Hexagon is a microservices' toolkit (not a framework) written in Kotlin. Its purpose is to ease the building of server applications (Web applications, APIs or queue consumers) that run inside a cloud platform.
The Hexagon Toolkit provides several libraries to build server applications. These libraries provide single standalone features and are referred to as "Ports".
The main ports are:
Each of these features or ports may have different implementations called "Adapters".
Hexagon is designed to fit in applications that conform to the Hexagonal Architecture (also called Clean Architecture or Ports and Adapters Architecture). Also, its design principles also fits in this architecture.
The Hexagon's goals and design principles are:
Put you in Charge: There is no code generation, no runtime annotation processing, no classpath based logic, and no implicit behaviour. You control your tools, not the other way around.
Modular: Each feature (Port) or adapter is isolated in its own module. Use only the modules you need without carrying unneeded dependencies.
Pluggable Adapters: Every Port may have many implementations (Adapters) using different technologies. You can swap adapters without changing the application code.
Batteries Included: It contains all the required pieces to make production-grade applications: logging utilities, serialization, resource handling and build helpers.
Kotlin First: Take full advantage of Kotlin instead of just calling Java code from Kotlin. The library is coded in Kotlin for coding with Kotlin. No strings attached to Java (as a Language).
Properly Tested: The project's coverage is checked in every Pull Request. It is also stress-tested at TechEmpower Frameworks Benchmark.
For more information check the Quick Start Guide.
You can clone a starter project (Gradle Starter or Maven Starter). Or you can create a project from scratch following these steps:
In Gradle. Import it inside build.gradle:
repositories { mavenCentral() } implementation("com.hexagonkt:http_server_jetty:$hexagonVersion")
In Maven. Declare the dependency in pom.xml:
<dependency> <groupId>com.hexagonkt</groupId> <artifactId>http_server_jetty</artifactId> <version>$hexagonVersion</version> </dependency>
src/main/kotlin/Hello.kt file:// hello_world import com.hexagonkt.core.media.TEXT_PLAIN import com.hexagonkt.http.model.ContentType import com.hexagonkt.http.server.HttpServer import com.hexagonkt.http.server.jetty.serve lateinit var server: HttpServer /** * Start a Hello World server, serving at path "/hello". */ fun main() { server = serve { get("/hello/{name}") { val name = pathParameters["name"] ok("Hello $name!", contentType = ContentType(TEXT_PLAIN)) } } } // hello_world
A simple CRUD example showing how to manage book resources. Here you can check the full test.
</details> <details> <summary>Error Handling Example</summary>// books data class Book(val author: String, val title: String) private val books: MutableMap<Int, Book> = linkedMapOf( 100 to Book("Miguel de Cervantes", "Don Quixote"), 101 to Book("William Shakespeare", "Hamlet"), 102 to Book("Homer", "The Odyssey") ) private val path: PathHandler = path { post("/books") { val author = queryParameters["author"]?.string() ?: return@post badRequest("Missing author") val title = queryParameters["title"]?.string() ?: return@post badRequest("Missing title") val id = (books.keys.maxOrNull() ?: 0) + 1 books += id to Book(author, title) created(id.toString()) } get("/books/{id}") { val bookId = pathParameters.require("id").toInt() val book = books[bookId] if (book != null) ok("Title: ${book.title}, Author: ${book.author}") else notFound("Book not found") } put("/books/{id}") { val bookId = pathParameters.require("id").toInt() val book = books[bookId] if (book != null) { books += bookId to book.copy( author = queryParameters["author"]?.string() ?: book.author, title = queryParameters["title"]?.string() ?: book.title ) ok("Book with id '$bookId' updated") } else { notFound("Book not found") } } delete("/books/{id}") { val bookId = pathParameters.require("id").toInt() val book = books[bookId] books -= bookId if (book != null) ok("Book with id '$bookId' deleted") else notFound("Book not found") } // Matches path's requests with *any* HTTP method as a fallback (return 405 instead 404) after(ALL - DELETE - PUT - GET, "/books/{id}", status = NOT_FOUND_404) { send(METHOD_NOT_ALLOWED_405) } get("/books") { ok(books.keys.joinToString(" ", transform = Int::toString)) } } // books
Code to show how to handle callback exceptions and HTTP error codes. Here you can check the full test.
