A simple tool to gather symbols provided or library calls and attribute usage based on static analysis of sources of Python applications.
See [the linked article describing this tool](https://developers.redhat.com/articles/2022/01/05/extracting-information-python-source-code).
Invectio can be installed from PyPI using:
$ pip3 install invectio $ invectio --help
You can use this library as a CLI tool or as a Python module:
invectio whatprovides project-dir/ # To scan all Python files recursively for symbols provided. invectio whatprovides app.py # To perform symbols gathering on app.py file. invectio whatuses project-dir/ # To scan all Python files recursively for symbols used from libraries. invectio whatuses app.py # To perform gather symbols used from libraries on app.py file.
from invectio import gather_library_calls from invectio import gather_symbols_provided result: dict = gather_library_usage("project-dir") result: dict = gather_library_usage("app.py") result: dict = gather_symbols_provided("project-dir") result: dict = gather_symbols_provided("app.py")
As Python is a dynamic programming language, it's not possible to obtain all library functions/attributes usage simply by performing static analysis of sources. One can still perfom "crazy" things like:
import tensorflow getattr(tensorflow, "const" + "ant")("Hello, Invectio")
This library does its best to detect all function/attributes being used inside Python sources, but usage like shown above cannot be detected simply by static analysis of source code.
To create a dev environment, clone the invectio repo and install all the dependencies:
git clone https://github.com/thoth-station/invectio && cd invectio pipenv install --dev
To perform checks against unit tests present in the tests/ directory, issue the following command from the root of the git repo:
pipenv run python3 setup.py test