Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Python packaging and distribution #26

Merged
merged 17 commits into from
Sep 18, 2024
Merged
Show file tree
Hide file tree
Changes from 15 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions docs/source/index.rst
Original file line number Diff line number Diff line change
Expand Up @@ -7,3 +7,4 @@ Welcome to the LASP Developer's Guide!
:maxdepth: 1

licensing
programming_languages/index
7 changes: 7 additions & 0 deletions docs/source/programming_languages/index.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
Programming Languages
=====================

.. toctree::
:maxdepth: 1

python/index
8 changes: 8 additions & 0 deletions docs/source/programming_languages/python/index.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,8 @@
Python
=====================

.. toctree::
:maxdepth: 1

terminology
packaging_and_distribution
185 changes: 185 additions & 0 deletions docs/source/programming_languages/python/packaging_and_distribution.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,185 @@
> **Warning:** More information is needed to complete this guideline

# Python Packaging and Distribution
Examples of Python packaging and distribution options and how use them.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Minor update: add "to" after "how"


## Purpose
> **Warning** Need to add an explanation of how this guideline supports DS workflows, meets internal and external
> policies, and aids in collaboration and our overall success

## Options
The options for Python packaging and distribution that we often see used at LASP are:
- [PyPI](#packaging-for-pypi--pip-install-)
- [Conda](#packaging-for-conda--conda-install-)

## Packaging for PyPI (`pip install`)

### PyPI resources:

- [PyPI Help Page](https://pypi.org/help/)

- [Setting up a PyPI account](https://pypi.org/account/register/)

- [Getting a PyPI access token](https://pypi.org/help/#apitoken)


### Built-In (`build` + `twine`)

> **Warning**: Need to add introductory paragraph that summarizes Built-In

#### How to use Built-In
Python Packaging User Guide: https://packaging.python.org/en/latest/
The link below is a fairly complete tutorial. There are also instructions there for using various other build tools:
https://packaging.python.org/en/latest/tutorials/packaging-projects/

#### Built-In resources

- [Python Packaging User Guide](https://packaging.python.org/en/latest/)

#### Setuptools Example – Library Package
<details>
<summary>setup.py</summary>

```
"""
Setup file for the science data processing pipeline.

The only required fields for setup are name, version, and packages. Other fields to consider (from looking at other
projects): keywords, include_package_data, requires, tests_require, package_data
"""
from setuptools import setup, find_packages

# Reads the requirements file
with open('requirements.txt') as f:
requirements = f.read().splitlines()

setup(
name='my_py_library',
version='0.1.0',
author='Jane Doe, John Doe, This is just a str',
author_email='[email protected]',
description='Science data processing pipeline for the instrument',
long_description=open('README.md', 'r').read(), # Reads the readme file
python_requires='>=3.8, <4',
url='https://some-git.url',
classifiers=[
"Natural Language :: English",
"Topic :: Scientific/Engineering",
"Topic :: Scientific/Engineering :: Astronomy",
"Programming Language :: Python :: 3.8",
"Operating System :: MacOS :: MacOS X",
"Operating System :: POSIX :: Linux",
],
packages=find_packages(exclude=('tests', 'tests.*')),
package_data={
"my_py_library": [
"some_necessary_config_data.json",
"calibration_data/*"
]
},
py_modules=['root_level_module_name',],
install_requires=requirements,
entry_points={
'console_scripts': [
'run-processing=my_py_library.cli:main', # package.module:function
]
}
)
```
</details>

### Poetry

> **Warning**: Need to add introductory paragraph that summarizes Built-In

[Poetry Build and Publish Docs](https://python-poetry.org/docs/cli/#build)

How to Publish to PyPI from Poetry
```
poetry lock
poetry install
poetry version
poetry build
PYPI_USERNAME=__token__
PYPI_TOKEN=<token-copied-from-pypi-account>
poetry publish # You will be prompted for your PyPI credentials if you don't provide the environment variables
```

