I'm working on Quickiebase my pure-Python NoSQL database which stores documents as JSON objects.
Why Quickiebase?
Python is renowned for being batteries-included. It also has access to sqlite as part of the standard library. I think it would make sense to have an NoSQL document database as part of the standard library; something that can you can set up and start using with a simple import statement.
(If not as part of the standard library, then it should be available as a package on the Python Package Index.)
According to the 2022 Stack Overflow developer survey, MongoDB is the most popular NoSQL database (and the 4th most popular database overall), and the most wanted NoSQL database which 17% of developers wanting to use it.
Because of this, the NoSQL database library should be based on MongoDB, that is its functionality should be a subset of MongoDB's, so that if a project using it grows to something big it can easily move up to a bigger database.
I'm also mindful that MongoDB stopped being open source with its Server Side Public License (SSPL).
Top comments (1)
I did this already. I modeled it after tinydb, which is also one. JSON is not efficient in data saving, lots of repeat data like semicolons. I had one that is more database-esque or how I imagine mysql to work internally with read/writes to a binary file.