Awesome GitHub RepositoriesObject Persistence

Instantiator is a PHP library designed to create class instances without invoking their constructors. It uses the PHP Reflection API to allocate objects in memory and initialize them in a predefined state, bypassing standard constructor logic. The project functions as a data hydration tool and a testing helper. It enables the population of PHP objects from external data sources by mapping values directly to properties, regardless of whether those properties are public, protected, or private. This capability allows for the generation of objects in specific internal states for unit tests withou

Doctrine ORM is a PHP object-relational mapper that connects application objects to relational database tables. It uses the data mapper and identity map patterns to decouple the in-memory object model from the database schema, allowing developers to manage data persistence without writing manual SQL. The project features a dedicated object-oriented query language and programmatic builder for retrieving data based on entities rather than tables. It implements a unit-of-work system to track object changes during a request and synchronize them via atomic transactions. The capability surface inc

Torch7 is a scientific computing environment and tensor computation library used for deep learning research and numerical analysis. It functions as a Lua-based framework for training neural networks and learning agents, providing a toolkit for implementing architectures and training through reinforcement learning algorithms. The project is distinguished by its tight integration with C, utilizing a binding layer to map high-level scripting to low-level C structures for direct memory access. It supports hardware-accelerated computation by offloading linear algebra and convolution operations to

MJExtension is a JSON serialization library and model mapping framework used to convert data between JSON strings and structured model objects. It functions as an object data mapper that handles the encoding and decoding of complex object hierarchies for network transmission and storage. The framework is a non-intrusive data mapper that uses reflection and runtime inspection to map raw data strings to application objects. This approach allows for data transformation without requiring base class inheritance, decorators, or extensions to the underlying model classes. The system supports recurs

This project is a Go library and runtime for loading and managing eBPF programs and maps. It provides a bytecode loader and kernel interface to inject instructions into kernel hooks for system-level execution and observability across both Linux and Windows operating systems. The library features a relocation engine and tooling to ensure program compatibility across different kernel versions and distributions. It supports portable deployment by embedding compiled objects for multiple CPU architectures into a single binary and provides the ability to load signed system drivers on Windows. The

Flashlight is a standalone C++ machine learning library and tensor library used for building and training neural networks. It functions as a comprehensive neural network framework and automatic differentiation engine, providing the tools to construct computation graphs and calculate gradients via backpropagation. The project serves as a distributed training framework, utilizing all-reduce operations to synchronize gradients and parameters across multiple compute nodes and devices. It distinguishes itself through deep integration of high-performance tensor manipulation, native device memory in

Flashlight is a C++ machine learning library and deep learning framework designed for building and training neural networks. It functions as a tensor manipulation library and an automatic differentiation engine that tracks operations to calculate gradients via backpropagation for model optimization. The project is distinguished by its role as a distributed training framework, utilizing all-reduce gradient synchronization and distributed environments to scale machine learning workloads across multiple nodes and devices. It features a backend-agnostic memory interface and RAII-based management

protobuf-net 是一个二进制序列化框架和 .NET 库，实现了 Protocol Buffers 规范。它作为一个基于模式的序列化工具和契约优先的数据映射器，将复杂的对象图转换为紧凑的二进制格式，用于存储和网络传输。 该库通过减小有效载荷大小和处理时间，实现了跨平台数据交换和高性能网络通信。它特别支持对象层次结构持久化，允许序列化和重构包含继承和派生类型的复杂类结构。 该项目涵盖了核心功能，包括二进制数据序列化和反序列化、模式驱动的类生成，以及通过唯一标识符管理类型继承。它提供了通过静态属性或动态运行时配置映射数据结构的机制。

Urho3D is a cross-platform 3D game engine written in C++. It uses a component-based scene graph to compose game objects from nodes and attached components, separating transforms from behaviors for modular design. The engine integrates AngelScript and Lua scripting for game logic, uses the Bullet library for physics simulation, and renders scenes with OpenGL or Direct3D through forward, deferred, or light pre-pass pipelines with customizable render passes and shadow mapping. The engine distinguishes itself with a built-in visual scene and UI editor for composing 3D worlds and interface layouts

ActiveAndroid is an object-relational mapper and data persistence library for Android. It serves as a high-level wrapper for SQLite databases, translating database records into class instances to facilitate data manipulation. The library enables data persistence without the need to write manual SQL queries. It maps database rows to Java objects, allowing records to be saved, retrieved, and deleted using object-oriented methods. The toolkit covers local mobile data storage and database management, providing a layer that links database table structures to specific object properties.

ObjectBox Java 是一个用于 Java 和 Android 的嵌入式 NoSQL 对象数据库，无需关系映射即可直接存储数据对象。它作为一个原生进程存储引擎，允许应用将普通 Java 或 Kotlin 类作为实体持久化。 该项目以其设备端向量数据库功能脱颖而出，利用 HNSW 索引执行近似最近邻搜索和语义相似度查询。它还包括一个本地托管的基于 Web 的浏览器，用于可视化数据对象、模式和依赖关系图。 该数据库涵盖了广泛的数据管理功能，包括符合 ACID 的原子事务、用于实时观察的响应式数据流以及跨平台数据同步。其存储范围支持具有延迟加载、基于属性的过滤的对象关系映射，以及在内存中初始化数据存储以进行缓存或自动化测试的能力。

