Scala 2.11 + Play Framework 2.3中的22个字段限制，Case类和函数

Question

Scala 2.11 + Play Framework 2.3中的22个字段限制，Case类和函数

37

Scala 2.11已经推出，而对于case class的22个字段限制似乎已经得到解决 (Scala Issue, 发布说明)。

这对我来说一直是个问题，因为我使用case class来建模在Play + Postgres Async环境下拥有超过22个字段的数据库实体。在Scala 2.10中，我的解决方案是将模型分成多个case class，但我发现这个解决方案很难维护和扩展。希望在切换到Play 2.3.0-RC1 + Scala 2.11.0后，我可以按照以下描述实现某些东西:

package entities

case class MyDbEntity(
  id: String,
  field1: String,
  field2: Boolean,
  field3: String,
  field4: String,
  field5: String,
  field6: String,
  field7: String,
  field8: String,
  field9: String,
  field10: String,
  field11: String,
  field12: String,
  field13: String,
  field14: String,
  field15: String,
  field16: String,
  field17: String,
  field18: String,
  field19: String,
  field20: String,
  field21: String,
  field22: String,
  field23: String,
) 

object MyDbEntity {
  import play.api.libs.json.Json
  import play.api.data._
  import play.api.data.Forms._

  implicit val entityReads = Json.reads[MyDbEntity]
  implicit val entityWrites = Json.writes[MyDbEntity]
}

上面的代码在"Reads"和"Writes"两个地方编译失败，错误信息如下：

找不到unapply函数

将"Reads"和"Writes"更新为以下内容：

  implicit val entityReads: Reads[MyDbEntity] = (
    (__ \ "id").read[Long] and
    (__ \ "field_1").read[String]
    ........
  )(MyDbEntity.apply _)  

  implicit val postWrites: Writes[MyDbEntity] = (
    (__ \ "id").write[Long] and
    (__ \ "user").write[String]
    ........
  )(unlift(MyDbEntity.unapply))

还是不起作用：

  implementation restricts functions to 22 parameters

  value unapply is not a member of object models.MyDbEntity

我的理解是，Scala 2.11 在函数方面仍然存在一些限制，像我上面描述的那样还不可能实现。这对我来说似乎很奇怪，因为如果一个主要用户案例的支持仍然没有得到支持，那么取消对 case classes 的限制就没有任何好处，所以我想知道是否有什么我没注意到的东西。

欢迎提供问题指针或实现细节！谢谢！

- luis

3

请看此相关拉取请求描述：第一件提到的限制是缺少>22个类的“unapply”函数，这是有原因的（据我所记，类文件大小会呈指数级增长）。 - om-nom-nom

6

一个具有超过22个参数的case类无法拥有unapply方法，因为它必须转换为具有X>22的TupleX类型，并且元组仍然限制为22。可悲的是，Json.format [MyCaseClass](基于宏的解决方案)并不需要有这种限制(它可以意识到它是一个case类，并直接提取字段，而无需使用unapply，就像模式匹配一样)，但目前它会寻找unapply并失败... - gourlaysama

2

对于那些感兴趣的人，这里有一个跟踪问题的票据，可以解决这个限制：https://github.com/playframework/playframework/issues/3174 - ppearcy

这个问题仍然没有解决。在使用Scala 2.11.1的2.3版本中，是否仍然存在大于22的问题？我正在为一个新项目研究不同的堆栈，想知道这是否会成为一个问题。谢谢。 - Foo L

8个回答

3

我们曾将模型拆分为多个case class，但这很快变得难以管理。我们使用Slick作为我们的对象关系映射器，而Slick 2.0带有一个代码生成器，我们用它来生成类（其带有应用程序方法和复制构造函数以模仿case classes），以及从Json实例化模型的方法（因为我们有太多特殊情况要处理，所以我们不会自动生成将模型转换为Json的方法）。使用Slick代码生成器不需要您使用Slick作为您的对象关系映射器。

