记录一下,怕以后忘记,有些还没有看懂,不懂的先猜以及跑例子,之后再补。
目录
准备
reduce
reduceLeft
reduce
reduceRight
Flod
未完待续
关于类型,在控制台运行可以方便观察到类型
scala> val one = 1 // 输出:one: Int = 1, 类型:Int
scala> val list = List(1, 2) // 输出:list: List[Int] = List(1, 2),类型List[Int],中括号里的是范型
scala> val g = (x: Int) => x *3 // 输出:g: Int => Int =
, 对照上面,则类型: Int => Int ,亦可以稍微补全,val g: Int => Int = (x: Int) => x * 3 表示输入Int,输出Int
scala> val product = (x: Int, y: Int) => x * y // 输出:product: (Int, Int) => Int =
,类型:(Int, Int) => Int,稍微补全一下,val product: (Int, Int) => Int = (x: Int, y: Int) => x * y 表示输入有两个参数,类型分别Int和Int,输出是Int类型
总结:就想一下运算的结合律,我们将序列中所有的元素,看成二元运算的两个操作数,由于是相同的运算,不用考虑优先级;
那么就剩左右结合,reduce和reduceLeft是左结合,看成()从左往右依次括起来,即从左往右算;
reduceRight是右结合,看成()从右往左依次括起来,即从右往左算。
reduceLeft,源码如下
def reduceLeft[B >: A](op: (B, A) => B): B = {
if (isEmpty)
throw new UnsupportedOperationException("empty.reduceLeft")
var first = true
var acc: B = 0.asInstanceOf[B]
for (x <- self) {
if (first) {
acc = x
first = false
}
else acc = op(acc, x)
}
acc
}
函数名:reduceLeft,范型[B >: A]
入参op,op的类型是(B,A) => B,参照前面的类型,可以理解为,op是一个函数,接受两个参数,类型分别为B和A,返回类型是B
reduceLeft的整个函数的返回类型也是B。
依据隐式转换以及网上资料搜索,B >: A,表示类型B是类型A的父类(也可以是A)。
self:暂时不知道是啥,目测是代表这个序列
举例:
val list = List(1, 2, 3), list.reduceLeft(_ - _)
1. acc = 0.asInstanceOf[Int],则acc = 0
2. 遍历下标为0的1,则acc = 1
3. 遍历下标为1的2, 则acc = acc - 2 = -1
4. 遍历下标为2的3, 则acc = acc - 3 = -4
5. 将acc作为结果返回,即-4
简而言之,结果=(1 - 2) - 3 = -4,从左往右,先有(1-2),再有(1-2)-3
scala> val list = List(1, 2, 3) // list: List[Int] = List(1, 2, 3)
scala> list.reduce(_ - _) //res120: Int = -4
scala> l // List[Int] = List(1, 2, 3, 4, 5, 6, 7, 8)
scala> l.reduceLeft(_ - _) // res150: Int = -34
可以看成:先有1 - 2, 再有(1 - 2) - 3, 再( (1 - 2) - 3) - 4,即(((((( (1 - 2) - 3) - 4) - 5) - 6) -7) - 8) = -34
reduce,源码如下
def reduce[A1 >: A](op: (A1, A1) => A1): A1 = reduceLeft(op)
直接调用的reduceLeft函数,所以结果同上,
scala> list // res127: List[Int] = List(1, 2, 3)
scala> list.reduce(_ - _) // res126: Int = -4
reduceRight,源码如下
def reduceRight[B >: A](op: (A, B) => B): B = {
if (isEmpty)
throw new UnsupportedOperationException("empty.reduceRight")
// 这边注意一下顺序,(x, y) => op(y, x)
reversed.reduceLeft[B]((x, y) => op(y, x))
}
override /*TraversableLike*/
def reduceLeft[B >: A](f: (B, A) => B): B =
if (isEmpty) throw new UnsupportedOperationException("empty.reduceLeft")
else tail.foldLeft[B](head)(f)
// currying函数
override /*TraversableLike*/
def foldLeft[B](z: B)(@deprecatedName('f) op: (B, A) => B): B = {
var acc = z
var these = this
while (!these.isEmpty) {
acc = op(acc, these.head)
these = these.tail
}
acc
}
其实就是一句话: reversed.reduceLeft[B]((x, y) => op(y, x)),我们先来看看reversed
reversed,源码如下
protected[this] def reversed = {
var elems: List[A] = Nil
self foreach (elems ::= _)
elems
}
这个reversed,生成翻转的序列,我们瞄一下
scala> l // List[Int] = List(1, 2, 3, 4, 5, 6, 7, 8)
scala> var nil: List[Int] = Nil // reversedL: List[Int] = List()
scala> l foreach (nil::= _) //
scala> reversedL // res136: List[Int] = List(8, 7, 6, 5, 4, 3, 2, 1)
再来看这一句话,reversed.reduceLeft[B]((x, y) => op(y, x))
与reduce也即reduceLeft的区别:1.在反转序列的基础上做的;2.调用的是LinearSeqOptimized类的foldLeft函数。
举例:
val list = List(1, 2, 3), list.reduceRight(_ - _)
A. 生成临时反转序列,reversedL, (3, 2, 1)
1. acc = 3,these=(2,1)
2. these不为空, acc= these.head - acc= 2 - 3 = -1, these = (1,)
3. these不为空, acc= these.head - acc= 1 - (-1) = 2, these = ()
4. theae为空,退出循环,结果返回,即2
简而言之,结果= 1 - (2 - 3) = 2,这边看成从右往左写,即先有2 - 3的结果,再有1 - (2 - 3)
scala> val list = List(1, 2, 3) // list: List[Int] = List(1, 2, 3)
scala> list.reduce(_ - _) //res120: Int = 2
scala> l // List[Int] = List(1, 2, 3, 4, 5, 6, 7, 8)
scala> l.reduceLeft(_ - _) // res150: Int = -4
可以看成:先有(7 - 8),再有6 - (7 - 8), 再有5 - (6 -( 7- 8)), 即 (1-(2-(3-(4-(5-(6-(7-8))))))) = - 4,