This project is composed of three components,

• Script for fetching categories from eBay using the trading api. Will automatically store each category in a local Sqlite database.
• Dynamic HTML rendering for on-the-fly graphical representations of our categories.
• REST api for fetching categories, includes category level filtering and by default filters expired resources.

The instructions express any additional modules aside from requests are unnecessary. Because I became a little carried away and fleshed out an entire backend, I've included other modules to help simplify development.

• Flask: REST API.
• Jinja: Template preprocessor for rendering HTML.
• xml2dict: Used for converting XML into dictionaries for easier management.
• xmler: Used for converting dictionaries into XML to help better structure our requested.

Rather than using a standard pip requirements file, I leveraged Pipenv which is a relatively new tool providing an NPM-like interface for Pypi. Pipenv is available on Windows, Mac, and Linux and can be installed with pip.

Cleanup and bootstrap our database.

python -m polymath --rebuild

Render an HTML page based on the category provided.

python -m polymath --render 100100

This page consumes the same data our API displays, showing how the api could be leveraged in an actual scenario. Realistically, the user would request the topmost level category and drill downwards, which the API also supports.

Alternatively use --renderl to render the top level category without the entire tree.

Dump json payload for provided category.

python -m polymath --json 100100

To test the backend core lib with coverage, you can leverage pytest installed with the dev dependencies from our Pipfile.

cd server
pipenv install && pipenv install --dev
pipenv shell
pytest --cov=polymath.core tests

The REST api is completely dockerized, and can be built and ran without any additional overhead.

docker build . -t polymath-ebay-rest-service
docker run --rm --name polymath-rest -p 3300:3300 polymath-ebay-rest-service

Copyright (c) 2018 - present John Nolette Licensed under the MIT license.