ID Schema

How to install it?

You can install ID Schema from this Github repository with python3 setup.py install, or just install it directly from pypi with pip3 install id-schema.

What is it?

ID Schema allows you to define your custom Schema, how an identifier string should look like. Each identifier string consists of one or more components, which are then concatenated into one large identifier string. Currently there are three components.

Components

RandomGroup

The RandomGroup component allows to generate a random sequence of a predefined alphabet with a predefined length. A length can either be a LengthRange or LengthSequence. A LengthRange has a minimum length and a maximum length. Valid LengthRanges are for example:

LengthRange(1) # always maps to 1
LengthRange(1, max_length=1) # always maps to 1
LengthRange(1, 3) # length range between 1 and 3, interval [1,3]
LengthRange(5,10) # interval [5,10]
LengthRange(10, max_length=1) # INVALID RANGE! , because min > max

A LengthSequence has a sequence of lengths that are allowed. Valid LengthSequences are for example:

LengthSequence(1) # always maps to 1
LengthSequence(1,2,3) # maps to lengths (1,2,3)
LengthSequence(1, 10, 20) # maps to lengths (1,10,20)
LengthSequence((1,10,20)) # also maps to lengths (1,10,20)

If you want the resulting random string sequence to contain only unique elements, you can use the keyword argument unique=True. The default is unique=False.

Here are some examples on how to use the RandomGroup component.

RandomGroup(alphabet="abcd", length=1)
RandomGroup(alphabet="abcd", length=LengthSequence(1,3))
RandomGroup(alphabet="abcd", length=LengthRange(1,5))
RandomGroup(alphabet="abcd", length=LengthSequence(1,3,5), unique=True)
RandomGroup(alphabet="ab", length=LengthRange(1,3), weights=[0.75,0.25]) # 'a' will occur 3/4 as much as 'b'

ExactLiteral

The ExactLiteral component allows to generate an exact literal sequence. The usage is very simple.

ExactLiteral("-") # will always generate '-'
ExactLiteral("abcd") # will always generate 'abcd'

OneOf

The OneOf component allows to generate exactly one other component based on a component subset. Specifying a pure string automatically transforms it to an ExactLiteral for you.

OneOf(components=(
	"gmail.com",
	"yahoo.com",
	"hotmail.com"
))
OneOf(components=(
	RandomGroup(alphabet="abcd", length=LengthRange(1,3)),
	RandomGroup(alphabet="efgh", length=LengthRange(1,3)),
), weights=(0.75, 0.25)) # first RandomGroup will occur 3/4 as much as second RandomGroup

ID Schema

To actually use these components, you have to create an IDSchema subclass. Just define your components under the Components class attribute and you are done.

from id_schema import OneOf, IDSchema, RandomGroup, ExactLiteral
import string

class EmailSchema(IDSchema):
	Components = (
		RandomGroup(alphabet=string.ascii_letter, length=LengthRange(1,20)),
		ExactLiteral("@"),
		OneOf(components=(
			"gmail.com",
			"yahoo.com",
			"hotmail.com"
		))
	)

class SerialNumberSchema(IDSchema):
	Components = (
		RandomGroup(alphabet=string.ascii_letter, length=5),
		ExactLiteral("-"),
		RandomGroup(alphabet=string.ascii_letter, length=LengthRange(5,10)),
		ExactLiteral("-"),
		RandomGroup(alphabet=string.ascii_letter, length=5),
	)

IDSchema exposes various functions:

• generate_one() : generate one string from the Schema
• generate_n(n): generate n strings from the same Schema
• generate_n_unique(n, max_retry=20) : generate n unique strings from the same Schema. Fail after max_retry attempts.
• validate(string, max_reductions=5000): validate if a given string matches the Schema (note that weights are not validated). DoS Protection implemented by max_reductions. Raises RuntimeError if max_reductions restriction hit.