validate.Validator

class validate.Validator(functions=None)

Validator is an object that allows you to register a set of ‘checks’. These checks take input and test that it conforms to the check.

This can also involve converting the value from a string into the correct datatype.

The check method takes an input string which configures which check is to be used and applies that check to a supplied value.

An example input string would be: ‘int_range(param1, param2)’

You would then provide something like:

>>> def int_range_check(value, min, max):
...     # turn min and max from strings to integers
...     min = int(min)
...     max = int(max)
...     # check that value is of the correct type.
...     # possible valid inputs are integers or strings
...     # that represent integers
...     if not isinstance(value, (int, long, string_type)):
...         raise VdtTypeError(value)
...     elif isinstance(value, string_type):
...         # if we are given a string
...         # attempt to convert to an integer
...         try:
...             value = int(value)
...         except ValueError:
...             raise VdtValueError(value)
...     # check the value is between our constraints
...     if not min <= value:
...          raise VdtValueTooSmallError(value)
...     if not value <= max:
...          raise VdtValueTooBigError(value)
...     return value

>>> fdict = {'int_range': int_range_check}
>>> vtr1 = Validator(fdict)
>>> vtr1.check('int_range(20, 40)', '30')
30
>>> vtr1.check('int_range(20, 40)', '60')
Traceback (most recent call last):
VdtValueTooBigError: the value "60" is too big.

New functions can be added with :

>>> vtr2 = Validator()       
>>> vtr2.functions['int_range'] = int_range_check

Or by passing in a dictionary of functions when Validator is instantiated.

Your functions can use keyword arguments, but the first argument should always be ‘value’.

If the function doesn’t take additional arguments, the parentheses are optional in the check. It can be written with either of :

keyword = function_name
keyword = function_name()

The first program to utilise Validator() was Michael Foord’s ConfigObj, an alternative to ConfigParser which supports lists and can validate a config file using a config schema. For more details on using Validator with ConfigObj see: https://configobj.readthedocs.org/en/latest/configobj.html

>>> vtri = Validator()

__init__(functions=None)

>>> vtri = Validator()

Methods

__init__([functions])	>>> vtri = Validator()
check(check, value[, missing])	Usage: check(check, value)
get_default_value(check)	Given a check, return the default value for the check (converted to the right type).

check(check, value, missing=False)

Usage: check(check, value)

Arguments:

check: string representing check to apply (including arguments) value: object to be checked

Returns value, converted to correct type if necessary

If the check fails, raises a ValidateError subclass.

>>> vtor.check('yoda', '')
Traceback (most recent call last):
VdtUnknownCheckError: the check "yoda" is unknown.
>>> vtor.check('yoda()', '')
Traceback (most recent call last):
VdtUnknownCheckError: the check "yoda" is unknown.

>>> vtor.check('string(default="")', '', missing=True)
''

get_default_value(check)

Given a check, return the default value for the check (converted to the right type).

If the check doesn’t specify a default value then a KeyError will be raised.