Introduction

Although we can get the symbolic representation of the Hamiltonians by our unitcell-description framework, there still remains a long way to implement concrete algorithms such as TBA, ED, etc. Despite the quite different technical details, algorithms shares common functionalities to be furnished with:

• provide tasks to be conducted with controlling parameters;
• record the results of some tasks for later use or analysis;
• update some parameters of the QuantumLattices to reconduct tasks;
• keep logs during code executions for debug;
• cache intermediate data to improve efficiency;
• ...

We thus provide a set of generic interfaces to resolve these problems, basically in the so called Engine-App mode. Specifically, algorithms are treated as Engine and tasks as App. Engine deals with the cores of algorithms along with file management, parameter updating and data caching, while App decides the concrete tasks to be conducted and provides hyper controlling parameters.