Make Project Installable

Introduction

In this lab, we will learn how to make a Python Flask project installable. This will make the project deployable and manageable using standard Python tools. This process is beneficial as it allows the project to be installed in different environments, manage dependencies, and isolate test environments.

Note: You need to create the code file yourself and run it in the environment. You can preview the Flask service status on Web 5000.

Skills Graph

Describe the Project

First, we need to create a pyproject.toml file to describe our project and how to build it.

The pyproject.toml file should look like this:

## pyproject.toml

[project]
name = "flaskr" ## name of the project
version = "1.0.0" ## version of the project
dependencies = [
    "flask", ## dependencies of the project
]

[build-system]
requires = ["setuptools"] ## required build system
build-backend = "setuptools.build_meta" ## backend build system

Include Necessary Files

The setuptools build backend needs another file named MANIFEST.in to include non-Python files in the project.

Create a MANIFEST.in with the following content:

## MANIFEST.in

include flaskr/schema.sql
graft flaskr/static
graft flaskr/templates
global-exclude *.pyc

This tells the build to copy everything in the static and templates directories, and the schema.sql file, while excluding all bytecode files.

Install the Project

Next, we will use pip to install the project in the virtual environment.

Run the following command in your terminal:

pip install -e .

This tells pip to find pyproject.toml in the current directory and install the project in editable or development mode. Editable mode means that as you make changes to your local code, you'll only need to re-install if you change the project's metadata.

To verify the installation, use the pip list command:

pip list

The output should show the installed project and its dependencies.

Summary

In this lab, we learned how to make a Python Flask project installable. We started by describing the project and including necessary files. We then installed the project in a virtual environment. Now, the project can be run from any location, not just the project directory.