这是代码生成器的输入部分 - 此方法接受JsObject并使用它来实例化新模型或更新现有模型。

private def getItem(original: Option[${name}], json: JsObject, trackingData: TrackingData)(implicit session: scala.slick.session.Session): Try[${name}] = {
  preProcess("$name", columnSet, json, trackingData).flatMap(updatedJson => {
    ${indent(indent(indent(entityColumnsSansId.map(c => s"""val ${c.name}_Parsed = parseJsonField[${c.exposedType}](original.map(_.${c.name}), "${c.name}", updatedJson, "${c.exposedType}")""").mkString("\n"))))}
    val errs = Seq(${indent(indent(indent(indent(entityColumnsSansId.map(c => s"${c.name}_Parsed.map(_ => ())").mkString(", ")))))}).condenseUnit
    for {
      _ <- errs
      ${indent(indent(indent(indent(entityColumnsSansId.map(c => s"${c.name}_Val <- ${c.name}_Parsed").mkString("\n")))))}
    } yield {
      original.map(_.copy(${entityColumnsSansId.map(c => s"${c.name} = ${c.name}_Val").mkString(", ")}))
        .getOrElse(${name}.apply(id = None, ${entityColumnsSansId.map(c => s"${c.name} = ${c.name}_Val").mkString(", ")}))
    }
  })
}

例如，对于我们的ActivityLog模型，这将产生以下代码。如果“original”是None，则这是从“createFromJson”方法中调用的并且我们实例化一个新模型；如果“original”是Some(activityLog)，则这是从“updateFromJson”方法中调用的，我们更新现有模型。“condenseUnit”方法在“val errs = ...”行上被调用，该方法需要Seq[Try[Unit]]并生成一个Try[Unit]；如果Seq有任何错误，则Try[Unit]将连接异常消息。parseJsonField和parseField方法不是由生成的代码生成的-它们只是从生成的代码中引用。

private def parseField[T](name: String, json: JsObject, tpe: String)(implicit r: Reads[T]): Try[T] = {
  Try((json \ name).as[T]).recoverWith {
    case e: Exception => Failure(new IllegalArgumentException("Failed to parse " + Json.stringify(json \ name) + " as " + name + " : " + tpe))
  }
}

def parseJsonField[T](default: Option[T], name: String, json: JsObject, tpe: String)(implicit r: Reads[T]): Try[T] = {
  default match {
    case Some(t) => if(json.keys.contains(name)) parseField(name, json, tpe)(r) else Try(t)
    case _ => parseField(name, json, tpe)(r)
  }
}

private def getItem(original: Option[ActivityLog], json: JsObject, trackingData: TrackingData)(implicit session: scala.slick.session.Session): Try[ActivityLog] = {
  preProcess("ActivityLog", columnSet, json, trackingData).flatMap(updatedJson => {
    val user_id_Parsed = parseJsonField[Option[Int]](original.map(_.user_id), "user_id", updatedJson, "Option[Int]")
    val user_name_Parsed = parseJsonField[Option[String]](original.map(_.user_name), "user_name", updatedJson, "Option[String]")
    val item_id_Parsed = parseJsonField[Option[String]](original.map(_.item_id), "item_id", updatedJson, "Option[String]")
    val item_item_type_Parsed = parseJsonField[Option[String]](original.map(_.item_item_type), "item_item_type", updatedJson, "Option[String]")
    val item_name_Parsed = parseJsonField[Option[String]](original.map(_.item_name), "item_name", updatedJson, "Option[String]")
    val modified_Parsed = parseJsonField[Option[String]](original.map(_.modified), "modified", updatedJson, "Option[String]")
    val action_name_Parsed = parseJsonField[Option[String]](original.map(_.action_name), "action_name", updatedJson, "Option[String]")
    val remote_ip_Parsed = parseJsonField[Option[String]](original.map(_.remote_ip), "remote_ip", updatedJson, "Option[String]")
    val item_key_Parsed = parseJsonField[Option[String]](original.map(_.item_key), "item_key", updatedJson, "Option[String]")
    val created_at_Parsed = parseJsonField[Option[java.sql.Timestamp]](original.map(_.created_at), "created_at", updatedJson, "Option[java.sql.Timestamp]")
    val as_of_date_Parsed = parseJsonField[Option[java.sql.Timestamp]](original.map(_.as_of_date), "as_of_date", updatedJson, "Option[java.sql.Timestamp]")
    val errs = Seq(user_id_Parsed.map(_ => ()), user_name_Parsed.map(_ => ()), item_id_Parsed.map(_ => ()), item_item_type_Parsed.map(_ => ()), item_name_Parsed.map(_ => ()), modified_Parsed.map(_ => ()), action_name_Parsed.map(_ => ()), remote_ip_Parsed.map(_ => ()), item_key_Parsed.map(_ => ()), created_at_Parsed.map(_ => ()), as_of_date_Parsed.map(_ => ())).condenseUnit
    for {
      _ <- errs
      user_id_Val <- user_id_Parsed
      user_name_Val <- user_name_Parsed
      item_id_Val <- item_id_Parsed
      item_item_type_Val <- item_item_type_Parsed
      item_name_Val <- item_name_Parsed
      modified_Val <- modified_Parsed
      action_name_Val <- action_name_Parsed
      remote_ip_Val <- remote_ip_Parsed
      item_key_Val <- item_key_Parsed
      created_at_Val <- created_at_Parsed
      as_of_date_Val <- as_of_date_Parsed
    } yield {
      original.map(_.copy(user_id = user_id_Val, user_name = user_name_Val, item_id = item_id_Val, item_item_type = item_item_type_Val, item_name = item_name_Val, modified = modified_Val, action_name = action_name_Val, remote_ip = remote_ip_Val, item_key = item_key_Val, created_at = created_at_Val, as_of_date = as_of_date_Val))
        .getOrElse(ActivityLog.apply(id = None, user_id = user_id_Val, user_name = user_name_Val, item_id = item_id_Val, item_item_type = item_item_type_Val, item_name = item_name_Val, modified = modified_Val, action_name = action_name_Val, remote_ip = remote_ip_Val, item_key = item_key_Val, created_at = created_at_Val, as_of_date = as_of_date_Val))
    }
  })
}

