Schematec

Schematec is a set of tools that makes input data validation easier. The purpose of this code is attempt to bring simplicity to applications logics using separation of data validation and actual data processing.

Quickstart

import schematec as s

schema = s.dictionary(
   id=s.integer & s.required,
   name=s.string,
   tags=s.array(s.string),
)

>>> data = {
...     'id': '1',
...     'name': 'Red Hot Chili Peppers',
...     'tags': ['funk', 'rock'],
...     'rank': '1',
... }
>>> schema(data)
{'id': 1, 'name': u'Red Hot Chili Peppers', 'tags': [u'funk', u'rock']}

Concepts

Schematec module is based on three basic concepts:

• Schema
• Validator
• Converter

Schema

Term "schema" is used to describe complex data struct such as dictionary(hashmap) or array(list). Schemas has two different types of validation (it is not related to array schemas):

• Default - Only values with required validator are required, other values are optional
• Weak - All values are optional

schematec.exc.SchemaError is raised in case provided data is incorrect.

Order of schema validations:

1. Unbound Validators
2. Schemas(inner)
3. Converters
4. Bound Validators

Validator

Term "validator" describes callable objects that perform different types of checks. There are two types of validators in schematec:

• Bound - type related, for example "max length" validator is bound to sized type.
• Unbound - universal, for example "required" validator.

Raises schematec.exc.ValidationError.

Schematec provides following validators:

required

check if value is provided

length

check iterable for max length

regex

check if given value is valid

Converter

Term "converter" is used to describe cast functions. Schematec supports subset of JSON data types.

Basic types:

• integer(int)
• string(str)
• boolean(bool)

Containers:

• array(list)
• dictionary(dict)

Raises schematec.exc.ConvertationError.

Convertation rules

integer

1. Any int or long value
2. Any suitable string/unicode
3. Boolean value

number

1. Any float or int or long value
2. Any suitable string/unicode
3. Boolean value

string

1. Any suitable string/unicode
2. Any int or long value

boolean

1. Boolean value
2. 0 or 1
3. '0' or '1'
4. u'0' or u'1'

dictionary

1. Any mapping value(collections.Mapping)

array

1. Any iterable value(collections.Iterable), but not a mapping

Complex Descriptors

"Schema", "validator" and "converter" are internally referenced as "descriptors". Common task is creation of complex validation rules for a field(or "complex descriptors"). To do this use bitwise "and" operator on descriptors:

>>> import schematec
>>> schematec.integer & schematec.required
<schematec.abc.ComplexDescriptor object at 0x10b05a0d0>

Sugar Schema

Schematec supports additional "magic" way to define your schemas. You can use simple dicts and lists to describe your data. For example:

>>> import schematec as s
>>> schema = {
...     'a': [{
...         'b': s.integer,
...     }]
... }
>>> data = {
...     'a': [{'b': 1}, {'b': '1'}, {}]
... }
>>> s.process(schema, data)
{'a': [{'b': 1}, {'b': 1}, {}]}

Examples

Recursive schema

import schematec as s

schema = s.dictionary(
    id=s.integer & s.required,
    entity=s.dictionary(
        name=s.string & s.required,
        value=s.string,
    )
)

>>> data = {
...     'id': 1,
...     'entity': {
...         'name': 'song',
...         'value': 'californication',
...     }
... }
>>> schema(data)
{'id': 1, 'entity': {'name': u'song', 'value': u'californication'}}

Errors handling

import schematec as s

schema = s.dictionary(
    id=s.integer & s.required,
    entity=s.dictionary(
        name=s.string & s.required,
        value=s.string,
    )
)

>>> data = {
...     'id': 1,
...     'entity': {
...         'value': 'californication',
...     }
... }
>>> schema(data)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "schematec/schema.py", line 44, in __call__
    value = schema(value, weak=weak)
  File "schematec/schema.py", line 32, in __call__
    validator(name, data)
  File "schematec/validators.py", line 12, in __call__
    raise exc.ValidationError(name)
schematec.exc.ValidationError: name