A fully featured twist on the classic REST API To-do list application, showcasing modern Python application development best practices and packages.
FastAPI is used to implement the REST API backend, which is a modern Python framework that is built on top of Starlette.
In addition to the ASGI performance benefits provided by Starlette, FastAPI has a very powerful dependency injection system that greatly simplifies the implementation of common patterns such as database sessions and authentication.
FastAPI also provides support for automatic OpenAPI document generation out of the box - no more manually creating Swagger manifests!
SQLAlchemy is used to implement the ORM layer
Alembic is used to manage database migrations
Pydantic is used to implement data models and schemas, which are used for request/response validation and serialization.
Docker is used to containerize the application, and a multi-stage build is used to cache mostly-static layers such as Poetry environment configuration.
Development activities are planned and tracked with Jira, starting with high-level epics which are broken down to user stories and tasks. Work is organized into sprints, where the tasks of each sprint are pulled from a backlog.
Jira automation is utilized for:
- Populating description fields of newly created tickets
- Automatically assigning tickets when the state is transitioned
- Populating Confluence pages when each sprint ends, and when an epic is completed
Additionally, Jira integrations with GitHub and Jenkins are used to track branches, commits, and build status.
Jenkins is used for CI/CD, with a multibranch pipeline approach that builds and tests each branch and PR. CI stages include package installation, linting, testing, and code scanning, driven by containerized agents which have a minimal set of dependencies required for each stage.
Package management is handled by Poetry, which is used to manage dependency versioning and virtual environments.
Dependency groups are implemented to separate dev and prod dependencies, e.g. [tool.poetry.group.test.dependencies],
allowing local development and CI to utilize packages such as pytest, while the production image only contains a
minimal set of dependencies required to run the application.
Linting is handled by Ruff, which is a wrapper around Flake8, Black, and isort. Ruff uses Rust under the hood and is ridiculously fast.
Testing is handled by Pytest, which is used to run unit and integration tests. coverage is used to generate coverage
reports, which are uploaded to SonarQube for analysis during CI.
One of the biggest benefits to using Pytest compared to other testing frameworks is the ability to use fixtures, which
can be scoped to the module, class, or function level. See tests/conftest.py for examples.
SonarQube is used to scan the codebase for bugs, vulnerabilities, and code smells. Each branch is analyzed during CI, and a webhook is configured to pass/fail CI execution based on the results of the scan.