typefit_hn

Hacker News API Client

This example app is a demo of an API client that can be used to parse and understand items coming from the Hacker News API.

All models are created according to the official documentation and describe the data in order to fit it into Python types.

The API client itself is a very simple API client that only supports getting items. The way it is defined will fit the data into the right class from models.

Usage

Then you can pretty simply use this library.

from typefit_hn import HackerNews

hn = HackerNews()
story = hn.get_item(42)

print(f"{story.title} ({story.score}): {story.url}")

Models

The models are based on the Python dataclasses module because they work quite well with inheritance and there is a lot of shared fields and logic between all types, since they all are defined as "items" in the original API.

There is a BaseItem class which will, in its __post_init__(), check if the type of the item corresponds to the statically defined TYPE that each sub-class holds. If not, a ValueError is raised which allows Typefit to reject fitting data into this class and to try another one, even if all the fields were matched.

Another subtle thing comes in the difference between Story and Ask. Both those types have the same type according to the API, however one will have an url and the other will have a text. For this reason, in the __post_init__() of the Story there is a check to make sure that a Story cannot have an URL by raising a ValueError if so. This will make sure that the difference between Story and Ask is well taken in account.