barriers

Python decorators for including/excluding type checks, value/bounds checks, and other code blocks within the compiled bytecode of functions and methods.

Installation and Usage

This library is available as a package on PyPI:

python -m pip install barriers

The library can be imported in the usual ways:

import barriers
from barriers import barriers

Examples

Consider the function below (defined within a file f.py). The body of this function contains a code block that raises an exception if either of the two inputs is a negative integer:

def f(x: int, y: int) -> int:

    if x < 0 or y < 0:
        raise ValueError('inputs must be nonnegative')

    return x + y

We can import the module above and invoke the function to observe its behavior:

>>> from f import f
>>> f(1, 2)
3
>>> f(-1, -2)
Traceback (most recent call last):
  ...
ValueError: inputs must be nonnegative

An instance of the barriers class should normally be introduced near the top of a Python module using a pair of statements such as those below:

>>> from barriers import barriers
>>> example = barriers(False) @ globals() # Remove marked code blocks (i.e., "disable barriers").

The barriers instance example defined above is a decorator that transforms any decorated function by removing any designated code blocks in the body of that function.

• The False argument in the expression barriers(False) above should be interpreted to mean that this barrier is disabled (i.e., that the marked code blocks in the bodies of functions decorated by this decorator should be removed). The default value for this optional argument is True; this should be interpreted to mean that this barrier is enabled (and, thus, that marked code blocks should not be removed from decorated functions).
• The notation @ globals() ensures that the namespace returned by globals is used when compiling the abstract syntax trees of transformed functions.

Consider the modified version of f.py below. A statement can be designated for automatic removal by placing a marker -- in this case, the example variable -- on the line directly above that statement. Note that in the body of the function f defined below, the if block is immediately preceded by a line that contains the variable example:

from barriers import barriers
example = barriers(False) @ globals()

@example
def f(x: int, y: int) -> int:

    example
    if x < 0 or y < 0:
        raise ValueError('inputs must be nonnegative')

    return x + y

The decorator @example automatically removes the if block in the function above. As a result, the function does not raise an exception when it is applied to negative inputs:

>>> from f import f
>>> f(1, 2)
3
>>> f(-1, -2)
-3

It is also possible to define and use individually named markers (which are created as attributes of the barriers instance):

>>> from barriers import barriers
>>> checks = barriers(types=True, bounds=False) @ globals()

Given the above definitions, it is possible to introduce named markers such as those in the variant of f.py presented below. When a marker definition has been assigned True, the statements immediately below that named marker are not removed (i.e., the marked barrier statements are enabled). When a marker definition has been assigned False, the corresponding marked statements are removed:

from barriers import barriers
checks = barriers(types=True, bounds=False) @ globals()

@checks
def f(x: int, y: int) -> int:

    checks.types
    if not isinstance(x, int) and not isinstance(y, int):
        raise TypeError('inputs must be integers')

    checks.bounds
    if x < 0 or y < 0:
        raise ValueError('inputs must be nonnegative')

    return x + y

The described behavior can be observed when evaluating the function:

>>> from f import f
>>> f('a', 'b')
Traceback (most recent call last):
  ...
TypeError: inputs must be integers
>>> f(-1, -2)
-3

Many additional details and examples are presented in the documentation.

Development

All installation and development dependencies are fully specified in pyproject.toml. The project.optional-dependencies object is used to specify optional requirements for various development tasks. This makes it possible to specify additional options (such as docs, lint, and so on) when performing installation using pip:

python -m pip install .[docs,lint]

Documentation

The documentation can be generated automatically from the source files using Sphinx:

python -m pip install .[docs]
cd docs
sphinx-apidoc -f -E --templatedir=_templates -o _source .. && make html

Testing and Conventions

All unit tests are executed and their coverage is measured when using pytest (see the pyproject.toml file for configuration details):

python -m pip install .[test]
python -m pytest

Alternatively, all unit tests are included in the module itself and can be executed using doctest:

python src/barriers/barriers.py -v

Style conventions are enforced using Pylint:

python -m pip install .[lint]
python -m pylint src/barriers

Contributions

In order to contribute to the source code, open an issue or submit a pull request on the GitHub page for this library.

Versioning

The version number format for this library and the changes to the library associated with version number increments conform with Semantic Versioning 2.0.0.

Publishing

This library can be published as a package on PyPI by a package maintainer. First, install the dependencies required for packaging and publishing:

python -m pip install .[publish]

Ensure that the correct version number appears in pyproject.toml, and that any links in this README document to the Read the Docs documentation of this package (or its dependencies) have appropriate version numbers. Also ensure that the Read the Docs project for this library has an automation rule that activates and sets as the default all tagged versions. Create and push a tag for this version (replacing ?.?.? with the version number):

git tag ?.?.?
git push origin ?.?.?

Remove any old build/distribution files. Then, package the source into a distribution archive:

rm -rf build dist src/*.egg-info
python -m build --sdist --wheel .

Finally, upload the package distribution archive to PyPI:

python -m twine upload dist/*