Hiku is a library to design Graph APIs. Why graphs? – They are simple, predictable, flexible, easy to compose and because of that, they are easy to reuse.

Documentation: https://hiku.readthedocs.io/

Licensed under BSD-3-Clause license. See LICENSE.txt

$ pip install hiku

• Not coupled to one specific query language
• Flexibility in result serialization, including binary format
• All concurrency models supported: async/await, threads, greenlets
• Parallel execution of the query itself for free
• No data under-fetching or over-fetching
• No extra data loading from databases, only what was needed to fulfill the query
• No N+1 problems, they are eliminated by design
• Even complex queries of any size are predictable in terms of performance impact
• Implements a concept of the Two-Level Graph in order to put your business-logic in the right place

Please feel free to open an issue if you are getting mysterious error messages. It will be much easier to track down bugs, if you will provide short graph definition to reproduce it. Pull requests are highly appreciated and awaited. Especially when it comes to a spelling errors in documentation.

Use Tox in order to test and lint your changes.