stonefish-code-shield 0.3.45

Python code protection

Industry-grade Python code protection.

For a comprehensive license management solution that is both developer- and user-friendly, check out the Stonefish License Manager.

Quickstart

Protecting Python Packages

To protect a Python project using the Stonefish Code Shield, ensure that it includes a minimal pyproject.toml file. Then simple replace your build system (e.g., setuptools) with stonefish_code_shield in the pyproject.toml file:

[build-system]
# requires = ["setuptools"]
# build-backend = "setuptools.build_meta"
requires = ["stonefish-code-shield"]
build-backend = "stonefish_code_shield.build_meta"

# ...
# additional project metadata as per PEP 621
# <https://peps.python.org/pep-0621/>
# (recommended)
# ...

Done! Your project is now protected with Stonefish Code Shield. Try it out by running

pip install .

or

(pip install build)
python -m build . --wheel

[!TIP]

In some situations, turning off code protection can be beneficial, for example when testing your software. To turn off Stonefish Code Shield, build with

python -m build . --wheel -Cstonefish=off

When testing with tox, a possible config is

[tox]
envlist = py3

[testenv:.pkg]
passenv = *
config_settings_build_wheel =
  stonefish-code-shield = off

[testenv]
passenv = *
package = wheel
deps =
    pytest
commands =
    pytest

Protecting Standalone Python Scripts

For individual Python files, you can use the scs command-line utility:

scs /path/to/file.py

Guidelines

Here are some guidelines to keep in mind while working with Stonefish Code Shield:

• Stonefish renames class and function names, so relying on the __name__ attribute in your code won't be possible.

• Stonefish cannot yet handle relative * imports, e.g.,

  from .utils import *
  

  Make all imports explicit:

  from .utils import div_to_mod, Extractor
  

• Stonefish discards private (underscored) names from the API. If you want your users to use these variables or functions, you'll have to rename them.

• Users must add *.dat files to their package data, e.g.,

  [tool.setuptools.package-data]
  "*" = ["*.dat"]
  

  in pyproject.toml. This is because the encrypted code must be shipped with the package.

• Local imports should be relative (from . import x) instead of using absolute imports (import x) if x/ or x.py is an internal folder or directory.

• Stonefish places all data files in a flat directory structure, so it cannot handle files that are read from two different Python paths.

  ./data.dat
  
  ./a.py
      Path(__file__).parent / "data.dat"
  
  ./b/b.py
      Path(__file__).parent / .. / "data.dat"
  

More info

For details on on Stonefish Code Shield protects your code, see here.

Contact [e-mail] for more info.