[Open Source] Machine Learning Zero-Code Unit Testing Platform with Comprehensive, Automated Testing
![[Open Source] Machine Learning Zero-Code Unit Testing Platform with Comprehensive, Automated Testing](https://yy123.ink/api/img/1782591336385-ho56cl.jpg)
1. Project Overview

UnitAuto
☀️ Machine Learning Zero-Code Unit Testing Platform
Zero-code, comprehensive, and automated testing of the correctness, availability, and performance of methods/functions.
2. Open Source License
Released under the Apache-2.0 open-source license.
3. Screenshots



4. Feature Overview
A machine learning unit testing platform that provides zero-code, comprehensive, and automated testing of the correctness, availability, and performance of methods/functions.
Currently used by multiple projects across several departments in Tencent's IEG (Interactive Entertainment Group) and WXG (Weixin Group), as well as Kuaishou Advertising.
Included in the National Engineering Laboratory for Internet Education Intelligent Technology and Applications.
Key Features and Advantages
Compared with various compiling testing tools such as JUnit, JTest, Mockito, and Mockk:
1. Other tools require developers to write extensive test code for every method, which increases development costs. They also require fixing bugs in test code and synchronizing test code changes whenever business code is modified.
UnitAuto requires no code writing. It directly reads method properties, automatically injects parameters, retrieves return values and class member variables, and uses machine learning for automated validation.
2. As a runtime testing tool, UnitAuto does not require a mocked environment (Application, Context, etc.).
You no longer need to worry about difficulties such as being unable to effectively mock environment-related classes or third-party login services that do not provide mock support.
A method can be tested as long as it meets at least one of these three conditions: has a return value, has an interface callback, or modifies a member variable (field).
5. Tech Stack
☀️ A machine learning zero-code unit testing platform that provides comprehensive, automated testing of method/function correctness, availability, and performance, developed by Tencent engineers.
How It Works
The project under test does not require any unit testing code (no logic code, annotation code, or anything else).
UnitAuto automatically generates test parameters, executes methods, retrieves return values, and performs validation.
Generics and interfaces are automatically simulated. Asynchronous method execution automatically records callback processes. Everything requires zero code.
UnitAuto-Admin is a machine learning zero-code testing tool similar to APIAuto (for viewing, uploading, executing, and testing methods in backend projects).
UnitAuto-Java provides two APIs: scanning all methods and executing a specific method (primarily by importing the MethodUtil.java class, which already implements these functions).