Goa takes a different approach to building services by making it possible to describe the design of the service API using a simple Go DSL. Goa uses the description to generate specialized service helper code, client code and documentation. Goa is extensible via plugins, for example the goakit plugin generates code that leverage the Go kit library.
The service design describes the transport independent layer of the services in the form of simple methods that accept a context and a payload and return a result and an error. The design also describes how the payloads, results and errors are serialized in the transport (HTTP or gRPC). For example a service method payload may be built from an HTTP request by extracting values from the request path, headers and body. This clean separation of layers makes it possible to expose the same service using multiple transports. It also promotes good design where the service business logic concerns are expressed and implemented separately from the transport logic.
The Goa DSL consists of Go functions so that it may be extended easily to avoid repetition and promote standards. The design code itself can easily be shared across multiple services by simply importing the corresponding Go package again promoting reuse and standardization across services.
The Goa tool accepts the Go design package import path as input and produces the interface as well as the glue that binds the service and client code with the underlying transport. The code is specific to the API so that for example there is no need to cast or "bind" any data structure prior to using the request payload or response result. The design may define validations in which case the generated code takes care of validating the incoming request payload prior to invoking the service method on the server, and validating the response prior to invoking the client code.
go install goa.design/goa/v3/cmd/goa@v3
Current Release: v3.18.2
Create a new Goa project:
mkdir -p calcsvc/design cd calcsvc go mod init calcsvc
Create the file design.go
in the design
directory with the following
content:
package design import . "goa.design/goa/v3/dsl" // API describes the global properties of the API server. var _ = API("calc", func() { Title("Calculator Service") Description("HTTP service for multiplying numbers, a goa teaser") Server("calc", func() { Host("localhost", func() { URI("http://localhost:8088") }) }) }) // Service describes a service var _ = Service("calc", func() { Description("The calc service performs operations on numbers") // Method describes a service method (endpoint) Method("multiply", func() { // Payload describes the method payload // Here the payload is an object that consists of two fields Payload(func() { // Attribute describes an object field Attribute("a", Int, "Left operand") Attribute("b", Int, "Right operand") // Both attributes must be provided when invoking "multiply" Required("a", "b") }) // Result describes the method result // Here the result is a simple integer value Result(Int) // HTTP describes the HTTP transport mapping HTTP(func() { // Requests to the service consist of HTTP GET requests // The payload fields are encoded as path parameters GET("/multiply/{a}/{b}") // Responses use a "200 OK" HTTP status // The result is encoded in the response body Response(StatusOK) }) }) })
This file contains the design for a calc
service which accepts HTTP GET
requests to /multiply/{a}/{b}
where {a}
and {b}
are placeholders for integer
values. The API returns the product of a
multiplied by b
in the HTTP response body.
Now that the design is done, let's run goa
on the design package.
In the calcsvc
directory run:
goa gen calcsvc/design
This produces a gen
directory with the following directory structure:
gen ├── calc │ ├── client.go │ ├── endpoints.go │ └─ ─ service.go └── http ├── calc │ ├── client │ │ ├── cli.go │ │ ├── client.go │ │ ├── encode_decode.go │ │ ├── paths.go │ │ └── types.go │ └── server │ ├── encode_decode.go │ ├── paths.go │ ├── server.go │ └── types.go ├── cli │ └── calc │ └── cli.go ├── openapi.json └── openapi.yaml 7 directories, 15 files
calc
contains the service endpoints and interface as well as a service
client.http
contains the HTTP transport layer. This layer maps the service
endpoints to HTTP handlers server side and HTTP client methods client side.
The http
directory also contains a complete
OpenAPI 3.0
spec for the service.The goa
tool can also generate example implementations for both the service
and client. These examples provide a good starting point:
goa example calcsvc/design calc.go cmd/calc-cli/http.go cmd/calc-cli/main.go cmd/calc/http.go cmd/calc/main.go
The tool generated the main
functions for two commands: one that runs the
server and one the client. The tool also generated a dummy service
implementation that prints a log message. Again note that the example
command
is intended to generate just that: an example, in particular it is not
intended to be re-run each time the design changes (as opposed to the gen
command which should be re-run each time the design changes).
