Akka(30): Http:High-Level-Api,Routing DSL

  在上篇我们介绍了Akka-http Low-Level-Api。实际上这个Api提供了Server对进来的Http-requests进行处理及反应的自定义Flow或者转换函数的接入界面。我们看看下面官方文档给出的例子:

import akka.actor.ActorSystem
import akka.http.scaladsl.Http
import akka.http.scaladsl.model.HttpMethods._
import akka.http.scaladsl.model._
import akka.stream.ActorMaterializer
import scala.io.StdIn

object WebServer {

  def main(args: Array[String]) {
    implicit val system = ActorSystem()
    implicit val materializer = ActorMaterializer()
    // needed for the future map/flatmap in the end
    implicit val executionContext = system.dispatcher

    val requestHandler: HttpRequest => HttpResponse = {
      case HttpRequest(GET, Uri.Path("/"), _, _, _) =>
        HttpResponse(entity = HttpEntity(
          ContentTypes.`text/html(UTF-8)`,
          "<html><body>Hello world!</body></html>"))

      case HttpRequest(GET, Uri.Path("/ping"), _, _, _) =>
        HttpResponse(entity = "PONG!")

      case HttpRequest(GET, Uri.Path("/crash"), _, _, _) =>
        sys.error("BOOM!")

      case r: HttpRequest =>
        r.discardEntityBytes() // important to drain incoming HTTP Entity stream
        HttpResponse(404, entity = "Unknown resource!")
    }

    val bindingFuture = Http().bindAndHandleSync(requestHandler, "localhost", 8080)
    println(s"Server online at http://localhost:8080/\nPress RETURN to stop...")
    StdIn.readLine() // let it run until user presses return
    bindingFuture
      .flatMap(_.unbind()) // trigger unbinding from the port
      .onComplete(_ => system.terminate()) // and shutdown when done
  }
}

  我们看到上面例子里的requestHandler函数用模式匹配方式对可能收到的HttpRequest进行了相关HttpResponse的对应。在对应的过程中可能还会按request要求进行一些Server端的运算作为例如Rest-Api这样的服务。不过对于大型的服务,模式匹配方式就会显得篇幅臃肿及模式僵化。Akka-http提供了一套routing DSL作为High-Level-Api的主要组成部分。用routing DSL代替Low-Level-Api的模式匹配方式可以更简练的编制HttpRequest到HttpResponse的转换服务,能更灵活高效的实现现代大型Rest-Api服务。routing DSL实现Rest-Api服务的方式是通过构建一个由组件Directives组合而成的多个多层三明治结构Route。Route是一个类型:

  type Route = RequestContext ⇒ Future[RouteResult]

下面是个Route例子:

    val route: Flow[HttpRequest, HttpResponse, NotUsed]=
      get {
        pathSingleSlash {
          complete(HttpEntity(ContentTypes.`text/html(UTF-8)`,"<html><body>Hello world!</body></html>"))
        } ~
          path("ping") {
            complete("PONG!")
          } ~
          path("crash") {
            sys.error("BOOM!")
          }
      }

在上期讨论的例子里我们可以这样使用route:

  val futBinding: Future[Http.ServerBinding] =
    connSource.to { Sink.foreach{ connection =>
      println(s"client address ${connection.remoteAddress}")
      //    connection handleWith flow
      //    connection handleWithSyncHandler syncHandler
      //connection handleWithAsyncHandler asyncHandler
      connection handleWith route
    }}.run()

handleWith(flow)的参数应该是Flow[HttpRequest,HttpResponse,_]才对呀?这个我们先看看RouteResult对象: 

/**
 * The result of handling a request.
 *
 * As a user you typically don't create RouteResult instances directly.
 * Instead, use the methods on the [[RequestContext]] to achieve the desired effect.
 */
sealed trait RouteResult extends javadsl.server.RouteResult

object RouteResult {
  final case class Complete(response: HttpResponse) extends javadsl.server.Complete with RouteResult {
    override def getResponse = response
  }
  final case class Rejected(rejections: immutable.Seq[Rejection]) extends javadsl.server.Rejected with RouteResult {
    override def getRejections = rejections.map(r ⇒ r: javadsl.server.Rejection).toIterable.asJava
  }

  implicit def route2HandlerFlow(route: Route)(implicit
    routingSettings: RoutingSettings,
                                               parserSettings:   ParserSettings,
                                               materializer:     Materializer,
                                               routingLog:       RoutingLog,
                                               executionContext: ExecutionContext = null,
                                               rejectionHandler: RejectionHandler = RejectionHandler.default,
                                               exceptionHandler: ExceptionHandler = null): Flow[HttpRequest, HttpResponse, NotUsed] =
    Route.handlerFlow(route)
}