#### Poetry Project Configuration Example – Library Package

<details>
<summary>pyproject.toml</summary>

```
# pyproject.toml
# See: https://python-poetry.org/docs/pyproject/

[tool.poetry]
name = "my_python_package"
version = "0.1.0"
description = "Science data processing library and applications for some instrument."
authors = [ # Alphabetical
"Jane Doe <[email protected]>",
"John Doe <[email protected]>"
]

# Configure private PyPI repo to download packages
[[tool.poetry.source]]
name = "lasp-pypi" # This name will be used in the configuration to retrieve the proper credentials
url = "https://artifacts.pdmz.lasp.colorado.edu/repository/lasp-pypi/simple" # URL used to download your private packages

# Dependency specification for core package
[tool.poetry.dependencies]
python = "^3.9"
astropy = "^4.2.1"
h5py = "^3.3.0"
numpy = "^1.21.0"
spiceypy = "^4.0.1"
lasp-packets = "1.2"
requests = "^2.26.0"
SQLAlchemy = "^1.4.27"
psycopg2 = "^2.9.2"
cloudpathlib = {extras = ["s3"], version = "^0.6.2"}

# Development dependencies
[tool.poetry.dev-dependencies]
pytest-cov = "^2.12.1"
pylint = "^2.9.3"
responses = "^0.14.0"
pytest-randomly = "^3.10.2"
moto = {extras = ["s3"], version = "^2.2.16"}

# Script entrypoints to put in installed bin directory
[tool.poetry.scripts]
sdp = 'my_python_package.cli:main'

# Poetry boilerplate
[build-system]
requires = ["poetry-core>=1.0.0"]
build-backend = "poetry.core.masonry.api"
```
</details>

## Packaging for Conda (`conda install`)
> **Warning**: Need a volunteer

### How to install and use Conda
https://conda.io/projects/conda-build/en/latest/user-guide/tutorials/build-pkgs.html

> Conda Develop:
> There is a conda subcommand called `conda develop`, but it is not actively maintained. The maintainers of
conda recommend using `pip install` to install an editable package in development mode.
> See: https://github.com/conda/conda-build/issues/1992

## Useful Links
Here are some helpful resources:

- [Python Packaging User's Guide](https://packaging.python.org/en/latest/)
- [The Hitchhiker's Guide to Python - Packaging your Code](https://docs.python-guide.org/shipping/packaging/)
- [The Sheer Joy of Packaging](https://python-packaging-tutorial.readthedocs.io/en/latest/index.html)
- [Package Python Projects the Proper Way with Poetry](https://hackersandslackers.com/python-poetry-package-manager/)
- [Poetry Documentation](https://python-poetry.org/docs/)
- [Setuptools Documentation](https://setuptools.pypa.io/en/latest/)
- [Building conda packages from scratch](https://conda.io/projects/conda-build/en/latest/user-guide/tutorials/build-pkgs.html)

Credit: Content taken from a Confluence guide written by Gavin Medley
40 changes: 40 additions & 0 deletions docs/source/programming_languages/python/terminology.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,40 @@
# Python Terminology

Some Python terminology that a user might encounter, particularly when working through this Python guide.

## Purpose
Like all programming languages, Python has some terminology that is unique to it and it is helpful to have that language
explained. This page may be updated over time so that it holds the most useful terminology to those that use this
developer's guide.

### Library Packages vs Applications vs Scripts

Python is a highly flexible interpreted language. This means it's easy to get started, quick to write, and easy to run.
Unfortunately, this also means it's very easy to write code that works in one context but not in others or code that is
robust but isn't designed to be inherited.

**Package**: A `python` package, intended for redistribution, containing objects that serve as building blocks for
other developers to use. Examples are `numpy`, `pytest`, and `sqlalchemy`.

> **Note** While most people think of a Python package as the above definition, a Python "package" is technically
> any directory containing an `__init__.py` file.

**Application**: A python project (which may or may not be a packaged distribution) that provides specific and possibly
configurable functionality. Examples are Poetry, the AWS CLI, the Conda CLI, the Green Unicorn WSGI HTTP server, any
Django "app".

**Script**: Pretty much anything else written in Python. One could arguably say that a Script is just a trivial Application
with little configuration or portability. Scripts are usually run with `python my_script.py argv`. They tend to be
difficult to maintain, update, or distribute.

**Packaging Tooling**: Not only managing a local environment, but also providing tooling for developing, building, and
distributing python packages for other users. While Conda does support this use case (it's how one creates and
distributes conda packages to `conda-forge`), Poetry and `setuptools` are much easier to develop with (for PyPI) and Poetry
boasts a similar dependency resolver to Conda. One major drawback to Conda in packaging is that there is no notion of an
"editable install" so the developer is forced to build and test end user functionality (e.g. script entrypoints) over
and over instead of simply making code changes in place.

## Useful Links
Helpful links to additional resources on the topic

Credit: Content taken from a Confluence guide written by Gavin Medley
4 changes: 3 additions & 1 deletion guideline_template.md
Original file line number Diff line number Diff line change
Expand Up @@ -13,4 +13,6 @@ An explanation of the options available for this guideline (could be one or more
Detailed instructions or general guidance for implementation of the guideline

## Useful Links
Helpful links to additional resources on the topic
Helpful links to additional resources on the topic

Credit: Content taken from a Confluence guide written by <add_name>
Loading