Types

Base

paranoid.types.base.TypeFactory(v)

Ensure v is a valid Type.

This function is used to convert user-specified types into internal types for the verification engine. It allows Type subclasses, Type subclass instances, Python type, and user-defined classes to be passed. Returns an instance of the type of v.

Users should never access this function directly.

class paranoid.types.base.Type(*args, **kwargs)

Bases: object

The base Type, from which all variable types should inherit.

What is a Type? While “types” can include standard types built into the programming language (e.g. int, float, string), they can also be more flexible. For example, you can have a “natural numbers” type, or a “range” type, or even a “file path” type.

All types must inherit from this class. They must define the “test” and “generate” functions.

generate()

Generate a list of values of this type.

test(v)

Check whether v is a valid value of this type. Throws an assertion error if v is not a valid value.

class paranoid.types.base.Constant(const)

Bases: paranoid.types.base.Type

Only one valid value, which is passed at initialization

class paranoid.types.base.Unchecked(typ=None)

Bases: paranoid.types.base.Type

Use type typ but do not check it.

class paranoid.types.base.Generic(typ)

Bases: paranoid.types.base.Type

Wraps Python classes to turn them into Types.

Classes may optionally contain the “_test_(v)” or “_generate()” static methods; adding these two functions gives them the same power as traditional Types. “_test(v)” should check if v is a valid member of the class using assert statements only, and “_generate()” should yield a finite number of instances of the class.

class paranoid.types.base.InitGeneric(typ)

Bases: paranoid.types.base.Type

For the self argument passed to __init__. Should not be used directly, use Self instead.

Before a class is initialized (i.e. __init__ is called), it may not be valid according to its _test method. Likewise, passing a fully initialized class value to __init__ through the self parameter may cause problems, as the self parameter for __init__ is the output of the __new__ method. Thus, the __init__ function must be handled separately. This tests only object identity, and generates types based on the __new__ method. This may fail for

class paranoid.types.base.Self(*args, **kwargs)

Bases: paranoid.types.base.Type

Used only as a placeholder for methods with a ‘self’ argument.

class paranoid.types.base.Nothing(*args, **kwargs)

Bases: paranoid.types.base.Type

The None type.

class paranoid.types.base.Function(*args, **kwargs)

Bases: paranoid.types.base.Type

Any function.

class paranoid.types.base.Boolean(*args, **kwargs)

Bases: paranoid.types.base.Type

True or False

class paranoid.types.base.And(*types)

Bases: paranoid.types.base.Type

Conforms to all of the given types.

Any number of Types can be passed to And. The And of these types is the logical AND of them, i.e. a value must conform to each of the types.

class paranoid.types.base.Or(*types)

Bases: paranoid.types.base.Type

Conforms to any of the given types.

Any number of Types can be passed to Or. The Or of these types is the logical OR of them, i.e. a value must conform to at least one the types.

class paranoid.types.base.Not(typ)

Bases: paranoid.types.base.Type

Valid if the given type fails.

Takes one type as an argument. Valid values are not of this type.

Note that this should be avoided when possible, as it cannot generate values. It is most useful within an And clause.

class paranoid.types.base.Maybe(typ)

Bases: paranoid.types.base.Or

Either the given type or None.

One Type may be passed to Maybe. Maybe checks to see whether it is either an element of the passed type or else None. This is useful for optional arguments which take None as the default parameter.

class paranoid.types.base.Void(*args, **kwargs)

Bases: paranoid.types.base.Type

Always fails.

Collections

class paranoid.types.collections.Set(els)

Bases: paranoid.types.base.Type

Any element which is a member of els.

els can be one of several standard Python types, including a list, a tuple, or a set. Any object with the __contains__ function is valid. This ensures that a value is contained within els.

class paranoid.types.collections.List(t)

Bases: paranoid.types.base.Type

A Python list.

class paranoid.types.collections.Tuple(*args)

Bases: paranoid.types.base.Type

A Python tuple.

class paranoid.types.collections.Dict(k, v)

Bases: paranoid.types.base.Type

A Python dictionary.

class paranoid.types.collections.ParametersDict(params, all_mandatory=False)

Bases: paranoid.types.base.Type

A Python dictionary with limited keys.

Represents a set of parameters. params, the single argument, should be a dictionary, where the keys are strings representing the parameter names the values are Types. The only keys allowed in a ParametersDict are the keys in params. The values for each key must be of the type specified. Note that not all of the keys in params must be specified for this type to be valid.

Numeric

class paranoid.types.numeric.Numeric(*args, **kwargs)

Bases: paranoid.types.base.Type

Any integer or float, including inf, -inf, and nan.

class paranoid.types.numeric.ExtendedReal(*args, **kwargs)

Bases: paranoid.types.base.Type

Any integer or float, excluding nan.

class paranoid.types.numeric.Number(*args, **kwargs)

Bases: paranoid.types.base.Type

Any integer or float, excluding inf, -inf, and nan.

class paranoid.types.numeric.Integer(*args, **kwargs)

Bases: paranoid.types.base.Type

Any integer.

class paranoid.types.numeric.Natural0(*args, **kwargs)

Bases: paranoid.types.numeric.Integer

Any natural number including 0.

class paranoid.types.numeric.Natural1(*args, **kwargs)

Bases: paranoid.types.numeric.Integer

Any natural number excluding 0.

class paranoid.types.numeric.Range(low, high)

Bases: paranoid.types.numeric.Number

Any integer or float from low to high, inclusive.

Note that this does NOT include correction for floating point roundoff errors. This is because, if there are floating point roundoff errors, some code may fail.

class paranoid.types.numeric.RangeClosedOpen(low, high)

Bases: paranoid.types.numeric.Range

A half open interval from low (closed) to high (open).

class paranoid.types.numeric.RangeOpenClosed(low, high)

Bases: paranoid.types.numeric.Range

A half open interval from low (open) to high (closed).

class paranoid.types.numeric.RangeOpen(low, high)

Bases: paranoid.types.numeric.Range

Any number in the open interval from low to high.

class paranoid.types.numeric.Positive0(*args, **kwargs)

Bases: paranoid.types.numeric.Number

A positive number, including zero.

class paranoid.types.numeric.Positive(*args, **kwargs)

Bases: paranoid.types.numeric.Number

A positive number, excluding zero.

class paranoid.types.numeric.NDArray(d=None, t=None)

Bases: paranoid.types.base.Type

A numpy ndarray of dimension d and type t.

String

class paranoid.types.string.String(*args, **kwargs)

Bases: paranoid.types.base.Type

Any string.

class paranoid.types.string.Identifier(*args, **kwargs)

Bases: paranoid.types.string.String

Any non-empty alphanumeric string with underscores and hyphens.

class paranoid.types.string.Alphanumeric(*args, **kwargs)

Bases: paranoid.types.string.Identifier

Any non-empty alphanumeric string

class paranoid.types.string.Latin(*args, **kwargs)

Bases: paranoid.types.string.Alphanumeric

Any non-empty string with latin characters only