这里有个隐式转换route2HandlerFlow把Route转换成Flow[HttpRequest,HttpResponse,NotUsed],问题解决了。

从type Route=RequestContext => Future[RouteResult]可以看到:Route就是一个把RequestContext转换成Future[RouteResult]的函数。RequestContext实质上封装了个Request以及对Request进行操作的环境、配置和工具:

/**
 * This class is not meant to be extended by user code.
 *
 * Immutable object encapsulating the context of an [[akka.http.scaladsl.model.HttpRequest]]
 * as it flows through a akka-http Route structure.
 */
@DoNotInherit
trait RequestContext {

  /** The request this context represents. Modelled as a `val` so as to enable an `import ctx.request._`. */
  val request: HttpRequest

  /** The unmatched path of this context. Modelled as a `val` so as to enable an `import ctx.unmatchedPath._`. */
  val unmatchedPath: Uri.Path

  /**
   * The default ExecutionContext to be used for scheduling asynchronous logic related to this request.
   */
  implicit def executionContext: ExecutionContextExecutor
...
}

Route是一种可组合组件。我们可以用简单的Route组合成更多层次的Route。下面是组合Route的几种方式:

1、Route转化:对输入的request,输出的response进行转化处理后把实际运算托付给下一层内部(inner)Route

2、筛选Route:只容许符合某种条件的Route通过并拒绝其它不符合条件的Route

3、链接Route:假如一个Route被拒绝,尝试下一个Route。这个是通过 ~ 操作符号实现的

在Akka-http的routing DSL里这些Route组合操作是通过Directive实现的。Akka-http提供了大量现成的Directive,我们也可以自定义一些特殊功能的Directive,详情可以查询官方文件或者api文件。

Directive的表达形式如下:

dirname(arguments) { extractions =>
  ... // 内层inner route
}

下面是Directive的一些用例: 

下面的三个route效果相等:

val route: Route = { ctx =>
  if (ctx.request.method == HttpMethods.GET)
    ctx.complete("Received GET")
  else
    ctx.complete("Received something else")
}

val route =
  get {
    complete("Received GET")
  } ~
  complete("Received something else")
  
val route =
  get { ctx =>
    ctx.complete("Received GET")
  } ~
  complete("Received something else")

下面列出一些Directive的组合例子:

val route: Route =
  path("order" / IntNumber) { id =>
    get {
      complete {
        "Received GET request for order " + id
      }
    } ~
    put {
      complete {
        "Received PUT request for order " + id
      }
    }
  }

def innerRoute(id: Int): Route =
  get {
    complete {
      "Received GET request for order " + id
    }
  } ~
  put {
    complete {
      "Received PUT request for order " + id
    }
  }
val route: Route = path("order" / IntNumber) { id => innerRoute(id) }

val route =
  path("order" / IntNumber) { id =>
    (get | put) { ctx =>
      ctx.complete(s"Received ${ctx.request.method.name} request for order $id")
    }
  }

val route =
  path("order" / IntNumber) { id =>
    (get | put) {
      extractMethod { m =>
        complete(s"Received ${m.name} request for order $id")
      }
    }
  }

val getOrPut = get | put
val route =
  path("order" / IntNumber) { id =>
    getOrPut {
      extractMethod { m =>
        complete(s"Received ${m.name} request for order $id")
      }
    }
  }

val route =
  (path("order" / IntNumber) & getOrPut & extractMethod) { (id, m) =>
    complete(s"Received ${m.name} request for order $id")
  }

val orderGetOrPutWithMethod =
  path("order" / IntNumber) & (get | put) & extractMethod
val route =
  orderGetOrPutWithMethod { (id, m) =>
    complete(s"Received ${m.name} request for order $id")
  }

上面例子里的~ & | 定义如下:

object RouteConcatenation extends RouteConcatenation {

  class RouteWithConcatenation(route: Route) {
    /**
     * Returns a Route that chains two Routes. If the first Route rejects the request the second route is given a
     * chance to act upon the request.
     */
    def ~(other: Route): Route = { ctx ⇒
      import ctx.executionContext
      route(ctx).fast.flatMap {
        case x: RouteResult.Complete ⇒ FastFuture.successful(x)
        case RouteResult.Rejected(outerRejections) ⇒
          other(ctx).fast.map {
            case x: RouteResult.Complete               ⇒ x
            case RouteResult.Rejected(innerRejections) ⇒ RouteResult.Rejected(outerRejections ++ innerRejections)
          }
      }
    }
  }
}

  /**
   * Joins two directives into one which runs the second directive if the first one rejects.
   */
  def |[R >: L](that: Directive[R]): Directive[R] =
    recover(rejections ⇒ directives.BasicDirectives.mapRejections(rejections ++ _) & that)(that.ev)

