ose-workbench-platform 0.1.0a9

Common platform for developing Open Source Ecology (OSE) workbenches.

OSE Workbench Platform

A platform for developing workbenches for Open Source Ecology (OSE).

OSE defines a "workbench" as a set of tools in CAD software to design and make a particular machine.

Each workbench OSE develops for one of it's machines has certain common development-time or "dev-time" needs and dependencies.

For example, running unit tests, making documentation, and generating code to streamline workbench development.

Rather than duplicate the approaches to each of these needs, ose-workbench-platform abstracts those needs into a common platform so they aren't the concern of individual OSE workbench maintainers.

Each workbench maintainer doesn't need to know or care about the particular versions and libraries we use to solve those needs, nor the particular configuration.

Having a common platform for OSE workbench development also makes it easier for developers to readily switch between workbenches by providing a common tool-set.

ose-workbench-platform provides a command-line interface (CLI), via the osewb command, containing commands for common dev-time tasks such as running all tests, making documentation, initializing new workbenches, and even generating code for common tasks.

Pre-Requisites

1. Install Git
2. Install Python
3. Install Docker

Installation

pip install ose-workbench-platform

After installing ose-workbench-platform, you'll have access to the osewb command.

See Commands for additional information.

Docker

We use Docker to create a reproducible virtualized OSE workbench development environment with requisite dependencies for development-time tasks like running unit tests, and generating documentation from source-code comments.

While having to install and learn Docker is a burden, our hope is that it's simpler than developers having to manually setup their own local development environment with the requisite dependencies for OSE workbench development.

Unit Tests

For running unit tests we use pytest.

For test coverage, we use coverage.py and pytest-cov.

Documentation

For building documentation, we use Sphinx.

For hosting documentation, we use a free service for open-source projects called Read the Docs.

For a modern and mobile-friendly look, we use Read the Docs Sphinx Theme.

Commands

The osewb command contains various sub-commands for performing common dev-time tasks of a OSE workbench.

$ osewb -h ↵
usage: osewb <command> [<args>]

A collection commands for OSE workbench development.

optional arguments:
  -h, --help            show this help message and exit

Commands:
  {init,test,docs,container}
    init                Initialize new workbench
    test                Run tests in workbench
    docs                Make documentation
    container           Commands for interacting with containers

Each sub-command may have flags and arguments, and additional information can be discovered via osewb <command> -h or --help.

Container

OSE Workbench Platform includes a container command to make interacting with the requisite container for running tests and building documentation easier.

$ osewb container -h ↵
usage: osewb container <command>

optional arguments:
  -h, --help      show this help message and exit

Commands:
  {image,create}
    image         Build image for container
    create        Create container -- must be in workbench repository

In order to run the test and docs commands, you must first:

1. Build the ose-workbench-platform image using osewb container image
2. Create a container for your workbench by running osewb container create within the workbench repository

Note, you only need one ose-workbench-platform image that acts as a blue-print for creating a container per workbench.

Workbench containers will be named after the base package in the workbench repository (e.g. ose3dprinter, osetractor, osepowercube, etc.).

Once the container for the workbench is created, you may interact with it using normal docker commands.

Some common operations you may wish to perform on the container are:

Description	Command
Starting the container	docker start <container name>
Stopping the container	docker stop <container name>
Removing the container	docker rm <container name>

Refer to the Docker CLI reference documentation for additional information.

Test

OSE Workbench Platform includes a test command for interacting with the test-suite of a workbench.

$ osewb test -h ↵
usage: osewb test

optional arguments:
  -h, --help      show this help message and exit
  -c, --coverage  Run tests with coverage, and generate report

To run the entire unit-test suite for a workbench, run:

osewb test

For running tests with coverage and generating a coverage report, pass the -c or --coverage flag to the test command:

osewb test --coverage

Docs

OSE Workbench Platform includes a docs command for building the documentation of a workbench.

osewb docs

The docs command will:

• Generate property tables for each Model class in the workbench and output them as .csv files in docs/property_table/
• Delete the docs/_build/ directory
• Delete the docs/<base package>/ directory
• Re-generate docs/_build/ and docs/<base_package></base_package>/ by running sphinx-build . _build within docs/ using the Sphinx configuration specified in docs/conf.py

For additional information, see sphinx-build and Sphinx Configuration.

Init

OSE Workbench Platform includes a init command for initializing a new workbench.

Navigate to where you want to initialize a directory for your new workbench. Then run:

osewb init

You'll be prompted to enter the machine name in Title Case.

machine_display_name [CEB Brick Press]: Tractor ↵

Then, you'll be prompted for several more values.

Each successive prompt derives it's default value from answers to previous prompts.

You can press the Enter key for most prompts to stick with the defaults.

repo_name [ose-tractor-workbench]: ↵
machine_title [OSE Tractor]: ↵
workbench_title [OSE Tractor Workbench]: ↵
base_package [osetractor]: ↵
command_registry_filename [OSE-Tractor]: ↵
command_namespace [OSETractor]: ↵
workbench_class_filename [tractor_workbench]: ↵
workbench_class_name [TractorWorkbench]: ↵

The above examples initializes a new workbench, in a ose-tractor-workbench directory, with the basic structure and files needed for the workbench.

ose-tractor-workbench
    ├── InitGui.py
    ├── osetractor
    │   ├── app
    │   │   ├── __init__.py
    │   │   └── part
    │   │       ├── box
    │   │       │   ├── box.py
    │   │       │   └── __init__.py
    │   │       └── __init__.py
    │   ├── gui
    │   │   ├── command
    │   │   │   ├── add_box
    │   │   │   │   ├── add_box_command.py
    │   │   │   │   └── __init__.py
    │   │   │   └── __init__.py
    │   │   ├── icon
    │   │   │   ├── Box.svg
    │   │   │   └── __init__.py
    │   │   ├── __init__.py
    │   │   ├── OSE-Tractor.py
    │   │   └── tractor_workbench.py
    │   └── __init__.py
    └── README.md

Generating Code

Within the root of a workbench repository, run the make command.

For example,

osewb make command AddMotor

Will generate a new AddMotor command class.

Developing Locally

Run the following command from the root of the repository:

pip install --editable .

See "Editable Installs" on the pip docs for additional information.

This will give you access to the osewb command locally for testing any changes to the source code.

Deploying to PyPi

From the root of repository:

1. python setup.py sdist

2. twine upload dist/*

   • pip install twine (if not already installed)

You'll be prompted for your PyPi username and password:

Enter your username: gbroques ↵
Enter your password: ↵