</details> <details> <summary>Filters Example</summary>// errors class CustomException : IllegalArgumentException() private val path: PathHandler = path { /* * Catching `Exception` handles any unhandled exception, has to be the last executed (first * declared) */ exception<Exception> { internalServerError("Root handler") } exception<IllegalArgumentException> { val error = exception?.message ?: exception?.javaClass?.name ?: fail val newHeaders = response.headers + Header("runtime-error", error) send(HttpStatus(598), "Runtime", headers = newHeaders) } exception<UnsupportedOperationException> { val error = exception?.message ?: exception?.javaClass?.name ?: fail val newHeaders = response.headers + Header("error", error) send(HttpStatus(599), "Unsupported", headers = newHeaders) } get("/exception") { throw UnsupportedOperationException("error message") } get("/baseException") { throw CustomException() } get("/unhandledException") { error("error message") } get("/invalidBody") { ok(LocalDateTime.now()) } get("/halt") { internalServerError("halted") } get("/588") { send(HttpStatus(588)) } // It is possible to execute a handler upon a given status code before returning before(pattern = "*", status = HttpStatus(588)) { send(HttpStatus(578), "588 -> 578") } } // errors
This example shows how to add filters before and after route execution. Here you can check the full test.
</details> <details> <summary>Files Example</summary>// filters private val users: Map<String, String> = mapOf( "Turing" to "London", "Dijkstra" to "Rotterdam" ) private val path: PathHandler = path { filter("*") { val start = System.nanoTime() // Call next and store result to chain it val next = next() val time = (System.nanoTime() - start).toString() // Copies result from chain with the extra data next.send(headers = response.headers + Header("time", time)) } filter("/protected/*") { val authorization = request.authorization ?: return@filter unauthorized("Unauthorized") val credentials = authorization.value val userPassword = String(credentials.decodeBase64()).split(":") // Parameters set in call attributes are accessible in other filters and routes send(attributes = attributes + ("username" to userPassword[0]) + ("password" to userPassword[1]) ).next() } // All matching filters are run in order unless call is halted filter("/protected/*") { if(users[attributes["username"]] != attributes["password"]) send(FORBIDDEN_403, "Forbidden") else next() } get("/protected/hi") { ok("Hello ${attributes["username"]}!") } path("/after") { after(PUT) { send(ALREADY_REPORTED_208) } after(PUT, "/second") { send(NO_CONTENT_204) } after("/second") { send(CREATED_201) } after { send(ACCEPTED_202) } } } // filters
The following code shows how to serve resources and receive files. Here you can check the full test.
</details>// files private val path: PathHandler = path { // Serve `public` resources folder on `/*` after( methods = setOf(GET), pattern = "/*", status = NOT_FOUND_404, callback = UrlCallback(urlOf("classpath:public")) ) path("/static") { get("/files/*", UrlCallback(urlOf("classpath:assets"))) get("/resources/*", FileCallback(File(directory))) } get("/html/*", UrlCallback(urlOf("classpath:assets"))) // Serve `assets` files on `/html/*` get("/pub/*", FileCallback(File(directory))) // Serve `test` folder on `/pub/*` post("/multipart") { val headers = parts.first().let { p -> val name = p.name val bodyString = p.bodyString() val size = p.size.toString() Headers( Header("name", name), Header("body", bodyString), Header("size", size), ) } ok(headers = headers) } post("/file") { val part = parts.first() val content = part.bodyString() val submittedFile = part.submittedFileName ?: "" ok(content, headers = response.headers + Header("submitted-file", submittedFile)) } post("/form") { fun <T : HttpField> serializeMap(map: Collection<T>): List<String> = listOf( map.joinToString("\n") { "${it.name}:${it.values.joinToString(",")}" } ) val queryParams = serializeMap(queryParameters.values) val formParams = serializeMap(formParameters.values) val headers = Headers(Header("query-params", queryParams), Header("form-params", formParams)) ok(headers = response.headers + headers) } } // files
You can check more sample projects and snippets at the examples page.
This project is supported by:
<a href="https://www.digitalocean.com/?utm_medium=opensource&utm_source=Hexagon-Toolkit"> <img height="128px" src= "https://opensource.nyc3.cdn.digitaloceanspaces.com/attribution/assets/SVG/DO_Logo_vertical_blue.svg"> </a> <a href="https://www.jetbrains.com/?from=Hexagon-Toolkit"> <img height="96px" src="https://hexagontk.com/img/sponsors/jetbrains-variant-4.svg"> </a>The toolkit is properly tested. This is the coverage report:
Performance is not the primary goal, but it is taken seriously. You can check performance numbers in the TechEmpower Web Framework Benchmarks.
If you like this project and want to support it, the easiest way is to give it a star :v:.
If you feel like you can do more. You can contribute to the project in different ways:
To know what issues are currently open and be aware of the next features you can check the Organization Board at GitHub.
You can ask any question, suggestion or complaint at the project's discussions. You can be up-to-date of project's news following @hexagontk on X (Twitter).
Thanks to all project's contributors!
The project is licensed under the [MIT License]. This license lets you use the source for free or commercial purposes as long as you provide attribution and don’t hold any project member liable.
[MIT License]:


