An experimental Python DSL for validation
Here is an example of the use of Violet:
from violet import is_type, cast, passes
validator = is_type(
[
{
"name": is_type(str) & passes(len),
"age": cast(int) & passes(lambda x: 18 <= x <= 99),
"sex": cast(str.lower) & passes(lambda s: s in ["male", "female"])
}
])
data = [{'name': 'Sue', 'age': '28', 'sex': 'FEMALE'},
{'name': 'Sam', 'age': '42', 'sex': 'Male'},
{'name': 'Sacha', 'age': '20', 'sex': 'Male'}]
assert [{'name': 'Sue', 'age': 28, 'sex': 'female'},
{'name': 'Sam', 'age': 42, 'sex': 'male'},
{'name': 'Sacha', 'age': 20, 'sex': 'male'}] == validator(data)
The idea behind violet is to create a composition of directives using the & and | operators as well as container types.
There are currently three directives, is_type, cast and passes.
This directive raises a TypeError if the type check fails.
Possible arguments for is_type are type classes such as int, str,
float, etc:
validator = is_type(int)
assert 10 == validator(10)
validator = is_type(str)
assert "foo" == validator("foo")
Constant values:
validator = is_type(10)
assert 10 == validator(10)
Other directives and composites of those directives:
validator = is_type(is_type(int) | is_type(str))
assert 10 == validator(10)
assert "foo" == validator("foo")
validator = is_type(cast(int) & passes(lambda x: x > 10))
assert 11 == validator("11")
Containers of types or directives:
# list
validator = is_type([int])
assert [10, 11, 12] == validator([10, 11, 12])
# set
validator = is_type({int})
assert {10, 11, 12} == validator({10, 11, 12})
# dict
validator = is_type({"name": str, "age": int})
assert {"name": "Eric", "age": 32} == validator({"name": "Eric", "age": 32})
validator = is_type({"name": str, "age": is_type(cast(int))})
assert {"name": "Eric", "age": 32} == validator({"name": "Eric", "age": "32"})
# nested containers
validator = is_type({"name": str, "friends": {str}})
assert {"name": "Eric", "friends": {"Glenn", "Mark"}}
== validator({"name": "Eric", "friends": {"Glenn", "Mark"}})
Simply converts the value as another:
validator = cast(int)
assert 10 == validator("10")
Raises an AssertionError if the predicate returns False,
passes the value through if the predicate returns True:
validator = passes(lambda x: x > 10)
assert 11 == validator(11)
Currently there is no way to support optional keys in dictionary
types. If a key is not found in the value being validated, a
KeyError is currently raised. Support for optional keys is a
planned feature.