zoukankan      html  css  js  c++  java
  • [Transducer] Lazyness in Transduer

    Transducers remove the requirement of being lazy to optimize for things like take(10). However, it can still be useful to "bind" a collection to a set of transformations and pass it around, without actually evaluating the transformations.

    As noted above, whenever you apply transformations to an iterator it does so lazily. It's easy convert array transformations into a lazy operation, just use the utility function iterator to grab an iterator of the array instead.

    The whole point here is using 'iterator' from transducer.js lib. So you get control when you need the data.

    import t from 'transducers.js';
    
    const doubleTheNumber = number => number * 2;
    export const evenOnly = number => number % 2 === 0;
    
    const doubleAndEven = t.compose(
      t.filter(evenOnly),
      t.map(doubleTheNumber),
    );
    
    const arr = [1,2,3,4,5,6,7,8,9,10];
    const res = t.seq(
      t.iterator(arr),
      t.compose(doubleAndEven, t.take(2)),
    );
    
    function* makeNumbers() {
      let num = 1;
      while (true) yield num++;
    }
    
    const lazyNums = t.seq(makeNumbers(), doubleAndEven);
    
    console.log(
      lazyNums.next(),
      lazyNums.next(),
      lazyNums.next(),
      lazyNums.next(),
      lazyNums.next(),
      lazyNums.next(),
    );
  • 相关阅读:
    VINS_Fusion 框架
    VINS_Fusion 前端源码解析
    堆与优先队列
    LSD-SLAM简介
    直接法和特征点法的区别与优缺点
    CV::Mat介绍
    C++ 位运算
    OPENCV重要函数
    C++ 优先队列
    特征点法的巅峰之作—ORBSLAM2
  • 原文地址:https://www.cnblogs.com/Answer1215/p/8306326.html
Copyright © 2011-2022 走看看