Golang:使用 httprouter 构建 API 服务器

https://medium.com/@gauravsingharoy/build-your-first-api-server-with-httprouter-in-golang-732b7b01f6ab

作者:Gaurav Singha Roy

译者:oopsguy.com

10 个月前,我成为了一名 Gopher,没有后悔。像许多其他 gopher 一样,我很快发现简单的语言特性对于快速构建快速、可扩展的软件非常有用。当我刚开始学习 Go 时,我正在捣鼓不同的路由器,它可以当做 API 服务器使用。如果你跟我一样有 Rails 背景,你也可能会在构建 Web 框架提供的所有功能方面遇到困难。回到路由器话题,我发现了 3 个是非常有用的好东西: Gorilla muxhttprouterbone(按性能从低到高)。即便 bone 的性能最高且有更简单的 handler 签名,但对我来说,它仍然不够成熟,无法应用于生产环境中。因此,我最终使用了 httprouter。在本教程中,我将使用 httprouter 构建一个简单的 REST API 服务器。

如果你想偷懒,只想获取源码,则可以在这里[4]直接克隆我的 github 仓库。

让我们开始吧。首先创建一个基本端点:

package main

import (
"fmt"
"log"
"net/http" "github.com/julienschmidt/httprouter"
) func Index(w http.ResponseWriter, r *http.Request, _ httprouter.Params) {
fmt.Fprint(w, "Welcome!\n")
} func main() {
router := httprouter.New()
router.GET("/", Index) log.Fatal(http.ListenAndServe(":8080", router))
}

在上面的代码段中,Index 是一个 handler 函数,需要传入三个参数。之后,该 handler 将在 main 函数中被注册到 GET / 路径上。现在编译并运行你的程序,转到 http:// localhost:8080 来查看你的 API 服务器。点击这里[1]获取当前代码。

我们可以让 API 变得复杂一点。现在有一个名为 Book 的实体,可以把 ISDN 字段作为唯一标识。让我们创建更多的 handler,即分表代表着 Index 和 Show 动作的 GET /books 和 GET /books/:isdnmain.go 文件此时如下:

package main

import (
"encoding/json"
"fmt"
"log"
"net/http" "github.com/julienschmidt/httprouter"
) func Index(w http.ResponseWriter, r *http.Request, _ httprouter.Params) {
fmt.Fprint(w, "Welcome!\n")
} type Book struct {
// The main identifier for the Book. This will be unique.
ISDN string `json:"isdn"`
Title string `json:"title"`
Author string `json:"author"`
Pages int `json:"pages"`
} type JsonResponse struct {
// Reserved field to add some meta information to the API response
Meta interface{} `json:"meta"`
Data interface{} `json:"data"`
} type JsonErrorResponse struct {
Error *ApiError `json:"error"`
} type ApiError struct {
Status int16 `json:"status"`
Title string `json:"title"`
} // A map to store the books with the ISDN as the key
// This acts as the storage in lieu of an actual database
var bookstore = make(map[string]*Book) // Handler for the books index action
// GET /books
func BookIndex(w http.ResponseWriter, r *http.Request, _ httprouter.Params) {
books := []*Book{}
for _, book := range bookstore {
books = append(books, book)
}
response := &JsonResponse{Data: &books}
w.Header().Set("Content-Type", "application/json; charset=UTF-8")
w.WriteHeader(http.StatusOK)
if err := json.NewEncoder(w).Encode(response); err != nil {
panic(err)
}
} // Handler for the books Show action
// GET /books/:isdn
func BookShow(w http.ResponseWriter, r *http.Request, params httprouter.Params) {
isdn := params.ByName("isdn")
book, ok := bookstore[isdn]
w.Header().Set("Content-Type", "application/json; charset=UTF-8")
if !ok {
// No book with the isdn in the url has been found
w.WriteHeader(http.StatusNotFound)
response := JsonErrorResponse{Error: &ApiError{Status: 404, Title: "Record Not Found"}}
if err := json.NewEncoder(w).Encode(response); err != nil {
panic(err)
}
}
response := JsonResponse{Data: book}
if err := json.NewEncoder(w).Encode(response); err != nil {
panic(err)
}
} func main() {
router := httprouter.New()
router.GET("/", Index)
router.GET("/books", BookIndex)
router.GET("/books/:isdn", BookShow) // Create a couple of sample Book entries
bookstore["123"] = &Book{
ISDN: "123",
Title: "Silence of the Lambs",
Author: "Thomas Harris",
Pages: 367,
} bookstore["124"] = &Book{
ISDN: "124",
Title: "To Kill a Mocking Bird",
Author: "Harper Lee",
Pages: 320,
} log.Fatal(http.ListenAndServe(":8080", router))
}

