in the documentation or in the tests.