- Zim-Zam O'Pootertoot

2

您可以使用Jackson的Scala模块。Play的json功能是基于Jackson scala构建的。我不知道为什么他们在这里设置了22个字段限制，而Jackson支持超过22个字段。一个函数调用可能永远不能使用超过22个参数，但我们可以在DB实体中有数百列，所以这里的限制是荒谬的，使Play成为一个不太高效的玩具。请看这个：

import com.fasterxml.jackson.databind.ObjectMapper
import com.fasterxml.jackson.module.scala.experimental.ScalaObjectMapper
import com.fasterxml.jackson.module.scala.DefaultScalaModule

object JacksonUtil extends App {
  val mapper = new ObjectMapper with ScalaObjectMapper
  mapper.registerModule(DefaultScalaModule)


  val t23 = T23("a","b","c","d","e","f","g","h","i","j","k","l","m","n","o","p","q","r","s","t","u","v","w")

  println(mapper.writeValueAsString(t23))
 }
case class T23(f1:String,f2:String,f3:String,f4:String,f5:String,f6:String,f7:String,
    f8:String,f9:String,f10:String,f11:String,f12:String,f13:String,f14:String,f15:String,
    f16:String,f17:String,f18:String,f19:String,f20:String,f21:String,f22:String,f23:String)

- CharlieQ

2

看起来这个处理得很好。 +22个字段的案例类格式化器等play-json扩展 https://github.com/xdotai/play-json-extensions 支持Scala 2.11.x、2.12.x和2.13.x以及play 2.3、2.4、2.5和2.7。

并且在play-json issue中被引用为首选解决方案（但尚未合并）。

- Shanness

1

有些情况下，案例类可能不适用；其中一个情况是案例类不能超过22个字段。另一种情况可能是您事先不知道模式。在这种方法中，数据被加载为行对象的RDD。使用StructType和StructField对象分别创建表示表格和字段的模式。将模式应用于行RDD以在Spark中创建DataFrame。

- Venu A Positive

1

我正在制作一个库。请尝试使用这个https://github.com/xuwei-k/play-twenty-three。

- Kenji Yoshida

没有读完所有的代码，但我至少看到了一个提到252个字段的引用!!! 哈哈 :) - pedrofurla

2

"640K应该足够每个人使用了" - Coder Guy

0

我尝试了另一个答案中提出的基于Shapeless“自动类型类派生”的解决方案，但它对我们的模型不起作用 - 会抛出StackOverflow异常（包含约30个字段和4个嵌套的包含4-10个字段的case class集合）。

