uniplate 的 `universeBi` 可以用于以广度优先的方式检索节点吗？

Question

是否可以使用 Uniplate 的 universeBi 以广度优先顺序获得输出？结果似乎以深度优先的方式返回。我想知道如何使用 uniplate 以广度优先的方式检索 universeBi。

为了说明，请考虑以下玩具程序：

{-# LANGUAGE DeriveDataTypeable #-}

import Data.Data
import Data.Generics.Uniplate.Data

data A = A B Int deriving (Data, Typeable)
data B = B Int   deriving (Data, Typeable)

val :: A
val = A (B 1) 2

ints :: [Int]
ints = universeBi val

我得到：

*Main> ints
[1,2]

但这是深度优先的，因为 1 是从 B 节点获得的。我宁愿按广度优先顺序获取它，即接收 [2,1]。这在 uniplate 中可以实现吗？

Answer 1

您可以深入研究 biplate 返回的 Str 的结构：

layers :: Str a -> [[a]]
layers Zero = []
layers (One x) = [[x]]
layers (Two f x) = catLayers (layers f) ([] : layers x)
  where catLayers [] ys = ys
        catLayers xs [] = xs
        catLayers (x : xs) (y : ys) = (x ++ y) : catLayers xs ys
layersBi :: Biplate from to => from -> [[to]]
layersBi = layers . fst . biplate
breadthBi :: Biplate from to => from -> [to]
breadthBi = concat . layersBi

所以现在

breadthBi (A (B 1) 2) :: [Int]
-- = [2, 1]

和

data Tree a = Branch (Tree a) a (Tree a) | Leaf deriving (Data, Typeable)
--       4
--   2       6
-- 1   3   5   7
example = Branch (Branch (Branch Leaf 1 Leaf) 2 (Branch Leaf 3 Leaf)) 4 (Branch (Branch Leaf 5 Leaf) 6 (Branch Leaf 7 Leaf))
(layersBi :: Data a => Tree a -> [[a]]) example
-- = [[],[4],[2,6],[1,3,5,7]]

我不确定是否真的保证 Str 准确反映了数据类型的结构，但看起来确实如此。如果必须的话，你可以用 Data 原语做一些东西。

uniplate 的 `universeBi` 可以用于以广度优先的方式检索节点吗？

Can uniplate's `universeBi` be used to retrieve nodes in a breadth-first fashion?

haskell

generic-programming

uniplate