Let's implement our service by providing a proper implementation for the multiply
method. Goa generated a payload struct for the multiply
method that contains both
fields. Goa also generated the transport layer that takes care of decoding the
request so all we have to do is to perform the actual multiplication. Edit the file
calc.go
and change the code of the multiply
function as follows:
// Multiply returns the multiplied value of attributes a and b of p. func (s *calcsrvc) Multiply(ctx context.Context, p *calc.MultiplyPayload) (res int, err error) { return p.A * p.B, nil }
That's it! we have now a full-fledged HTTP service with a corresponding OpenAPI specification and a client tool.
Now let's compile and run the service:
cd cmd/calc go build ./calc [calcapi] 16:10:47 HTTP "Multiply" mounted on GET /multiply/{a}/{b} [calcapi] 16:10:47 HTTP server listening on "localhost:8088"
Open a new console and compile the generated CLI tool:
cd calcsvc/cmd/calc-cli go build
and run it:
./calc-cli calc multiply -a 2 -b 3 6
The tool includes contextual help:
./calc-cli --help
Help is also available on each command:
./calc-cli calc multiply --help
Now let's see how robust our code is and try to use non integer values:
./calc-cli calc multiply -a 1 -b foo invalid value for b, must be INT run './calccli --help' for detailed usage.
The generated code validates the command line arguments against the types defined in the design. The server also validates the types when decoding incoming requests so that your code only has to deal with the business logic.
The service now returns an integer, but most OpenAPI services expect JSON. Lets fix that now!
In design.go
, change Result(Int)
so it reads like this:
Result(func() { Attribute("result", Int) Required("result") })
Inside of calc.go
, replace the func
block:
func (s *calcsrvc) Multiply(ctx context.Context, p *calc.MultiplyPayload) (res *calc.MultiplyResult, err error) { return &calc.MultiplyResult{Result: p.A * p.B}, nil }
Finally rebuild the app by running the build parts again:
goa gen calcsvc/design cd cmd/calc go build ./calc
You can now test and verify that your service is returning JSON:
curl -X 'GET' 'http://localhost:8088/multiply/10/10' -H 'accept: application/json' | jq .
If all goes well, you should see:
{ "result": 100 }
The http
directory contains OpenAPI 2.0 and 3.0 specifications in both YAML
and JSON format.
The specification can easily be served from the service itself using a file
server. The Files DSL
function makes it possible to serve a static file. Edit the file
design/design.go
and add:
var _ = Service("openapi", func() { // Serve the file gen/http/openapi3.json for requests sent to // /openapi.json. The HTTP file system is created below. Files("/openapi.json", "openapi3.json") })
Re-run goa gen calcsvc/design
and note the new directory gen/openapi
and
gen/http/openapi
which contain the implementation for a HTTP handler that
serves the openapi.json
file.
All we need to do is mount the handler on the service mux. Add the corresponding
import statement to cmd/calc/http.go
:
import openapisvr "calcsvc/gen/http/openapi/server"
and mount the handler by adding the following line in the same file and after
the mux creation (e.g. one the line after the // Configure the mux.
comment):
svr := openapisvr.New(nil, mux, dec, enc, nil, nil, http.Dir("../../gen/http")) openapisvr.Mount(mux, svr)
That's it! we now have a self-documenting service. Stop the running service
with CTRL-C. Rebuild and re-run it then make requests to the newly added
/openapi.json
endpoint:
^C[calcapi] 16:17:37 exiting (interrupt) [calcapi] 16:17:37 shutting down HTTP server at "localhost:8088" [calcapi] 16:17:37 exited go build ./calc
In a different console:
curl localhost:8088/openapi.json {"openapi":"3.0.3","info":{"title":"Calculator Service","description":...
