This article explores how AI-driven front‑end code generation—from design‑to‑code pipelines, product analyses, technical challenges, and layered recognition models—to improve developer efficiency, detailing the D2C system, its real‑world deployment during 2019 Double‑11, and future roadmap.
This tutorial explains how to install the EasyCode plugin, set up a MySQL database, configure IDEA, generate entity, controller, service, DAO, and mapper code for a Spring Boot project, and adjust Maven and YAML settings to run the application successfully.
This tutorial explains how Java developers can use IntelliJ IDEA Live Templates to automatically generate repetitive code such as private fields, loggers, and Spring beans, covering basic usage, custom templates, advanced functions, and practical examples that dramatically improve coding efficiency.
This article recounts how a developer transformed a massive three‑thousand‑line C# data‑center monitoring client into a concise fifteen‑line solution by applying reflection, generics, and systematic refactoring, while sharing practical advice on avoiding code generators, reinventing the wheel, and using unit tests to foster clean architecture.
This article explains how Alibaba's Frontend Intelligent Project automatically generated 79.34% of the Double‑11 page code by recognizing business modules from visual drafts using data‑augmented samples, traditional multi‑class machine‑learning models, and a pipeline of preprocessing, model training, deployment, and OOD handling.
The article introduces the concept of an App Factory—a code‑generation platform that, based on a unified component library and dependency graph, can produce multiple iOS applications on demand while eliminating unused code, and details its architecture, implementation methods, quality controls, and real‑world results within 58.com’s ecosystem.
This article explains how Alibaba's Frontend Intelligent project uses design‑to‑code techniques, component recognition with YOLO, and a full pipeline of sample creation, model training, evaluation, and prediction refinement to automatically generate a large portion of the Double‑11 event code.
This article explains how Alibaba's Front‑End Intelligent project automatically converts UI design images into production‑ready code, covering layout analysis, background and foreground processing, a fusion of traditional image algorithms with deep‑learning detection, GAN‑based complex‑background extraction, experimental results and real‑world deployment.
This article explains Alibaba's front‑end intelligent project that automatically generated 79.34% of the Double‑11 UI code, detailing why images are used as input, the layered image‑processing pipeline, background and foreground analysis, traditional versus deep‑learning methods, fusion techniques, evaluation results, and real‑world deployments.
This article explores the rise of front‑end intelligence, analyzing background trends, competitive solutions, problem decomposition, and technical approaches for automatically converting design drafts into HTML, CSS, and JavaScript, while discussing challenges, model accuracy, data quality, and future roadmap for the D2C (Design‑to‑Code) system.
This article explores six practical ways to accelerate Java development—from hand‑writing and copy‑pasting to Excel formulas, code‑generation tools, and custom generators—detailing their workflows, sample code, advantages, drawbacks, and how to choose the right method for any project.
This article explains the concept of PSI (Program Structure Interface) in IntelliJ IDEA, demonstrates how to obtain PsiFile objects, manipulate PsiJavaFile, PsiClass, PsiField, and add annotations and parameters, providing a practical guide for plugin developers.
Metaprogramming in Dart leverages the Dill intermediate representation and its abstract syntax tree to programmatically generate, transform, or inject code at compile time, enabling features such as custom serialization, AOP hooks, and reflection‑like capabilities without runtime overhead.
This article explains how to parse a Swagger JSON document and automatically generate maintainable Java test code for each API endpoint, detailing the handling of GET query parameters, POST form‑data, RESTful path variables, and header conventions.
This article explains the inner workings of Tapable—the hook library behind Webpack—by detailing its synchronous and asynchronous hook types, registration and invocation methods, the lazy‑compilation code‑generation process, and how these mechanisms give Webpack a performance edge over naïve event loops.
This article explains the concept and practical application of the Annotation Processing Tool (APT) in Android development, demonstrating how to set up an APT project, define annotations, process them with custom processors, and automatically generate Java code for activity navigation using JavaPoet.
AspectD is a compile‑time Aspect‑Oriented Programming framework for Dart that traverses Flutter’s dill intermediate files to inject, call, or execute advice code, allowing non‑intrusive instrumentation, logging, and UI event injection without modifying source, and integrates into the build pipeline via a toolchain patch.
This article explains why Lombok was created to eliminate repetitive Java code, shows how to add the Maven dependency, configure IDE support, describes its annotation‑processing mechanism, and demonstrates the most common Lombok annotations with before‑and‑after code examples.
The UI2CODE pipeline converts image‑derived JSON DSL layouts into Flutter code by constructing a hierarchical DSL tree, partitioning it into rows, columns or stacks, applying unsupervised text‑alignment and padding heuristics, generating templated widget code with separate data models, cutting development effort by roughly two‑thirds and now used internally at Xianyu.
