nautica 3.2.6

A service management framework

Nautica V3

Nautica3 is a backend platform for Python. It gives you a managed runtime environment with a service registry, lifecycle system, CLI, and built-in tools, saving you time from putting your app together and giving you more time to build it.

Features

• Service Registry A System for managing plugins and their lifecycle and runtime. You can install 3rd party packages from the Package Registry, or drop in .py files into plugins/.

• Built-in Services Includes the HTTP and WebSocket server, scheduler and more. Everything can be configured or disabled entirely.

• Config Manager A TOML configuration system with auto key management, live read and writes.

• Log Manager Provides a readable console and file output. Automatically resolves module from which it is being called.

• Shell Gives you a way to interact with the server. Additionally you can use the TUI to view all the workers and running threads.

And more! Learn more

---

I made Nautica because I was solving the same problems in my projects, those being: configs, logging, figuring out startup orders, not to mention the validation boilerplate on every route. Because of this I made Nautica V2, which solved many of these issues, and V3 to improve on the idea.

Get Started

Firstly, install Nautica:

pip install nautica

To create a project, use the Nautica CLI (docs):

nautica create my-project --demo
cd my-project
nautica run

Or continue working on an already existing project by installing it:

cd my-project
nautica install

This will automatically download all needed packages and create associated configuration files/

---

How Nautica Compares

Benchmark Performance

Framework	Requests/Second	Avg Latency	Overall
Nautica3	3929	2.5ms	Baseline
FastAPI	3154	3.2ms	24% slower, 19% higher latency
Flask	75	132.6ms	50x slower, 98% higher latency

More in-depth report is available here

Code Complexity

Similarly to SvelteKit, Nautica defines the route names for you. This helps to reduce boilerplate (see Flask and FastAPI examples), improve readability, and helps with naming conventions

Nautica3

# file: src/http/api/v1/auth.py
from napi.http import HTTP, Context, Reply, Error
from somewhere import username, password

@HTTP.POST()
@HTTP.Require(body = {"username": str, "password": str})
def login(ctx: Context):
    if ctx.body["username"] == username and ctx.body["password"] == password:
        return Reply(ok=True)
    raise Error(401, "Invalid credentials")

Flask

# file: main.py
from flask import Flask, request
from flask.blueprints import Blueprint
import json
from somewhere import username, password

app = Flask(__name__)
v1auth = Blueprint("v1auth", __name__, url_prefix="/api/v1/auth")

@v1auth.post("/login")
def login():
    data = request.get_json(silent=True) or {}
    if data.get("username") == username and data.get("password") == password:
        return json.dumps({"ok": True})

    return json.dumps({"ok": False, "error": "Invalid credentials"}), 401

app.register_blueprint(v1auth)
app.run(port=8101)

Flexibility

Request Validation

FastAPIs Pydantic models are fine..

from pydantic import BaseModel

class ListUsersModel(BaseModel):
    format: str
    limit: int

@app.get("/api/users")
async def list_users(body: ListUsersModel):
    ...

But Nautica's Requirement system gives you more freedom:

from napi.http import HTTP, Context, Require

@HTTP.POST()
@HTTP.Require( body = {"format": Require.AnyOf("dict", "list"), "limit": int} )
async def users(ctx: Context):
    ...

Which gives you access to default type checking + 6 Requirement validators:

AnyOf(*options: any), AnyTypeOf(*types: type), ExactMatch(match: any), RegExMatch(regex: str), File(max_size?: int, mime?: str) and ListOf(obj: type | dict | Requirement, max_length?: int, max_length?: int).

You can implement your custom validators on top of the system, learn more on the wiki.

Response Validation

Similarly, FastAPI uses type hints to determine the response of a route..

from pydantic import BaseModel

class AuthResponse:
    session: str

@app.post("api/auth/login")
def login(body: LoginBody) -> AuthResponse:
    ...

Nautica instead uses a decorator-based approach:

from napi.http import HTTP, Context, Error, ReplyModel

@HTTP.POST()
@HTTP.Responses(
    ReplyModel(200, {"session": str}),
    Error(401)
)
def login(ctx: Context):
    ...

If you want to prevent faulty responses, you can use the strict mode:

@HTTP.Responses(
    ...,
    strict = True
)
# Cancels all responses that don't conform to a schema defined above