pyschool

The pyschool repository demonstrates

Best Practices, Pythonic Practices, and Design Patterns implemented in Python.

Best Practices

• D.R.Y. > W.E.T.
• Convention > Configuration
• Code Smell
• Verb Smell
  • Get and Set
  • C.R.U.D.
• Client and Service
• Production
  • Git Summary
  • Style (with Black)
  • Test (with Pytest)
  • Coverage
  • Continuous Integration (CI)
  • Deployment (with PyPI)
• Docstrings
• Object Oriented Programming (OOP)
  • Open-Closed Principle (OCP)
  • Separation of concern
  • Interface versus implementation
• Singleton versus Collections, part of REST
• Test Driven Design (TDD) by Paul Everitt, 2020-06-10, PyCharm Guide

Pythonic Practices

• Iterator
• Decorator (of Python)
• List comprehension
• Error handling

Patterns

• Adapter
• Singleton
• Decorator
• Factory
  • Static Factory Method
  • Builder
  • Abstract Factory
• Publish-Subscribe (aka "PubSub" and Observer)

Examples

• Anatomy of a figure
• Animation
• Colors
• Element scale versus density
• Growth charts
• Imports
• Least squares
• Midpoint differentiation
• Oscillator
• Quartiles
• Scientific notation

Toolbox

• VIM macros

Pythonic Patterns

Attributes

• See the get_set example, which goes through a short code example, with evolution from bad, to better, to best implementations.

Client-Service

• To come.

Decorators

• An example of a logging decorator pattern, which can help D.R.Y. out code (based on this example).
• An example, using factorial, tic-toc timing convention, and memoization.
• See also Real Python, Wiki Python, Yasoob Khalid

Docstrings

• See Google docstring format
• Nice examples of Google docstring format

Error Checking (ask for forgiveness, not permission)

• See examples.

Factory

Publish-Subscribe (aka Observer)

Import

• When importing modules from non-local directories. Used often in client-service patterns.

Inheritance

• A compact example, showing a verb astraction (e.g., speak) and inheritance of behavior, with a simple client-service architecture.
• shapes

Model, View, Controller (MVC)

• shapes_mvc

Unit Test

• An example, using the unittest.mock library for testing in Python.
• Real Python article.

Examples from xyfigure_test.py:

$ python xyfigure_test.py                      # for terse interaction,
$ python -m unittest xyfigure_test             # for default interaction,
$ python -m unittest -v xyfigure_test          # for higher verbosity, and

$ python -m unittest xyfigure_test.TestImageDiff.test_same  # e.g., to test the test_same() method

Examples

Conda

• virual environment (venv)

# from [Apollo/sibl] with all folders containin __init__.py file:
(base) conda create -n temp
(base) conda activate temp
(temp) conda install numpy scipy matplotlib
(temp) conda install pip
(temp) pip install -e .  # -e is development mode, if code updates, new pip install is not required
(temp) conda list
(temp) 
(temp) conda deactivate
(base) conda remove -n temp --all
# list all environments:
(base) conda info --envs

• also, pip virtual environment (venv)

(base) python -m venv fire

Computation

• Ordinary least squares
• Several examples in the monkey_see folder

Debugger

Example: From ~/sibl/xyfigure/test in VS Code, Run | Open Configurations and add to launch.json the following:

{
    "version": "0.2.0",
    "configurations": [
        {
            "name": "Python: Current File",
            "type": "python",
            "request": "launch",
            "program": "/Users/Apollo/sibl/xyfigure/xyfigure.py",
            "args": ["signal_process_serialize.json"],
            "console": "integratedTerminal"
        }
    ]
}

Production

Once the code base has sufficient development, and it is ready for production, use the following steps:

• TLC - timing, leaks, coverage

• Test

  • Unit Test
  • Integration Test

• Packaging/Distribution/PyPI (Build, Installer), and packaging guide, and distribution with PyPI.

• Packaging and Distribution with Conda to create a Conda Package. Building conda packages from scratch.

• Client Use

  • Client Installation
  • Client Bug Reports (https://github.com/hovey/pyschool/issues)

  Production/Distribution Notes

# -----------------
# production server
# -----------------
$ cd ~/sibl/xyfigure
$ rm -r xyfigure.egg-info/
$ vim setup.py   # update setup.py, typically increment the version, located parent file README.md

# update server if necessary
$ python -m pip install --user --upgrade setuptools wheel
$ python -m pip install --user --upgrade twine

# build to the dist/ subdirectory
$ python setup.py sdist bdist_wheel

# assure the PyPI API token for the server is created on pypi.org and saved on the server at ~/.pypirc

# remove any old .gz and .whl files in dist/ subdirectory
$ cd dist/  #rm old .gz and old .whl
$ cd ../  # back to the ~/sibl/xyfigure directory

# deploy
$ python -m twine upload dist/*

# ------
# client
# ------
$ pip list
$ pip uninstall sibllib  # the old name
$ cd /nscratch/chovey/casco_sim/temp/
$ pip install --user xyfigure-0.0.4-py3-none-any.whl
$ pip list

References

• Effective-Python
• GitHub build badge
• Google Python Style Guide
• Example Google Style Python Docstrings