Categories on Y-AX

ES6 Array.prototype.fill For Object Fix

Sun, 23 Oct 2016 00:00:00 +0000

In ES6, we have Array.prototype.fill method, which will help us to initialize arrays with same values. for example: new Array(5).fill(0) will get [0,0,0,0,0]. it’s a very handy method.

But we when you use it with object, like new Array(5).fill([]), or new Array(5).fill({}), you will probaly got hurt by this method:

let arr = new Array(5).fill({});
arr[0].name = 'test';
console.log(arr);

you will got:

[{name:'test'}, {name:'test'}, {name:'test'}, {name:'test'}, {name:'test'}]

the same are for arrays, as array is object in JS too.

As in JS, objects are passed by reference, and Array.prototype.fill do not consider this diffirence, so it’s impossible to just use fill to create an array of empty objects or arrays.

We can use map to do the same thing actually:

new Array(5).fill(null).map(()=>{})

Solve NumberOfDiscIntersections

Fri, 14 Oct 2016 00:00:00 +0000

For the NumberOfDiscIntersections problem of Codility site. it’s a little hard to figure it out.

The Question

At first, I use bruce force method to solve it, the complexity is O(n^^2)

function solution(A) {
    // write your code in JavaScript (Node.js 6.4.0)
    A = A.map((x, i) => [i-x, i+x])
    let c = 0;
    for(i = 0; i< A.length; i++){
        for(j = i+1; j < A.length; j++){
            if(A[i][1] >= A[j][0]){
                c++
                if(c > 10000000) return -1;
            }
        }
    }
    return c;
}

On the web page, they require the worst complexity is O(nlogn), so we need find a better solution.

I searched on web, many people have a solution which sort the boundary of disc, and its complexity is O(nlogn), it suggest to use binary search whose complexity is O(logn), so I try and got this:

function solution(A) {
    // write your code in JavaScript (Node.js 6.4.0)
    A = A.map((x, i) => [i-x, i+x])
    A = A.sort((a,b) => a[0] - b[0])
    let lefts = A.map((x)=>x[0]);
    
    let search = function(list, x){
        let beg = 0;
        let end = list.length -1;
        let result = -1;
        while(beg <= end){
            let mid = Math.floor((beg+end)/2)
            if(list[mid] <= x){
                beg = mid + 1;
                result = mid;
            }else{
                end = mid -1;
            }
        }
        return result;
    }
    
    let c = 0;
    for(i = 0; i< A.length; i++){
        let rank = search(lefts, A[i][1])
        c += (rank+1) - i - 1;
        if(c > 10000000) return -1;
    }
    return c;
}

We first sort the left boundarys, and and then in a loop to find each right boundary’s position in it(using binary search), as the loop is O(n), and binary search is O(logn), so the total complexity is O(nlogn)

But do we still have a better solution?

We can consider the discs as events. for each disc, so each disc has 2 events: open event, and close event. And then we can use stack to trace the events. when a disc open, we push a record, when another disc open, we check how many records are there in the stack, that means the intersection, and when a disc close, we pop the stack.

function solution(A) {
    // write your code in JavaScript (Node.js 6.4.0)
    var events = []
    A.forEach(function(x, i){
        events.push({type: 'left', value: i-x})
        events.push({type: 'right', value: i+x})
    })
    events = events.sort(function(a,b){
        if(a.value != b.value) return a.value - b.value;
        else return (a.type == 'right' ? 1 : -1);
    })

    var c= 0;
    var stack = [];
    for(i = 0; i < events.length; i++){
        let event = events[i];
        
        if(event.type == 'left'){
            c += stack.length;
            if(c > 10000000) return -1;
            stack.push(1);
        }else{
            stack.pop()
        }
    }

    return c;
}

Now the complexity is O(n) now.

Sample Markdown

Sat, 01 Oct 2016 00:00:00 +0000

An h1 header

Paragraphs are separated by a blank line.

2nd paragraph. Italic, bold, monospace. Itemized lists look like:

this one
that one
the other one

Note that — not considering the asterisk — the actual text content starts at 4-columns in.

Block quotes are written like so.

They can span multiple paragraphs, if you like.

Use 3 dashes for an em-dash. Use 2 dashes for ranges (ex. “it’s all in chapters 12–14”). Three dots … will be converted to an ellipsis.

An h2 header

Here’s a numbered list:

first item
second item
third item

Note again how the actual text starts at 4 columns in (4 characters from the left side). Here’s a code sample:

# Let me re-iterate ...
for i in 1 .. 10 { do-something(i) }

As you probably guessed, indented 4 spaces. By the way, instead of indenting the block, you can use delimited blocks, if you like:

define foobar() {
    print "Welcome to flavor country!";
}

(which makes copying & pasting easier). You can optionally mark the delimited block for Pandoc to syntax highlight it:

import time
# Quick, count to ten!
for i in range(10):
    # (but not *too* quick)
    time.sleep(0.5)
    print i

An h3 header

Now a nested list:

First, get these ingredients:
- carrots
- celery
- lentils
Boil some water.

Dump everything in the pot and follow this algorithm:

find wooden spoon
uncover pot
stir
cover pot
balance wooden spoon precariously on pot handle
wait 10 minutes
goto first step (or shut off burner when done)

Do not bump wooden spoon or it will fall.

Notice again how text always lines up on 4-space indents (including that last line which continues item 3 above). Here’s a link to a website. Here’s a link to a local doc. Here’s a footnote ¹.

Done.

