Welcome to Pyskema’s documentation!
About
Pyskema is a small module used to define schema for structured data.
The schema can be used to validate, document and generate data.
Installation
Use pip install pyskema.
Usage
Here is a simple example of defining a model of data:
from pyskema import Node, AtomType
model = Node.of_record({
"number of schleem": Node.of_atom(AtomType.INT),
"length of dinglepop": Node.of_atom(AtomType.FLOAT),
"color of fleeb": Node.of_atom(AtomType.OPTION, [
"pink",
"red",
"octarine",
]),
})
The model can be used to validate a concrete piece of data from any source (see pyskema.validate).
from pyskema.validate import validate, InvalidDataError
from json import load
with open("plumbus.json") as f:
data = load(f)
try:
validate(data, model, fail=True):
except InvalidDataError as e:
print("The data does not match:")
explanation, = e.args
print(explanation)
else:
print("all good!")
For documentation purpose, you can automatically describe a schema (see pyskema.describe).
from pyskema import Node, AtomType, describe
model = Node.of_record(
{
"n_schleem": Node.of_atom(
AtomType.STR,
description="number of dangling schleems",
),
"l_ding": Node.of_atom(
AtomType.FLOAT,
description="length of a the dinglepop in mm",
),
"c_fleeb": Node.of_atom(
AtomType.OPTION,
[
"pink",
"red",
"octarine",
],
description="color of the fleeb",
),
},
description="a plumbus specification"
)
# >>> describe(model)
# a plumbus specification
# n_schleem: number of dangling schleems
# value of type str
# l_ding: length of a the dinglepop in mm
# value of type float
# c_fleeb: color of the fleeb
# one of 'pink', 'red', or 'octarine'
You can also clone an existing model with modifications with methods pyskema.schema.Node.delete() and pyskema.schema.Node.inject().
More details can be found in the reference.