你如果现在尝试请求 GET https:// localhost:8080/books,将得到以下响应:

{
"meta": null,
"data": [
{
"isdn": "123",
"title": "Silence of the Lambs",
"author": "Thomas Harris",
"pages": 367
},
{
"isdn": "124",
"title": "To Kill a Mocking Bird",
"author": "Harper Lee",
"pages": 320
}
]
}

我们在 main 函数中硬编码了这两个 book 实体。点击这里[2]获取当前代码。

让我们来重构一下代码。到目前为止,所有的代码都放置在同一个文件中:main.go。我们可以把它们划分移到不同的文件中。此时我们有一个目录:

.
├── handlers.go
├── main.go
├── models.go
└── responses.go

我们把所有与 JSON 响应相关的结构体移动到 responses.go,将 handler 函数移动到 Handlers.go,将 Book 结构体移动到 models.go。点击这里[3]查看当前代码。现在,我们跳过来写一些测试。在 Go 中,*_test.go 文件是用作测试用途。因此让我们创建一个 handlers_test.go

package main

import (
"net/http"
"net/http/httptest"
"testing" "github.com/julienschmidt/httprouter"
) func TestBookIndex(t *testing.T) {
// Create an entry of the book to the bookstore map
testBook := &Book{
ISDN: "111",
Title: "test title",
Author: "test author",
Pages: 42,
}
bookstore["111"] = testBook
// A request with an existing isdn
req1, err := http.NewRequest("GET", "/books", nil)
if err != nil {
t.Fatal(err)
}
rr1 := newRequestRecorder(req1, "GET", "/books", BookIndex)
if rr1.Code != 200 {
t.Error("Expected response code to be 200")
}
// expected response
er1 := "{\"meta\":null,\"data\":[{\"isdn\":\"111\",\"title\":\"test title\",\"author\":\"test author\",\"pages\":42}]}\n"
if rr1.Body.String() != er1 {
t.Error("Response body does not match")
}
} // Mocks a handler and returns a httptest.ResponseRecorder
func newRequestRecorder(req *http.Request, method string, strPath string, fnHandler func(w http.ResponseWriter, r *http.Request, param httprouter.Params)) *httptest.ResponseRecorder {
router := httprouter.New()
router.Handle(method, strPath, fnHandler)
// We create a ResponseRecorder (which satisfies http.ResponseWriter) to record the response.
rr := httptest.NewRecorder()
// Our handlers satisfy http.Handler, so we can call their ServeHTTP method
// directly and pass in our Request and ResponseRecorder.
router.ServeHTTP(rr, req)
return rr
}

我们使用 httptest 包的 Recorder 来 mock handler。同样,你也可以为 handler BookShow 编写测试用例。

让我们稍微做些重构。我们仍然把所有路由都定义在了 main 函数中,handler 看起来有点臃肿,需要做点 DRY。我们仍然在终端中输出一些日志消息,并且可以添加一个 BookCreate handler 来创建一个新的 Book。

首先先解决 handlers.go

package main

import (
"encoding/json"
"fmt"
"io"
"io/ioutil"
"net/http" "github.com/julienschmidt/httprouter"
) func Index(w http.ResponseWriter, r *http.Request, _ httprouter.Params) {
fmt.Fprint(w, "Welcome!\n")
} // Handler for the books Create action
// POST /books
func BookCreate(w http.ResponseWriter, r *http.Request, params httprouter.Params) {
book := &Book{}
if err := populateModelFromHandler(w, r, params, book); err != nil {
writeErrorResponse(w, http.StatusUnprocessableEntity, "Unprocessible Entity")
return
}
bookstore[book.ISDN] = book
writeOKResponse(w, book)
} // Handler for the books index action
// GET /books
func BookIndex(w http.ResponseWriter, r *http.Request, _ httprouter.Params) {
books := []*Book{}
for _, book := range bookstore {
books = append(books, book)
}
writeOKResponse(w, books)
} // Handler for the books Show action
// GET /books/:isdn
func BookShow(w http.ResponseWriter, r *http.Request, params httprouter.Params) {
isdn := params.ByName("isdn")
book, ok := bookstore[isdn]
if !ok {
// No book with the isdn in the url has been found
writeErrorResponse(w, http.StatusNotFound, "Record Not Found")
return
}
writeOKResponse(w, book)
} // Writes the response as a standard JSON response with StatusOK
func writeOKResponse(w http.ResponseWriter, m interface{}) {
w.Header().Set("Content-Type", "application/json; charset=UTF-8")
w.WriteHeader(http.StatusOK)
if err := json.NewEncoder(w).Encode(&JsonResponse{Data: m}); err != nil {
writeErrorResponse(w, http.StatusInternalServerError, "Internal Server Error")
}
} // Writes the error response as a Standard API JSON response with a response code
func writeErrorResponse(w http.ResponseWriter, errorCode int, errorMsg string) {
w.Header().Set("Content-Type", "application/json; charset=UTF-8")
w.WriteHeader(errorCode)
json.
NewEncoder(w).
Encode(&JsonErrorResponse{Error: &ApiError{Status: errorCode, Title: errorMsg}})
} //Populates a model from the params in the Handler
func populateModelFromHandler(w http.ResponseWriter, r *http.Request, params httprouter.Params, model interface{}) error {
body, err := ioutil.ReadAll(io.LimitReader(r.Body, 1048576))
if err != nil {
return err
}
if err := r.Body.Close(); err != nil {
return err
}
if err := json.Unmarshal(body, model); err != nil {
return err
}
return nil
}