  /**
   * Joins two directives into one which extracts the concatenation of its base directive extractions.
   * NOTE: Extraction joining is an O(N) operation with N being the number of extractions on the right-side.
   */
  def &(magnet: ConjunctionMagnet[L]): magnet.Out = magnet(this)

我们可以从上面这些示范例子得出结论:Directive的组合能力是routing DSL的核心。来看看Directive的组合能力是如何实现的。Directive类定义如下:

//#basic
abstract class Directive[L](implicit val ev: Tuple[L]) {

  /**
   * Calls the inner route with a tuple of extracted values of type `L`.
   *
   * `tapply` is short for "tuple-apply". Usually, you will use the regular `apply` method instead,
   * which is added by an implicit conversion (see `Directive.addDirectiveApply`).
   */
  def tapply(f: L ⇒ Route): Route
  ...
}
  /**
   * Constructs a directive from a function literal.
   */
  def apply[T: Tuple](f: (T ⇒ Route) ⇒ Route): Directive[T] =
    new Directive[T] { def tapply(inner: T ⇒ Route) = f(inner) }

  /**
   * A Directive that always passes the request on to its inner route (i.e. does nothing).
   */
  val Empty: Directive0 = Directive(_(()))
...
  implicit class SingleValueModifiers[T](underlying: Directive1[T]) extends AnyRef {
    def map[R](f: T ⇒ R)(implicit tupler: Tupler[R]): Directive[tupler.Out] =
      underlying.tmap { case Tuple1(value) ⇒ f(value) }

    def flatMap[R: Tuple](f: T ⇒ Directive[R]): Directive[R] =
      underlying.tflatMap { case Tuple1(value) ⇒ f(value) }

    def require(predicate: T ⇒ Boolean, rejections: Rejection*): Directive0 =
      underlying.filter(predicate, rejections: _*).tflatMap(_ ⇒ Empty)

    def filter(predicate: T ⇒ Boolean, rejections: Rejection*): Directive1[T] =
      underlying.tfilter({ case Tuple1(value) ⇒ predicate(value) }, rejections: _*)
  }
}

注意implicit ev: Tuple[L]是给compiler的证例,它要求Tuple[L]存在于可视域。Akka-http提供了所有22个TupleXX[L]的隐形实例。再注意implicit class singleValueModifiers[T]:它提供了多层Directive的自动展平,能够实现下面的自动转换结果:

Directive1[T] = Directive[Tuple1[T]]
Directive1[Tuple2[M,N]] = Directive[Tuple1[Tuple2[M,N]]] = Directive[Tuple2[M,N]]
Directive1[Tuple3[M,N,G]] = ... = Directive[Tuple3[M,N,G]]
Directive1[Tuple4[M1,M2,M3,M4]] = ... = Directive[Tuple4[M1,M2,M3,M4]]
...
Directive1[Unit] = Directive0

Directive1,Directive0:

  type Directive0 = Directive[Unit]
  type Directive1[T] = Directive[Tuple1[T]]

下面是这几种Directive的使用模式:

  dirname { route }                  //Directive0
  dirname[L] { L => route }          //Directive1[L]
  dirname[T] { (T1,T2...) => route}  //Directive[T]

任何类型值到Tuple的自动转换是通过Tupler类实现的:

/**
 * Provides a way to convert a value into an Tuple.
 * If the value is already a Tuple then it is returned unchanged, otherwise it's wrapped in a Tuple1 instance.
 */
trait Tupler[T] {
  type Out
  def OutIsTuple: Tuple[Out]
  def apply(value: T): Out
}

object Tupler extends LowerPriorityTupler {
  implicit def forTuple[T: Tuple]: Tupler[T] { type Out = T } =
    new Tupler[T] {
      type Out = T
      def OutIsTuple = implicitly[Tuple[Out]]
      def apply(value: T) = value
    }
}

private[server] abstract class LowerPriorityTupler {
  implicit def forAnyRef[T]: Tupler[T] { type Out = Tuple1[T] } =
    new Tupler[T] {
      type Out = Tuple1[T]
      def OutIsTuple = implicitly[Tuple[Out]]
      def apply(value: T) = Tuple1(value)
    }
}

我的理解是:Route里Directive的主要功能可以分成两部分:一是如程序菜单拣选,二是对Request,Response,Entity的读写。我们把第二项功能放在以后的讨论里,下面就提供一些RestApi的菜单拣选样例:

trait UsersApi extends JsonMappings{
  val usersApi =
    (path("users") & get ) {
       complete (UsersDao.findAll.map(_.toJson))
    }~
    (path("users"/IntNumber) & get) { id =>
        complete (UsersDao.findById(id).map(_.toJson))
    }~
    (path("users") & post) { entity(as[User]) { user =>
        complete (UsersDao.create(user).map(_.toJson))
      }
    }~
    (path("users"/IntNumber) & put) { id => entity(as[User]) { user =>
        complete (UsersDao.update(user, id).map(_.toJson))
      }
    }~
    (path("users"/IntNumber) & delete) { userId =>
      complete (UsersDao.delete(userId).map(_.toJson))
    }
}