To meet Xianyu’s rapid UI change demands and curb its swelling Android package, the team built a Flutter‑specific dynamic template system that compiles Dart widget code into JSON, streams it to the client, reconstructs the widget tree at runtime, and maintains 55‑60 fps performance while planning further widget support and caching.
UI2CODE, an AI‑powered tool from Xianyu’s tech team, automatically transforms UI screenshots into editable Flutter code by extracting visual elements with machine‑vision, classifying them via deep‑learning, generating a domain‑specific language through a recursive neural network, and mapping this DSL onto syntax‑tree templates to achieve pixel‑level precision, full accuracy, and maintainable output.
This article demystifies how programming languages are transformed by compilers, covering the roles of lexical analysis, parsing, abstract syntax trees, code generation, and linking, with practical Rust examples, diagrams, and references to deepen your understanding of language implementation.
This article explains how to replace cumbersome if‑else or manual map‑based routing in Flutter with a lightweight, annotation‑driven solution that automatically generates a page‑to‑URL mapping at compile time, using Dart's source_gen, build, and analyser packages.
The UI2CODE framework from Alibaba's Xianyu team uses machine‑vision and deep‑learning techniques to automatically interpret design screenshots and generate ready‑to‑use Android, iOS, and web code, dramatically cutting UI development effort.
This guide shows how to use a PHP class that randomly creates realistic Chinese surnames and given names, explains its methods, and provides a ready‑to‑run example for backend developers.
ui‑automation, developed by Alibaba’s Xianyu team, automatically converts Sketch files or PNG images into clean, maintainable UI code for platforms such as Flutter or Weex by extracting visual information, building a hierarchical DSL, and applying template generation, delivering senior‑engineer‑level quality while overcoming the maintainability limits of earlier scanning‑based approaches.
This article provides a detailed tutorial for creating a Spring Boot project with MyBatis, including environment setup, Maven dependencies, configuration files, database creation, MyBatis generator usage, and sample controller, service, and mapper code, enabling pagination with PageHelper and testing via HTTP requests.
This article introduces Lombok—a Java library that eliminates boilerplate code—by showing how to install it in IntelliJ, demonstrating the most common annotations with runnable examples, and revealing the annotation‑processing mechanism that generates the bytecode during compilation.
This article introduces the React‑Sketchapp tool for generating Sketch design assets from code, explains why code‑based design generation is beneficial, provides step‑by‑step setup commands, showcases example components and API usage, and reflects on its impact on design‑development workflows.
By walking through the creation of a custom Android Studio (IntelliJ) plugin that reads protobuf files and auto‑generates Service and Presenter classes, the author shows how VFS, PSI, and project‑model APIs can replace tedious manual coding, delivering roughly 5.5% time savings per requirement.
This article describes the pain points of manual API test script writing, proposes a workflow that parses interface documentation, automatically generates both interface classes and robust test cases using Python and Swagger, and demonstrates significant efficiency gains and future improvement directions.
This article demonstrates how to use Google's ZXing Core library in a Java Maven project to generate QR code images, create random verification codes, and decode QR codes from files, providing complete code examples and step-by-step instructions.
This article explains why a template engine is needed for web applications that mix large static HTML with dynamic data, describes the supported template syntax based on Django, and walks through the implementation of a Python‑based engine that parses templates, compiles them into Python code, and renders the final HTML efficiently.
This article introduces Lombok as a Java code‑simplification tool, explains how to install the IDEA plugin and add the Maven dependency, demonstrates basic and advanced annotation usage, discusses domain‑driven design models, and shows how Lombok can be combined with FreeMarker and MyBatis for automatic code generation.
Recent advances in deep learning have enabled machines to automatically generate code, with research divided into black‑box methods that learn input‑output transformations and code‑generation approaches that produce explicit program fragments, exemplified by systems such as Neural Program Interpreters and hierarchical generative CNN models.
The article presents a serverless‑style solution that replaces a heavy backend with a lightweight, automatically generated CRUD layer, detailing RESTful API conventions, extensions for batch operations, query parameters, foreign‑key handling, and a Rails‑based code‑generation pipeline to streamline development of a high‑availability supply‑chain system.
The article introduces Ctrip DAL, an open‑source Java/C# database access framework that supports sharding across MySQL and SQL Server, provides a code generator and client library, and helps large enterprises manage multi‑team, multi‑database environments while ensuring scalability, consistency, and security.
Ctrip DAL is an open‑source, multi‑language database access framework that supports sharding, code generation, and cross‑team management, enabling large‑scale enterprises to protect existing investments while rapidly scaling their data layer across MySQL and SQL Server.
The article recounts how the author transformed a massive, repetitive three‑tier factory class in a C# data‑center management system into a clean, reflection‑based solution of just a few lines, while sharing practical refactoring lessons, pitfalls of code generators, and advice on unit‑testing and architecture design.
