typechecks

Check Python values at runtime and raise named errors when arguments have the wrong type. This is a small legacy utility, rather than a static type checker.

Install

python -m pip install typechecks

Check and validate arguments

from typechecks import is_integer, is_string, require_integer, require_string

print(is_string("sample"), is_integer(3))
require_string("sample", name="label", nonempty=True)
require_integer(3, name="count")
try:
    require_string(3, name="label")
except TypeError as error:
    print(type(error).__name__)

True True
TypeError

The is_* functions return booleans. The require_* functions return None on success and raise TypeError for a mismatched type. An empty string with nonempty=True raises ValueError. name labels the argument in the error.

Validate a collection

from typechecks import require_iterable_of

labels = ["tumor", "normal"]
require_iterable_of(labels, str, name="labels")
print("valid")

valid

Use a reusable sequence such as a list: this implementation may traverse the input twice when a value is invalid. It is unsuitable for validating one-shot iterators without first materializing them.

Type semantics

Checks use isinstance(), so subclasses pass. Python booleans are integer subclasses: is_integer(True) returns True, and require_integer(True) succeeds. Integer validation does not check positivity or reject zero; make those value checks in your application.

Use require_instance(value, types, name=...) for another class or a tuple of classes. The API reference gives the actual supported arguments.

Documentation development

Install requirements-docs.txt and run ./docs.sh; execute examples with python scripts/check_docs_examples.py.