Footnote text goes here. ↩︎

A Javascript Fibonacci (Generator) Function

Thu, 30 May 2013 00:00:00 +0000

function fibonacci () {
	var f1 = 1;
	fibonacci = function () {
		var f2 = 1;
		fibonacci = function () {
			var f3 = f1 + f2;
			f1 = f2;
			f2 = f3;
			return f3;
		}
		return f2;
	}
	return f1;
}

in ES6(Harmony), the generator keyword: yield will be introduced, a generator function will be like this:

function fibonacci(){
	var fn1 = 1;
	var fn2 = 1;
	while (1){
		var current = fn2;
		fn2 = fn1;
		fn1 = fn1 + current;
		yield current;
	}
}

Asynchronous closure call in loop in Javascript

Wed, 15 May 2013 00:00:00 +0000

Problem: Async closure call in loop may cause mistake, Consider following code

for(var i = 0; i<4; i++){
	setTimeout(function(){
		console.log(i);
	},1000)
}

what will the output be? 0 1 2 3? no, actually it will be 4 4 4 4

The reason is Async, when the closure: function(){console.log(i)} been called, the loop already finished, and the i is been changed to 4, so when the closure been called, all the i is 4

How to solve this problem

we need to avoid the change of the i in the closure, this time, we need use another closure:

for(var i = 0; i<4; i++){
	(function(i){
		setTimeout(function(){
			console.log(i);
		},1000)
	})(i)
}

we make a closure and call it instantly in the loop, so when the closure is been called each time, the i is different, so in the closure inside this closure, the i is different.

Async and Closure are both great features of Javascript, when using them in a correct way, they are very powerful and flexiable. And when you are using them together, you must be careful.

Reference: Closures

Auto start node.js application on linux server

Thu, 09 May 2013 00:00:00 +0000

So we are developing node.js application and we want to keep it runing on server

What we need?

node.js application auto start on linux server boot
node.js application auto restart when it crashed
node.js application auto restart when files change
developer can check the output of node.js application anytime

Easy problem first

The problem 2,3 are easily solved by the nodemon module. This module can auto restart node.js app when it crashed and also when files of the app changed. OK, let’s install it and get it running first:

npm install -g nodemon
cd /path/to/app
nodemon app.js

And we see the outputs of the app, and also we have auto restart feature now.

OK, if we close our terminal (for me, it’s putty), the nodemon process will be termiated as its parent process termiated. so let’s introduce screen to keep the session. to use it:

screen -S nodeapp

And now we are in the new screen session (whose name is nodeapp). run the nodemon command:

nodemon app.js

Now we can press Ctrl+A,D to detach the screen session, and even we close putty, the node is still runing. and later, when we login to the server again,and use:

screen -r nodeapp

to attach into the screen session again.

so we solved problem 2,3,4 now

we make this into a shell script so we can run it anytime

touch ~/nodeapp.sh
chmod u+x ~/nodeapp.sh

add add following shell command into it

#!/bin/bash
cd /home/william
nodemon /path/to/app.js

Auto start on server boot

Next, let’s focus on problem 1. to auto start the node app when server boot, in our situation, we want the server to automaticlly create a screen session, and start nodemon in it.

When search on internet, i found many different solution for this, some one suggest systemd, upstart, and Monit etc.

But consider diffirent linux distro have diffirent boot machanism, it’s a little bit hard to do that.

My solution is a lot easier. and much flexiable. We want:

it auto start a screen session so that the developer can attach to that screen session again to see the outputs.

In most linux distro, we can find the /etc/init.d/rc.local file. this file is designed to be a custom shell script which will run on everytime the server boot. And in most case, that file will source the /etc/rc.local file too, so let’s add our command there.

the command to add is:

su - william -c 'screen -d -m -S nodeapp /home/william/nodeapp.sh'

Let me explain this command, we can split it into 3 parts:

su - william: this command switch into the william user who is the developer of the node.js app
-c '....': this part run the xxxx command once it logged into the developer user
screen -d -m -S nodeapp /home/william/nodeapp.sh: this part start a deattached screen session whose name is nodeapp and run the nodeapp.sh shell script

Difference of module.exports and exports in node.js

Mon, 06 May 2013 00:00:00 +0000

If you are new in node.js, you may get confused by the difference of module.exports and exports in node.js’s module system.

Docs

Following 2 pages in official documents didn’t explain them very well:

Globals Modules

Conclusion

Let me give you the conclusion first:

module.exports is the real object that been used

exports is just a reference

Details

Then let’s explain them in details:

in Javascript, when you use = to assign an object to a variable, it’s passed it’s reference, not value so:

 var a = {name:'a'};
 var b = a;
 b.name = 'b';
 console.log(a.name); // this will output 'b'

in node.js, for each module, there is a module object been automaticlly created. and in this module object, an exports attribute been automatically created too, and assigned a initial value: {}, so in code, it look like this:

module.exports = {};

and then, Nodejs added a convinient way to use this module.exports object

var exports = module.exports;

So that when you want to add a attribute into module.exports object, you can do it like this:

exports.test = function(){
	//do something;
}

What if i want to assign the whole exports object as a function? so you may try it like this way:

exports = function(){
	//do something
}

This will not work. because when you do this, the exports variable is been assigned to another object (a function), and it’s now not the same object as module.exports, you can add following code to check it

module.exports === exports // will return false

and as I mentioned before: module.exports is the real object that been used

so when you want to do make the whole module just been a function. do it like this:

module.exports = function(){
	//do something	
};

Hello World!

Sun, 05 May 2013 00:00:00 +0000

echo 'Hello World!';

module.exports = exports = function(){
	console.log('Hello World!');
};