REFramework is a runtime modding framework for games built on the RE Engine, providing a C# plugin system with typed proxy code generation, a Lua scripting interface, an ImGui-based UI system, a runtime debugging toolkit, and a VR integration layer. At its core, it intercepts and modifies internal game engine functions and state at runtime through hooking and scripting APIs, enabling deep inspection and manipulation of live game objects. The framework distinguishes itself through typed proxy code generation that auto-creates C# interfaces from the game's type database, giving compile-time saf

Joblib 是一套用于并行化计算工作负载和优化大型数值数据集及函数结果存储的实用工具。它作为并行计算库和多进程包装器，将函数执行分配到多个 CPU 核心上，以加速独立任务和计算循环。 该项目提供了一个磁盘缓存框架，将昂贵的函数输出持久化到文件系统，仅在输入参数发生变化时才重新评估。它进一步专注于大型数值数组的序列化，利用高效的压缩和内存映射来优化海量数据集的存储和检索。 该工具包包括并行函数映射功能，并使用可插拔的执行后端来控制任务如何在可用硬件上分配。其存储层涵盖了复杂对象持久化和序列化数据的透明压缩。

MongoEngine 是一个 Python 对象文档映射器（ODM），它将数据库记录转换为对象，从而为数据持久化提供面向对象的接口。它作为 MongoDB 的文档管理器和模式验证器，将类映射到文档以强制执行数据类型和验证规则。 该项目提供了一个延迟加载的 QuerySet 系统，用于使用 Pythonic 语法过滤、排序和聚合集合。它通过文档继承、递归嵌入式文档处理和基于引用的对象链接等功能管理复杂数据结构。 该库涵盖了广泛的功能，包括模式迁移、全文搜索以及通过 GridFS 文件系统管理大二进制文件。它还包括用于数据库索引优化、查询性能分析和基于信号的生命周期钩子的工具，以在文档事件期间自动化逻辑。

AdalFlow 是一个自主 AI 代理框架和 LLM 应用库，旨在构建模块化工作流。它作为一个模型无关的接口和 RAG 流水线编排器，允许用户开发 ReAct 代理，利用迭代推理和外部工具执行来解决复杂任务。 该项目通过一个提示词优化系统脱颖而出，该系统使用文本梯度下降自动优化提示词模板和少样本示例。它将模型反馈视为可微分信号，实现了一种 LLM 反向传播形式，从而根据评估指标迭代提高输出质量。 该框架涵盖了广泛的功能面，包括带有语义向量搜索和重排序的检索增强生成、用于可观测性的基于跨度的执行追踪，以及模式驱动的结构化解析。它为众多专有和开源模型提供商提供了统一的通信层，并支持将 Python 函数转换为标准化的工具接口。 该系统使用 Python 实现，并与 MLflow 集成以进行工作流跟踪和分析。

该项目是一个基于 Markdown 的个人知识库和数字学习日志，用于存储书籍和技术文献的笔记与摘要。它作为一个阅读摘要仓库和技术参考库，用于归档非虚构类和专业材料的关键要点与见解。 该集合作为一个数字花园，用于整理来自书籍、文章和视频的见解。它专门专注于将复杂的架构模式、技术概念和专业领导力原则提炼成可搜索的格式，以供长期参考。 该仓库还通过收集可复用的思维框架、心智模型和专业决策工具来组织个人知识。它包含一个用于通过管理已完成和进行中作品列表来跟踪阅读进度的系统。

这是一个对象持久化库和数据映射器抽象层。它提供了一组共享接口和基类，旨在将领域逻辑与特定的对象映射器实现解耦，从而将业务领域层与底层数据访问实现分离开来。 该库包含一个数据库模式管理和迁移工具，用于定义、版本化和部署跨不同环境的数据库结构增量更新。它还可用作文档数据库映射器，将对象状态转换为结构化文档格式以进行存储和检索。 该系统涵盖了广泛的持久化功能，包括对象集合的管理以及在不调用构造函数的情况下实例化对象以恢复状态的能力。它还提供了用于数据库模式版本化和调度自定义事件的实用程序，以允许解耦组件对持久化更改做出反应。

Factory Boy is a dynamic test fixture framework and data generation tool for Python. It serves as a replacement for static fixtures by providing a system to create complex Python objects and database models through programmable blueprints. The project differentiates itself by offering specialized integration with Object Relational Mappers to manage persistence within database sessions. It enables the creation of complex object hierarchies using sub-factories and recursive composition to resolve dependent related objects. The framework provides capabilities for synthesizing realistic random d

This project is a C++ learning resource and study guide consisting of structured notes and programming examples. It provides practical implementations and exercise solutions covering core language syntax, data types, and control flow. The repository features specialized samples for object-oriented design, including class inheritance, polymorphism, and abstract classes. It includes demonstrations of memory management techniques such as dynamic allocation, move semantics, and placement new, as well as template programming examples for creating generic functions and data structures. The codebas

Odin Serializer 是一个为 Unity 游戏引擎设计的框架，用于将复杂的对象图转换为持久化的二进制或 JSON 格式。它提供了一个用于保存和加载复杂类结构的系统，包括嵌套泛型和多态类型，同时保持数据的结构完整性。 该框架通过其对提前（AOT）编译的支持脱颖而出，在构建过程中生成必要的类型元数据，以确保在限制运行时反射的平台上功能正常。它直接与 Unity 引擎生命周期集成，允许序列化引擎特定的类型，并通过唯一的持久化标识符管理外部资产引用。 用户可以定义细粒度的序列化策略，以控制在转换过程中包含或排除哪些类成员。此功能允许通过确保仅处理所需信息来优化数据存储和传输，而底层引擎处理复杂继承层次结构的递归映射。