Skip to content

Latest commit

 

History

History
66 lines (44 loc) · 3.17 KB

DEVELOPMENT.md

File metadata and controls

66 lines (44 loc) · 3.17 KB
Raw

Development

Zeno's backend is a Python FastAPI server.

Development Setup

Note: We recommend you use Docker as detailed in the overall development setup instead of setting this up locally.

Please read the overall development setup before reading this. Then follow the steps outlined here to run the backend locally.

The backend works with a postgres database. To install postgres on macOS, use Homebrew: brew install postgresql. Then, start the postgres service using brew services start postgresql.

To set up the server:

  1. Enter a psql env psql postgres.
  2. Add a new user CREATE ROLE [username] WITH LOGIN PASSWORD '[password]';.
  3. Give the user rights to create a database ALTER ROLE [username] CREATEDB;.
  4. Exit psql \q and re-enter it from that user psql postgres -U [username].
  5. Create a database CREATE DATABASE zeno;.
  6. Grant your user privileges GRANT ALL PRIVILEGES ON DATABASE zeno TO [username];.
  7. Bootstrap the initial tables by pasting the commands in init.sql into the PSQL terminal.

To inspect your data, you can install the Postico GUI with brew brew install --cask postico. The default host is localhost and port is 5432.

Once this backend database is set up, the backend needs information on how to connect to it. Therefore, add a database.ini in ./zeno_backend/database as follows:

[postgresql]
host=[whereever your DB is running, e.g. localhost:5432]
dbname=[the name of your DB]
user=[usename for DB login]
password=[password for DB login]

For the backend to be able to verify user status, the ZENO_USER_POOL_AUTH_REGION, ZENO_USER_POOL_ID, and ZENO_USER_POOL_CLIENT_ID environment variables are needed. To simplify the dev setup, the zeno backend will try to read these from ../frontend/.env.

For a debug setup, you can run uvicorn --factory zeno_backend.server:get_server --reload from within the poetry shell.

Static Analysis Tools

Ruff (linting), black (formatting), mypy (typechecking)

Testing

pytest

Guidelines:

  • Use Type Annotation: Extensive type annotation is a key way to catch bugs in Python programs.
  • Prefer Data Structures over Dicts: Dicts are flexible but don’t allow for type checking using static analysis. Instead, prefer using pydantic classes, which allows you to easily specify data structures.
  • Documentation: All code should be documented (we will use the Google docstring standard). This will be enforced by linting.

Deployment

To deploy the backend, you can either use the same setup as outlined for local development, or set environment variables instead of creating a database.ini. You will have to set the following variables: DB_HOST, DB_NAME, DB_USER, and DB_PASSWORD.

Additionally, you can configure the host and port for the backend using BACKEND_HOST and BACKEND_PORT, if not set, these default to 0.0.0.0 and 80, respectively.

To make sure that your frontend can access content from this backend even if running on another origin, use the CORS_ORIGIN environment variable to set CORS headers when requests come from the specified origin.