因此，我们采用了this的解决方案，并且它完美地工作。通过编写ScalaCheck测试进行了确认。请注意，它需要Play Json 2.4。

- Tvaroh

0

在 dotty（Scala 3）中，现在您可以在 Case 类中使用超过22个字段。

- John

网页内容由stack overflow 提供, 点击上面的

可以查看英文原文，
原文链接

- mchv · Accepted Answer

由于几个原因，这不可能直接实现：

首先，正如gourlaysama所指出的那样，play-json库使用scala宏来避免样板代码，而当前代码依赖于unapply和apply方法来检索字段。这解释了你问题中的第一个错误信息。
其次，play-json库依赖于一个功能库，该库目前仅使用固定数量的参数与先前的案例类字段度数极限相对应。这解释了你问题中的第二个错误消息。

但是可以通过以下方式之一绕过第二点：

使用shapeless的自动类型类推导功能。 Naveen Gattu编写了一份非常好的概述。
覆盖默认的功能构建器

首先，创建缺失的FunctionalBuilder：

class CustomFunctionalBuilder[M[_]](canBuild: FunctionalCanBuild[M]) extends FunctionalBuilder {

    class CustomCanBuild22[A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22](m1: M[A1 ~ A2 ~ A3 ~ A4 ~ A5 ~ A6 ~ A7 ~ A8 ~ A9 ~ A10 ~ A11 ~ A12 ~ A13 ~ A14 ~ A15 ~ A16 ~ A17 ~ A18 ~ A19 ~ A20 ~ A21], m2: M[A22]) {
def ~[A23](m3: M[A23]) = new CustomCanBuild23[A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22, A23](canBuild(m1, m2), m3)

  def and[A23](m3: M[A23]) = this.~(m3)

  def apply[B](f: (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22) => B)(implicit fu: Functor[M]): M[B] =
  fu.fmap[A1 ~ A2 ~ A3 ~ A4 ~ A5 ~ A6 ~ A7 ~ A8 ~ A9 ~ A10 ~ A11 ~ A12 ~ A13 ~ A14 ~ A15 ~ A16 ~ A17 ~ A18 ~ A19 ~ A20 ~ A21 ~ A22, B](canBuild(m1, m2), { case a1 ~ a2 ~ a3 ~ a4 ~ a5 ~ a6 ~ a7 ~ a8 ~ a9 ~ a10 ~ a11 ~ a12 ~ a13 ~ a14 ~ a15 ~ a16 ~ a17 ~ a18 ~ a19 ~ a20 ~ a21 ~ a22 => f(a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22) })

  def apply[B](f: B => (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22))(implicit fu: ContravariantFunctor[M]): M[B] =
  fu.contramap(canBuild(m1, m2), (b: B) => { val (a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22) = f(b); new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(a1, a2), a3), a4), a5), a6), a7), a8), a9), a10), a11), a12), a13), a14), a15), a16), a17), a18), a19), a20), a21), a22) })

  def apply[B](f1: (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22) => B, f2: B => (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22))(implicit fu: InvariantFunctor[M]): M[B] =
  fu.inmap[A1 ~ A2 ~ A3 ~ A4 ~ A5 ~ A6 ~ A7 ~ A8 ~ A9 ~ A10 ~ A11 ~ A12 ~ A13 ~ A14 ~ A15 ~ A16 ~ A17 ~ A18 ~ A19 ~ A20 ~ A21 ~ A22, B](
    canBuild(m1, m2), { case a1 ~ a2 ~ a3 ~ a4 ~ a5 ~ a6 ~ a7 ~ a8 ~ a9 ~ a10 ~ a11 ~ a12 ~ a13 ~ a14 ~ a15 ~ a16 ~ a17 ~ a18 ~ a19 ~ a20 ~ a21 ~ a22 => f1(a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22) },
    (b: B) => { val (a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22) = f2(b); new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(new ~(a1, a2), a3), a4), a5), a6), a7), a8), a9), a10), a11), a12), a13), a14), a15), a16), a17), a18), a19), a20), a21), a22) }
  )

  def join[A >: A1](implicit witness1: <:<[A, A1], witness2: <:<[A, A2], witness3: <:<[A, A3], witness4: <:<[A, A4], witness5: <:<[A, A5], witness6: <:<[A, A6], witness7: <:<[A, A7], witness8: <:<[A, A8], witness9: <:<[A, A9], witness10: <:<[A, A10], witness11: <:<[A, A11], witness12: <:<[A, A12], witness13: <:<[A, A13], witness14: <:<[A, A14], witness15: <:<[A, A15], witness16: <:<[A, A16], witness17: <:<[A, A17], witness18: <:<[A, A18], witness19: <:<[A, A19], witness20: <:<[A, A20], witness21: <:<[A, A21], witness22: <:<[A, A22], fu: ContravariantFunctor[M]): M[A] =
  apply[A]((a: A) => (a: A1, a: A2, a: A3, a: A4, a: A5, a: A6, a: A7, a: A8, a: A9, a: A10, a: A11, a: A12, a: A13, a: A14, a: A15, a: A16, a: A17, a: A18, a: A19, a: A20, a: A21, a: A22))(fu)