The goa.design website provides a high level overview of Goa and the DSL.
In particular the page Implementing a Goa Service explains how to leverage the generated code to implement an HTTP or gRPC service.
The DSL Go Doc contains a fully documented reference of all the DSL functions.
The clue project provides observability packages that work in tandem with Goa. The packages
一键生成PPT和Word,让学习生活更轻松
讯飞智文是一个利用 AI 技术的项目,能够帮助用户生成 PPT 以及各类文档。无论是商业领域的市 场分析报告、年度目标制定,还是学生群体的职业生涯规划、实习避坑指南,亦或是活动策划、旅游攻略等内容,它都能提供支持,帮助用户精准表达,轻松呈现各种信息。
深度推理能力全新升级,全面对标OpenAI o1
科大讯飞的星火大模型,支持语言理解、知识问答和文本创作等多功能,适用于多种文件和业务场景,提升办公和日常生活的效率。讯飞星火是一个提供丰富智能服务的平台,涵盖科技资讯、图像创作、写作辅助、编程解答、科研文献解读等功能,能为不同需求的用户提供便捷高效的帮助,助力用户轻松获取信息、解决问题,满足多样化使用场景。
一种基于大语言模型的高效单流解耦语音令牌文本到语音合成模型
Spark-TTS 是一个基于 PyTorch 的开源文本到语音合成项目,由多个知名机构联合参与。该项目提供了高效的 LLM(大语言模型)驱动的语音合成方案,支持语音克隆和语音创建功能,可通过命令行界面(CLI)和 Web UI 两种方式使用。用户可以根据需求调整语音的性别、音高、速度等参数,生成高质量的语音。该项目适用于多种场景,如有声读物制作、智能语音助手开发等。
字节跳动发布的AI编程神器IDE
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。
AI助力,做PPT更简单!
咔片是一款轻量化在线演示设计工具,借助 AI 技术,实现从内容生成到智能设计的一站式 PPT 制作服务。支持多种文档格式导入生成 PPT,提供海量模板、智能美化、素材替换等功能,适用于销售、教师、学生等各类人群,能高效制作出高品质 PPT,满足不同场景演示需求。
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。
专业的AI公文写作平台,公文写作神器
AI 材料星,专业的 AI 公文写作辅助平台,为体制内工作人员提供高效的公文写作解决方案。拥有海量公文文库、9 大核心 AI 功能,支持 30 + 文稿类型生成,助力快速完成领导讲话、工作总结、述职报告等材料,提升办公效率,是体制打工人的得力写作神器。
OpenAI Agents SDK,助力开发者便捷使用 OpenAI 相关功能。
openai-agents-python 是 OpenAI 推出的一款强大 Python SDK, 它为开发者提供了与 OpenAI 模型交互的高效工具,支持工具调用、结果处理、追踪等功能,涵盖多种应用场景,如研究助手、财务研究等,能显著提升开发效率,让开发者更轻松地利用 OpenAI 的技术优势。
高分辨率纹理 3D 资产生成
Hunyuan3D-2 是腾讯开发的用于 3D 资产生成的强大工具,支持从文本描述、单张图片或多视角图片生成 3D 模型,具备快速形状生成能力,可生成带纹理的高质量 3D 模型,适用于多个领域,为 3D 创作提供了高效解决方案。
一个具备存储、管理和客户端操作等多种功能的分布式文件系统相关项目。
3FS 是一个功能强大的分布式文件系统项目,涵盖了存储引擎、元数据管理、客户端工具等多个模块。它支持多种文件操作,如创建文件和目录、设置布局等,同时具备高效的事件循环、节点选择和协程池管理等特性。适用于需要大规模数据存储和管理的场景,能够提高系统的性能和可靠性,是分布式存储领域的优质解决方案。
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