前言:
什么？砖升本？不行，戒毒后只想学习代码。跟随Geemo狗的节奏，学习下他的stream深入篇，好的直接略过博文看最后的参考资料 
，美团前端的文章，那么我就跟随着来上那么一发。
美团原文: http://fe.meituan.com/stream-internals.html
流的应用:
先来一发，我写了个复制java学习资料的栗子。代码如下。(知乎上有参考)，clearLine，cursorTo等方法已经从process.stdout移到了Readline那里，具体可以看极客的API或者英文API。
"use strict"
const readline = require('readline');
const fs = require('fs');
const filePath = 'snis716.avi';
const reFilePath = 'javlibrary.avi';
let passedLength = 0;
let lastSize = 0;
let readStream = fs.createReadStream(filePath);
let writeStream = fs.createWriteStream(reFilePath);
//返回stat数组
let stat=fs.statSync(filePath);
const totalSize=stat.size;
console.log(stat);
readStream.on('data',chunk =>{
passedLength+= chunk.length;
if(writeStream.write(chunk) === false ){
readStream.pause();
}
})
readStream.on('end', ()=> {
writeStream.end();
})
//drain事件：可以向流中写入更多数据
writeStream.on('drain', ()=> {
readStream.resume();//继续触发data
})
console.info(`正在复制${filePath}到${reFilePath}`);
console.time('copyTime');
setTimeout(function show() {
let present = Math.ceil((passedLength/totalSize)*100);
let size = Math.ceil((passedLength/1048576));
let currentSize = size - lastSize;
lastSize = size;
//清除本行
readline.clearLine(process.stdout, 0)
//光标移向本行初始位置
readline.cursorTo(process.stdout, 0, null)
process.stdout.write(`正在复制,写入速率${currentSize*2}MB/s,已写入${size}MB`);
if (passedLength < totalSize) {
setTimeout(show, 500);
} else {
console.log('\n');
console.timeEnd('copyTime');
}},500)
这段代码就复制了G级文件，当然也可以用pipe 实现自动控制。
fs.createReadStream(‘snis716.avi’).pipe(fs.createWriteStream(‘javlibrary.avi’));
Stream 潜入
先上API http://nodeapi.ucdok.com/#/api/stream.html
var Stream = require(‘stream’);
var Readable = Stream.Readable;//可读流
var Writable = Stream.Writable;//可写流
var Duplex = Stream.Duplex;//可读可写流
var Transform = Stream.Transform;//因果关系的双工流
var Readable = Stream.Readable;//可读流
var Writable = Stream.Writable;//可写流
var Duplex = Stream.Duplex;//可读可写流
var Transform = Stream.Transform;//因果关系的双工流
Readable:
下游请求数据，会调用_read()方法，这东西一方面从数据源获取数据，一方面Push并触发data事件。
Read实现:
//Readable局部源码
var doRead = state.needReadable;
// if we currently have less than the highWaterMark, then also read some
if (state.length === 0 || state.length - n < state.highWaterMark) {
doRead = true;
}
if (state.ended || state.reading) {
doRead = false;
debug('reading or ended', doRead);
}
if (doRead) {
debug('do read');
state.reading = true;
state.sync = true;
if (state.length === 0)
state.needReadable = true;
// call internal read method
this._read(state.highWaterMark);
state.sync = false;
}
// 如果_read是同步的那么reading为false
// 从缓存获取的实际数据量
if (doRead && !state.reading){
n = howMuchToRead(nOrig, state);
state.sync = false
}
state.highWaterMark 缓存阈值
state.length 当前缓存数据量
state.flowing 流动模式
sync push方法的同/异步
state.needReadable 缓存是否充足，不充足从底层来数据
reading 上一次从底层获取数据是否结束
后记
今天到这了，记录下，后面的内容看Geemo和美团了，懒得码字了，还是这是别人理解出来的东西，接下来就要自己看源码，结合下别人的理解，去“看懂”源码了。有时候了解个原理过程很简单，但是能设计并应用上去却很难，剩下的看源码了，我这就不重复搬弄了。
