输出和输出

在序列工作流程中,yield 将单个项目添加到正在构建的序列中。 (在 monadic 术语中,它是 return。)

> seq { yield 1; yield 2; yield 3 }
val it: seq<int> = seq [1; 2; 3]

> let homogenousTup2ToSeq (a, b) = seq { yield a; yield b }
> tup2Seq ("foo", "bar")
val homogenousTup2ToSeq: 'a * 'a -> seq<'a>
val it: seq<string> = seq ["foo"; "bar"]

yield!(发音为 yield bang )将另一个序列的所有项插入到正在构建的序列中。或者,换句话说,它附加一个序列。 (关于 monad,它是 bind。)

> seq { yield 1; yield! [10;11;12]; yield 20 }
val it: seq<int> = seq [1; 10; 11; 12; 20]

// Creates a sequence containing the items of seq1 and seq2 in order
> let concat seq1 seq2 = seq { yield! seq1; yield! seq2 }
> concat ['a'..'c'] ['x'..'z']
val concat: seq<'a> -> seq<'a> -> seq<'a>
val it: seq<int> = seq ['a'; 'b'; 'c'; 'x'; 'y'; 'z']

由序列工作流创建的序列也是惰性的,这意味着序列的项目在需要之前不会被实际评估。强制项目的几种方法包括调用 Seq.take(将前 n 个项目拉入序列),Seq.iter(将函数应用于每个项目以执行副作用)或 Seq.toList(将序列转换为列表)。将此与递归相结合是 yield! 真正开始闪耀的地方。

> let rec numbersFrom n = seq { yield n; yield! numbersFrom (n + 1) }
> let naturals = numbersFrom 0
val numbersFrom: int -> seq<int>
val naturals: seq<int> = seq [0; 1; 2; ...]

// Just like Seq.map: applies a mapping function to each item in a sequence to build a new sequence
> let rec map f seq1 =
      if Seq.isEmpty seq1 then Seq.empty
      else seq { yield f (Seq.head seq1); yield! map f (Seq.tail seq1) }
> map (fun x -> x * x) [1..10]
val map: ('a -> 'b) -> seq<'a> -> 'b
val it: seq<int> = seq [1; 4; 9; 16; 25; 36; 49; 64; 81; 100]