  def reduce[A >: A1, B](implicit witness1: <:<[A1, A], witness2: <:<[A2, A], witness3: <:<[A3, A], witness4: <:<[A4, A], witness5: <:<[A5, A], witness6: <:<[A6, A], witness7: <:<[A7, A], witness8: <:<[A8, A], witness9: <:<[A9, A], witness10: <:<[A10, A], witness11: <:<[A11, A], witness12: <:<[A12, A], witness13: <:<[A13, A], witness14: <:<[A14, A], witness15: <:<[A15, A], witness16: <:<[A16, A], witness17: <:<[A17, A], witness18: <:<[A18, A], witness19: <:<[A19, A], witness20: <:<[A20, A], witness21: <:<[A21, A], witness22: <:<[A22, A], fu: Functor[M], reducer: Reducer[A, B]): M[B] =
  apply[B]((a1: A1, a2: A2, a3: A3, a4: A4, a5: A5, a6: A6, a7: A7, a8: A8, a9: A9, a10: A10, a11: A11, a12: A12, a13: A13, a14: A14, a15: A15, a16: A16, a17: A17, a18: A18, a19: A19, a20: A20, a21: A21, a22: A22) =>  reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.append(reducer.unit(a1: A), a2: A), a3: A), a4: A), a5: A), a6: A), a7: A), a8: A), a9: A), a10: A), a11: A), a12: A), a13: A), a14: A), a15: A), a16: A), a17: A), a18: A), a19: A), a20: A), a21: A), a22: A))(fu)

  def tupled(implicit v: VariantExtractor[M]): M[(A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22)] =
  v match {
    case FunctorExtractor(fu) => apply { (a1: A1, a2: A2, a3: A3, a4: A4, a5: A5, a6: A6, a7: A7, a8: A8, a9: A9, a10: A10, a11: A11, a12: A12, a13: A13, a14: A14, a15: A15, a16: A16, a17: A17, a18: A18, a19: A19, a20: A20, a21: A21, a22: A22) => (a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22) }(fu)
    case ContravariantFunctorExtractor(fu) => apply[(A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22)] { (a: (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22)) => (a._1, a._2, a._3, a._4, a._5, a._6, a._7, a._8, a._9, a._10, a._11, a._12, a._13, a._14, a._15, a._16, a._17, a._18, a._19, a._20, a._21, a._22) }(fu)
    case InvariantFunctorExtractor(fu) => apply[(A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22)]({ (a1: A1, a2: A2, a3: A3, a4: A4, a5: A5, a6: A6, a7: A7, a8: A8, a9: A9, a10: A10, a11: A11, a12: A12, a13: A13, a14: A14, a15: A15, a16: A16, a17: A17, a18: A18, a19: A19, a20: A20, a21: A21, a22: A22) => (a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22) }, { (a: (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22)) => (a._1, a._2, a._3, a._4, a._5, a._6, a._7, a._8, a._9, a._10, a._11, a._12, a._13, a._14, a._15, a._16, a._17, a._18, a._19, a._20, a._21, a._22) })(fu)
    }

  }

  class CustomCanBuild23[A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A20, A21, A22, A23](m1: M[A1 ~ A2 ~ A3 ~ A4 ~ A5 ~ A6 ~ A7 ~ A8 ~ A9 ~ A10 ~ A11 ~ A12 ~ A13 ~ A14 ~ A15 ~ A16 ~ A17 ~ A18 ~ A19 ~ A20 ~ A21 ~ A22], m2: M[A23]) {
  }

}

然后通过提供您自己的FunctionalBuilderOps实例：

implicit def customToFunctionalBuilderOps[M[_], A](a: M[A])(implicit fcb: FunctionalCanBuild[M]) = new CustomFunctionalBuilderOps[M, A](a)(fcb)

最后，关于第一点，我已经发送了一个拉取请求（pull request），试图简化当前的实现。