我创建了两个函数,writeOKResponse 用于将 StatusOK 写入响应,其返回一个 model 或一个 model slice,writeErrorResponse 将在发生预期或意外错误时将 JSON 错误作为响应。像任何一个优秀的 gopher 一样,我们不应该 panic。我还添加了一个名为 populateModelFromHandler 的函数,它将内容从 body 中解析成所需的 model(struct)。在这种情况下,我们在 BookCreate handler 中使用它来填充一个 Book

现在来看看日志。我们简单地创建一个 Logger 函数,它包装了 handler 函数,并在执行 handler 函数之前和之后打印日志消息。

package main

import (
"log"
"net/http"
"time" "github.com/julienschmidt/httprouter"
) // A Logger function which simply wraps the handler function around some log messages
func Logger(fn func(w http.ResponseWriter, r *http.Request, param httprouter.Params)) func(w http.ResponseWriter, r *http.Request, param httprouter.Params) {
return func(w http.ResponseWriter, r *http.Request, param httprouter.Params) {
start := time.Now()
log.Printf("%s %s", r.Method, r.URL.Path)
fn(w, r, param)
log.Printf("Done in %v (%s %s)", time.Since(start), r.Method, r.URL.Path)
}
}

之后来看看路由。首先,在统一一个地方定义所有路由,比如 routes.go

package main

import "github.com/julienschmidt/httprouter"

/*
Define all the routes here.
A new Route entry passed to the routes slice will be automatically
translated to a handler with the NewRouter() function
*/
type Route struct {
Name string
Method string
Path string
HandlerFunc httprouter.Handle
} type Routes []Route func AllRoutes() Routes {
routes := Routes{
Route{"Index", "GET", "/", Index},
Route{"BookIndex", "GET", "/books", BookIndex},
Route{"Bookshow", "GET", "/books/:isdn", BookShow},
Route{"Bookshow", "POST", "/books", BookCreate},
}
return routes
}

让我们创建了一个 NewRouter 函数,它可以在 main 函数中调用,它读取上面定义的所有路由,并返回一个可用的 httprouter.Router。因此创建一个文件 router.go。我们还将使用新创建的 Logger 函数来包装 handler。

package main

import "github.com/julienschmidt/httprouter"

//Reads from the routes slice to translate the values to httprouter.Handle
func NewRouter(routes Routes) *httprouter.Router { router := httprouter.New()
for _, route := range routes {
var handle httprouter.Handle handle = route.HandlerFunc
handle = Logger(handle) router.Handle(route.Method, route.Path, handle)
} return router
}

你的目录此时应该像这样:

.
├── handlers.go
├── handlers_test.go
├── logger.go
├── main.go
├── models.go
├── responses.go
├── router.go
└── routes.go

这里[4]查看完整代码。

到这里,这些代码应该可以帮助你开始编写自己的 API 服务器了。当然,你需要把你的功能放在不同的包中,以下目录结构是一个参照:

.
├── LICENSE
├── README.md
├── handlers
│ ├── books_test.go
│ └── books.go
├── models
│ ├── book.go
│ └── *
├── store
│ ├── *
└── lib
| ├── *
├── main.go
├── router.go
├── rotes.go

如果你有一个大的单体服务,你还可以将 handlersmodels 和所有路由功能都放在另一个名为 app 的包中。你只要记住,go 不像 Java 或 Scala 那样可以有循环的包引用。因此你必须格外注意你的包结构。

这就是全部内容,希望本教程能帮助到你。

上一篇:利用SOLR搭建企业搜索平台 之——模式配置Schema.xml


下一篇:RHEL7通过Rsyslog搭建集中日志服务器