sckott / habanero

client for Crossref search API

Home Page:https://habanero.readthedocs.io

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Possibly use module package architecture instead of classes

sckott opened this issue · comments

might be simpler, like did for pygbif

This would be suitable for static helper functions (like date transform) but I think classes is best, especially when you might start to get state involved (e.g. deep paging or query builders).

Suggest probably a combination of both: module functions for despatching queries, but class hierarchies for results, and operations that can happen on results. Getting state mixed up with modules is a bad idea.

@afandian Thanks for the feedback. Yeah, esp. maintaining state alone is a good reason for using classes.