Pokie 1.1.0

A high-level Python REST web framework based on Flask, Rick and RickDb.

Welcome to Pokie

Pokie is an REST web framework built on top of Flask, Rick and Rick-db libraries, following three-layer and clean architecture design principles.

It features an object-oriented design, borrowing from common patterns found in other languages, such as dependency injection, service location, factories and object composition. It also offers the following functionality:

• Modular design;
• Dependency registry and factories;
• Configuration via environment variables or JSON files;
• CLI command support;
• Events and signal handlers;
• Caching (Redis, in-memory, dummy);
• Jobs with per-job intervals, retry/backoff, and timeouts;
• CORS and rate limiting via built-in factories;
• OpenAPI 3.0 spec generation from registered routes;
• Fixtures;
• Unit testing support with pytest;
• Code generation;
• Automatic REST endpoint generation;
• REST-oriented service design;
• Compatibility with Flask;
• Forward-only SQL migrations;
• PostgreSQL support;

For detailed information, please check the Documentation

Automatic REST from Database Tables

Generate full CRUD endpoints directly from a database table — no DTO, no RequestRecord, no service class needed:

from pokie.core import BaseModule
from pokie.http import AutoRouter
from pokie.rest.auto import Auto


class Module(BaseModule):
    name = "my_module"

    def build(self, parent=None):
        app = parent.app

        # generate a full REST view from the "customers" table
        view = Auto.view(app, "customers")
        AutoRouter.resource(app, "customer", view, id_type="string")

This introspects the customers table at startup and registers:

URL	Method	Operation
/customer	GET	List records
/customer/<string:id_record>	GET	Get by id
/customer	POST	Create
/customer/<string:id_record>	PUT, PATCH	Update
/customer/<string:id_record>	DELETE	Delete

Listing supports server-side pagination, sorting, filtering and free-text search out of the box via query parameters (offset, limit, sort, match, search).

Automatic REST from DTO Records

For more control, use Auto.rest() with a DTO Record — Pokie generates the RequestRecord and service automatically:

from rick_db import fieldmapper
from pokie.core import BaseModule
from pokie.rest.auto import Auto


@fieldmapper(tablename="customers", pk="customer_id")
class CustomerRecord:
    id = "customer_id"
    company_name = "company_name"
    contact_name = "contact_name"


class Module(BaseModule):
    name = "my_module"

    def build(self, parent=None):
        Auto.rest(
            parent.app,
            "customer",
            CustomerRecord,
            search_fields=[CustomerRecord.company_name, CustomerRecord.contact_name],
            id_type="string",
        )

Both approaches can be incrementally customized — add a custom RequestRecord for input validation, a custom service for business logic, or a custom base class for authentication.

Getting Started

1. Create the application entrypoint, called main.py:

from rick.resource.config import EnvironmentConfig
from pokie.config import PokieConfig
from pokie.core import FlaskApplication
from pokie.core.factories.pgsql import PgSqlFactory


class Config(EnvironmentConfig, PokieConfig):
    pass


def build_pokie():
    cfg = Config().build()

    modules = [
        # add your modules here
    ]

    factories = [
        PgSqlFactory,
    ]

    pokie_app = FlaskApplication(cfg)
    flask_app = pokie_app.build(modules, factories)
    return pokie_app, flask_app


main, app = build_pokie()

if __name__ == '__main__':
    main.cli()

2. Scaffold a module:

$ python3 main.py codegen:module my_module_name .

3. Add the module to the module list on main.py:

    modules = [
        'my_module_name',
    ]

4. Implement the desired logic in the module

Running tests with tox

1. Install tox & tox-docker:

$ pip install -r requirements-test.txt

2. Run tests:

$ tox [-e py<XX>]