Setting up a modular Python package under a large codebase involves organizing the package structure, managing dependencies, and setting up test environments. Here are some good options to consider:

1. Organize your package structure: Follow the standard Python package structure to maintain a clean and organized codebase. At the root level of your package, create a folder with your package name, and include an __init__.py file to make it a Python package:

home/
    my/
        project/
            my_package/
                __init__.py
                module1.py
                module2.py
            setup.py
            README.md
            .gitignore

2. Use setup.py or pyproject.toml: To manage package installations, create a setup.py file or a pyproject.toml file at the root level of your package. The setup.py file is used with setuptools, while the pyproject.toml file is used with modern build systems like Poetry or Flit.

Here's an example of a simple setup.py file:

from setuptools import setup, find_packages

setup(
    name="my-package",
    version="0.1.0",
    packages=find_packages(),
    install_requires=[
        # Add your package dependencies here
        "numpy",
        "pandas",
    ],
)

3. Use virtual environments: Create an isolated virtual environment for your package to avoid conflicts with other packages in your system. You can use venv (part of the standard library) or conda:

• For venv:

python -m venv venv
source venv/bin/activate  # Linux/Mac
venv\Scripts\activate     # Windows

• For conda:

conda create --name my-package-env python=3.8
conda activate my-package-env

4. Manage dependencies with requirements.txt or Pipfile: To keep track of your package's dependencies, you can use a requirements.txt file or a Pipfile. The requirements.txt file is a simple list of packages, while the Pipfile is used with Pipenv and provides more advanced features like specifying package versions, managing development dependencies, and automatically creating a virtual environment.

• For requirements.txt:

numpy
pandas

• For Pipfile:

[[source]]
url = "https://pypi.python.org/simple"
verify_ssl = true
name = "pypi"

[packages]
numpy = "*"
pandas = "*"

[dev-packages]
pytest = "*"

[requires]
python_version = "3.8"

5. Configure test environments: Use a testing library like pytest to write and run tests for your package. To configure the test environment, create a pytest.ini or tox.ini file at the root level of your package.

• For pytest.ini:

[pytest]
addopts = --verbose
testpaths = tests

• For tox.ini:

[tox]
envlist = py36, py37, py38

[testenv]
deps = pytest
commands = pytest

Remember to include your testing library (e.g. pytest) in your requirements.txt, Pipfile, or pyproject.toml file as a development dependency.

By following these guidelines, you'll be able to set up a modular Python package with a clean structure, properly managed dependencies, and a well-configured test environment.

Yes, using setup.py and tox.ini is a common practice and should work well for your project. Here's an example of how you can set up your Python package using these files.

1. Create a setup.py file: At the root level of your package, create a setup.py file to manage package installations and dependencies:

from setuptools import setup, find_packages

setup(
    name="my-package",
    version="0.1.0",
    packages=find_packages(),
    install_requires=[
        # Add your package dependencies here
        "numpy",
        "pandas",
    ],
    extras_require={
        "dev": [
            "pytest",
            "tox",
        ]
    },
)

In this example, the extras_require field specifies development dependencies such as pytest and tox. Users can install these dependencies by running pip install -e .[dev].

2. Create a tox.ini file: At the root level of your package, create a tox.ini file to configure the test environment and specify the Python versions you want to test against:

[tox]
envlist = py36, py37, py38

[testenv]
deps = 
    .[dev]
commands = 
    pytest

In this example, the envlist specifies the Python versions to test (3.6, 3.7, and 3.8). The deps field indicates that the test environment should install the development dependencies specified in the setup.py file. The commands field specifies that pytest should be run to execute the tests.

With this setup, you can manage your package installations and dependencies using setup.py and configure your test environments using tox.ini. To run your tests with multiple Python versions, simply execute tox in your terminal, and it will take care of setting up the virtual environments, installing dependencies, and running the tests for each Python version specified in tox.ini.

To locally check if your tox.ini file is correct, you can follow these steps:

1. Install Tox: If you haven't already installed Tox, you can do so using pip. To install Tox globally or in your current virtual environment, run:

pip install tox

2. Run Tox: In your terminal, navigate to the root directory of your project (where the tox.ini file is located), and run:

tox

Tox will read the tox.ini file, create virtual environments for each Python version specified in your envlist, install the specified dependencies, and run the tests using the specified command (e.g., pytest).

If your tox.ini file is correct, you should see output indicating the test results for each Python version. If there are any issues with the file, Tox will report errors or warnings, which you can use to debug and correct your tox.ini configuration.

3. Check for warnings or errors: Carefully review the output from Tox to ensure there are no warnings or errors related to your tox.ini file. If you find any issues, fix them in the configuration file and rerun Tox to confirm that everything is working correctly.

Keep in mind that you need to have the specified Python versions (e.g., Python 3.6, 3.7, 3.8) installed on your system for Tox to create the virtual environments and run tests. If you don't have a specific Python version installed, Tox will report an error indicating the missing interpreter.

If you're using a Continuous Integration (CI) system, make sure to include Tox in your CI configuration so that your tests are automatically run for each commit or pull request. This will help ensure that your tox.ini file and test suite remain correct and up-to-date.