poormanschema 0.0.0.1

poormanschema

Simple and effective schema checker:

>>> from poormanschema import *
>>> check({'first_name': 1}. [{'first_name': str, 'last_name': str}])
ValueError: {first_name} should be of type "str"

It can also normalize data:

>>> check(' 2016-03-23T12:23:12 ', OR(STRIP, ISO8601))
'2016-03-23T12:23:12'

Or convert them on entry:

>>> repr(check(['1.3'], [DECIMAL]))
[Decimal('1.3')]

List of predicates

• check(data, schema, path=’’) - check data agains schema, prefix error messages with path

• OR(*schemas) - first schema to match is returned

• ANY - can be anything

• AND(*schemas) - all schemas must match

• MANDATORY(schema) - only useful on dict values, indicate the corresponding key is mandatory, all other keys are optional

• RE(regexp, repl=None, count=0, flag=0) - value must be a string matching regexp

• ISO8601 - value must be a string like ‘2016-09-07T12:12:34’

• NORMALIZE(schema, convert) - apply schema then apply converter to the result

• STRIP - remove starting and ending spaces

• LOWER - lowercase the string

• UPPER - uppercase the string

• DECIMAL - convert a string into a decimal.Decimal object

Exemple: SCIM 1.0 Core Schema of an User object

schema = {
    'schemas': MANDATORY([basestring]),
    'id': MANDATORY(basestring),
    'externalId': MANDATORY(basestring),
    'userName': MANDATORY(unicode),
    'name': MANDATORY({
        'formatted': MANDATORY(str),
        'familyName': str,
        'givenName': str,
        'middleName': str,
        'honorificPrefix': str,
        'honorificSuffix': str,
    }),
    'urn:ietf:params:scim:schemas:extension:enterprise:2.0:User': MANDATORY({
        'employeeNumber': OR(STRIP, RE(r'^\d+$')),
        'costCenter': OR(STRIP, RE(r'^\d+$')),
    }),
    'meta': {
        'resourceType': 'User',
        'created': ISO8601,
        'lastModified': ISO8601,
        'version': RE(r'^(W\\)?"[^"]"$'),
        'location': str,
    }
}