Skip to content

structure.md

Latest commit

 

History

History
72 lines (49 loc) · 2.27 KB

structure.md

File metadata and controls

72 lines (49 loc) · 2.27 KB
Raw

Structure of the project

In this list I'll try to explain how the project works and what is where.

What are you using?

The server side is running Falcon a framework to build APIs using python. I've chosen Falcon because of how fast it is and because of it extensibility.

As ORM we run peewee. It's a mature library, well manteined and documentated.

To manage different asyncronous task we run huey. It's small, simple and easy to setup. It runs over redis so we also need redis.

DB structure

The main objects of the project are users and photos. We implement different classes to represent each type of object in the db in the folder named models.

In addition one can find the representation for:

  • Albums
  • Hashtags
  • liked_posts
  • Following/Follower relations
  • Tokens
  • Users
  • Photos

API endpoints

The api endpoints are available under the api. Different subfolders represent different versions of the API.

Users

In this file we implement all the method relative to actions for users.

  • Authenticate an user.
  • Get the followers for an user.
  • Get the different statuses of an user.
  • Logout the current user.
  • Get basic information of the current user.

Statuses

Security

To keep users passwords safe we only keep a hashed version of the passwords using the Aragon2 algorithm by a third-party library implementing it for python.

To give access to different apps we use the token auth method. Over the falcon library we use falcon_auth that implements several auth methods and we only have to track the validity of the tokens and not about the logic to allow or not access to different api endpoints.

Some pipelines

Upload images

When one starts the process to upload an image the steps are the following:

  • First we call /api/v1/statuses with a POST http request following the info in the API documentation. (main.py)
  • An object representating the image is created in the db assigning a unique filename to the image uploaded. (api/v1/photos.py)
  • Then the request to store the image is send to the corresponding queue and a 200 HTTP response is created. (tasks/tasks.py)
  • When the queue is ready it dispatches the process and stores the image