AI赋能电商视觉革命,一站式智能商拍平台
潮际好麦深耕服装行业,是国内AI试衣效果最好的软件。使用先进AIGC能力为电商卖家批量提供优质的、低成本的商拍图。合作品牌有Shein、Lazada、安踏、百丽等65个国内外头部品牌,以及国内10万+淘宝、天猫、京东等主流平台的品牌商家,为卖家节省将近85%的出图成本,提升约3倍出图效率,让品牌能够快速上架。


企业专属的AI法律顾问
iTerms是法大大集团旗下法律子品牌,基于最先进的大语言模型(LLM)、专业的法律知识库和强大的智能体架构,帮助企业扫清合规障碍,筑牢风控防线,成为您企业专属的AI法律顾问。


稳定高效的流量提升解决方案,助力品牌曝光
稳定高效的流量提升解决方案,助力品牌曝光


最新版Sora2模型免费使用,一键生成无水印视频
最新版Sora2模型免费使用,一键生成无水印视频


实时语音翻译/同声传译工具
Transly是一个多场景的AI大语言模型驱动的同声传译、专业翻译助手,它拥有超精准的音频识别翻译能力,几乎零延迟的使用体验和支持多国语言可以让你带它走遍全球,无论你是留学生、商务人士、韩剧美剧爱好者,还是出国游玩、多国会议、跨国追星等等,都可以满足你所有需要同传的场景需求,线上线下通用,扫除语言障碍,让全世界的语言交流不再有国界。


选题、配图、成文,一站式创作,让内容运营更高效
讯飞绘文,一个AI集成平台,支持写作、选题、配图、排版和发布。高效生成适用于各类媒体的定制内容,加速品牌传播,提升内容营销效果。


AI辅助编程,代码自动修复
Trae是一种自适应的集成开发环境(IDE),通过自动化和多元协作改变开发流程。利用Trae,团队能够更快速、精确地编写和部署代码,从而提高编程效率和项目交付速度。Trae具备上下文感知和代码自动完成功能,是提升开发效率的理想工具。


最强AI数据分析助手
小浣熊家族Raccoon,您的AI智能助手,致力于通过先进的人工智能技术,为用户提供高效、便捷的智能服务。无论是日常咨询还是专业问题解答,小浣熊都能以快速、准确的响应满足您的需求,让您的生活更加智能便捷。


像人一样思考的AI智能体
imini 是一款超级AI智能体,能根据人类指令,自主思考、自主完成、并且交付结果的AI智能体。


AI数字人视频创作平台
Keevx 一款开箱即用的AI数字人视频创作平台,广泛适用于电商广告、企业培训与社媒宣传,让全球企业与个人创作者无需拍摄剪辑,就能快速生成多语言、高质量的专业视频。
最新AI工具、AI资讯
独家AI资源、AI项目落地

微信扫一扫关注公众号