trait CommentsApi extends JsonMappings{
  val commentsApi =
    (path("users"/IntNumber/"posts"/IntNumber/"comments") & get ) {(userId, postId) =>
       complete (CommentsDao.findAll(userId, postId).map(_.toJson))
    }~
      (path("users"/IntNumber/"posts"/IntNumber/"comments"/IntNumber) & get) { (userId, postId, commentId) =>
        complete (CommentsDao.findById(userId, postId, commentId).map(_.toJson))
    }~
      (path("comments") & post) { entity(as[Comment]) { comment =>
        complete (CommentsDao.create(comment).map(_.toJson))
      }
    }~
      (path("users"/IntNumber/"posts"/IntNumber/"comments"/IntNumber) & put) { (userId, postId, commentId) => entity(as[Comment]) { comment =>
        complete (CommentsDao.update(comment, commentId).map(_.toJson))
      }
    }~
      (path("comments"/IntNumber) & delete) { commentId =>
        complete (CommentsDao.delete(commentId).map(_.toJson))
    }
}

trait PostsApi extends JsonMappings{
  val postsApi =
    (path("users"/IntNumber/"posts") & get){ userId =>
      complete (PostsDao.findUserPosts(userId).map(_.toJson))
    }~
    (path("users"/IntNumber/"posts"/IntNumber) & get) { (userId,postId) =>
      complete (PostsDao.findByUserIdAndId(userId, postId).map(_.toJson))
    }~
    (path("users"/IntNumber/"posts") & post) { userId => entity(as[Post]) { post =>
      complete (PostsDao.create(post).map(_.toJson))
    }}~
    (path("users"/IntNumber/"posts"/IntNumber) & put) { (userId, id) => entity(as[Post]) { post =>
      complete (PostsDao.update(post, id).map(_.toJson))
    }}~
    (path("users"/IntNumber/"posts"/IntNumber) & delete) { (userId, postId) =>
      complete (PostsDao.delete(postId).map(_.toJson))
    }
}

  val routes =
    pathPrefix("v1") {
      usersApi ~
      postsApi ~
      commentsApi
    } ~ path("")(getFromResource("public/index.html"))

本文参与腾讯云自媒体分享计划,欢迎正在阅读的你也加入,一起分享。

发表于

我来说两句

0 条评论
登录 后参与评论

相关文章

来自专栏算法修养

PAT 甲级 1021 Deepest Root (并查集,树的遍历)

1021. Deepest Root (25) 时间限制 1500 ms 内存限制 65536 kB 代码长度限制 16000 B ...

2667
来自专栏机器学习入门

LWC 63:746. Min Cost Climbing Stairs

Problem: On a staircase, the i-th step has some non-negative cost cost[i] assig...

1949
来自专栏小鹏的专栏

ubuntu下C++如何调用python程序,gdb调试C++代码

Linux下gdb调试C++代码:http://jingyan.baidu.com/article/acf728fd464984f8e410a369.html ...

2409
来自专栏c#开发者

为什么nhibernate 不能保存on-to-many的结构

下面是主类文件 Code namespace EasyTalk.Module {     /// <summary>     /// SiteAddre...

2515
来自专栏.Net Core 开发记录

[译]Dapper教程

Dapper是一个简单的.NET对象映射器,在速度方面具有"King of Micro ORM"的头衔,几乎与使用原始的ADO.NET数据读取器一样快。ORM是...

482
来自专栏史上最简单的Spring Cloud教程

常见的安全算法

本文整理了常见的安全算法,包括MD5、SHA、DES、AES、RSA等,并写了完整的工具类(Java 版),工具类包含测试。 一、数字摘要算法 数字摘要也称为...

2207
来自专栏Spark学习技巧

Spark高级操作之json复杂和嵌套数据结构的操作一

一,基本介绍 本文主要讲spark2.0版本以后存在的Sparksql的一些实用的函数,帮助解决复杂嵌套的json数据格式,比如,map和嵌套结构。Spark2...

1.9K5
来自专栏Petrichor的专栏

leetcode: 81. Search in Rotated Sorted Array II

leetcode: 33. Search in Rotated Sorted Array

803
来自专栏ml

hdu---(3555)Bomb(数位dp(入门))

Bomb Time Limit: 2000/1000 MS (Java/Others)    Memory Limit: 131072/65536 K (Jav...

3258
来自专栏算法修养

PAT 甲级 1003Emergency(Dijkstra最短路)

1003. Emergency (25) 时间限制 400 ms 内存限制 65536 kB 代码长度限制 16000 B 判题程序...

4007

扫码关注云+社区