Python local development environment construction template [Flask / Django / Jupyter with Docker + VS Code]

Overview

Building a local environment comes with some costs. Of course, I really understand the opinion that you should have some trouble here. However, I think it's a waste to lose my feelings. Therefore, I created a template for Flask / Django / Jupyter Notebook that allows you to complete ** Python development environment construction with 3 commands **. (Excluding cd)

Prerequisites

--Being MacOS --The git and make commands are valid (should be valid by default)

• If conda is installed, batting with pip may occur and it may fail in the middle.

Local build image

You will run the app on the Docker container. Even if you're not familiar with it, you can say that you're developing with a Docker container **.

--The directory is provided from the host PC to the container by bind mount. This makes it possible to read the contents of the file edited on the host PC side on the container side. --For access from the host PC, port forwarding is adopted. The connection port on the container side has been changed depending on the project, but all the ports on the host side to forward have been unified to 8000.

Construction procedure

1. Download template

Describes how to download with git clone.

Terminal

git clone [email protected]:mintak21/local-py-dev-template.git

2. Extract the template to the development directory

Specify the extraction destination directory and extract the project. Use the make command properly according to the created project.

Terminal

cd local-py-dev-template
#When creating a jupyter project
make create_jupyter_pj
#When creating a flask project
make create_flask_pj
#When creating a django project
make create_django_pj

--Specifiable arguments

Command example with arguments

Terminal

#One level higher flask_Extract the flask project to the dir directory
make create_flask_pj TARGET_DIR=../flask_dir
# ~/Expand the django project in the workspace directory and name the django project blog
make create_django_pj TARGET_DIR=~/workspace DJANGO_PJ_NAME=blog

3. Start container

Move to the directory extracted in 2, create a Docker image and container, and start it.

Terminal

cd ${TARGET_DIR}
make run

4. Connection

Since both containers are port forwarding on port 8000 of the host PC, you can connect if you can access the following, and if the contents described are displayed, the setting is completed.

Access: http: // localhost: 8000 /

5. Shut down container

Since it is running in Foreground, stopping the process on the terminal will stop the container.

Ctrl + C in the terminal

Bonus. Visual-Studio-Code setup

At this point, we have started and connected, so we will proceed with development from now on. There are also development editors such as Vim and pytorch, but here we are using Visual-Studio-Code to proceed with development. Installation is completed with one command.

Terminal

make setup_vscode # local-py-dev-Run in template directory

Setting outline

Not only installing Visual-Studio-Code, but also setting static analysis and pytest, and introducing necessary & useful extensions at the same time.

1. Install Visual-Studio-Code with brew cask
2. Package introduction (pip) --Static analysis: flake8 --import alignment: isort --Formatter: autopep8 --type check: mypy --Test: pytest
3. Extension installation (code --install-extension) --Microsoft-Python: Python extension tool (required) --Bracket Pair Colorizer: Corresponding brackets are easier to see in color --Whitespace +: Visualize tabs and whitespace --Docker: Docker can be handled with vscode --vscode-icons: Make icons easier to see --autoDocString: DocString creation assistance

Others The local side should also use a virtual environment such as pipenv, but since many of the pythonic seniors I knew were using pure Python, I introduced a virtual environment following this. Is not ...

Reference Github repository