- Create src directory consisting of all application code
- Create test directory consisting of all the tests
- create requirements file consisting of all the requirements
- Creating a make file, so when someone else copies the source code into their machine they can check everything is working(github action file but for local machines)
- Docker file(with necessary ports exposed)
- Creating a utilscli.py, where we can test the application with python cli commands
- [Optional]Creating a *.sh file where we have command prompt code (like curl) to test the server
- [Optional] If we are to deploy it to a cloud service then we can have Iac(Infrastructure as a code file )
- Create .github/workflows/python-app.yml file with CI/CD code
- the code in the yml file can have multiple jobs and for each job there are multiple tasks
- We will have once job and the tasks are
-
- deep clone our repository from github to the ubuntu machine,
- download and setup python on the machine,
- install the requirements.txt packages on the machine
- install pytest; export our src directory to python path; run the tests
- we can chose a cloud platform, (Heroku in this case) make sure it supports infrastructure as code
- Get the api keys to interact with the cloud account through code
- create an app on the cloud platform
- Upload the name of the app and the secret key in GitHub secrets
- in yml file
- set environment variables of our secret key and heroku app name
- If the status up and until the above process is true and the branch is main perform the below actions
- Add link of the github repo of the heroku app as remote link to our project
- push the code from main branch of our repo to main branch of heroku repo, Forcefully
<name of the action>
on:
pull_request:
push:
branches:
- <branch name>
jobs
<name of job>
runs-on: ubuntu-latest
steps
- uses
- name
- with:
python-version:3.6
- name:
run: <command>
- name:
env:
var1=value
run: |
<command>
<command>
- Make sure to remove linting checks if not required