Skip to content

Latest commit

 

History

History
3961 lines (2641 loc) · 132 KB

README.md

File metadata and controls

3961 lines (2641 loc) · 132 KB

In JS we trust - The best way to learn is by building/coding and teaching. I create the challenges to help my friends learn JavaScript and in return it helps me embrace the language in much deeper level. Feel free to clone, fork and pull.


1. What's the output?
function a(x) {
  x++;
  return function () {
    console.log(++x);
  };
}

a(1)();
a(1)();
a(1)();

let x = a(1);
x();
x();
x();
  • A: 1, 2, 3 and 1, 2, 3
  • B: 3, 3, 3 and 3, 4, 5
  • C: 3, 3, 3 and 1, 2, 3
  • D: 1, 2, 3 and 3, 3, 3
Answer

Answer: B

This question revisits closure - one of the most confusing concepts in JavaScript. Closure allows us to create a stateful function and such a function can access to the variable outside of its scope. In a nutshell, a closure can have access to the global variable (scope), father function scope and its own scope.

We have here, the only one correct answer, 3, 3, 3 and 3, 4, 5 because first we simply call the function a(). It works like a normal function and we have not seen anything so-called stateful yet. In the following code, we declare a variable x and it stores the value of function a(1), that is why we get 3. 4. 5 rather than 3, 3, 3.

This kind of gotcha gives me the feeling of static variable in PHP world.


2. What's the output?
function Name(a, b) {
  this.a = a;
  this.b = b;
}

const me = Name("Vuong", "Nguyen");

console.log(!(a.length - window.a.length));
  • A: undefined
  • B: NaN
  • C: true
  • D: false
Answer

Answer: C

We get true in the console. The tricky part is when we create an object from the constructor function Name but we DO NOT USE new keywork. That makes the variable a global one and get the value "Vuong". Remember that it is actually a property of the global object window (in the browser) or global in the nodejs.

We then get a.length ~ 5 and window.a.length ~ 5 which return 0. !0 returns true.

Imagine what would happen when we create the instance me with the new keywork. That is an interesting inquire!


3. What's the output?
const x = function (...x) {
  let k = (typeof x).length;
  let y = () => "freetut".length;
  let z = { y: y };

  return k - z.y();
};

console.log(Boolean(x()));
  • A: true
  • B: 1
  • C: -1
  • D: false
Answer

Answer: A

The spread operator ...x might help us obtain the parameter in the function in the form of array. Yet, in Javascript the typeof array return "object" rather than "array". It is totally odd if you are coming from PHP.

That is said, we now have the length of the string object which returns 6. z.y() simply returns the length of the string 'freetut' (7).

Be aware that the function x() (in the form of function express or anonymous function (if you are coming from PHP) return -1 when being called and when converted to bool with Boolean(-1) return true instead of false. Noted that Boolean(0) return false.


4. What's the output?
(function js(x) {
  const y = (j) => j * x;

  console.log(y(s()));

  function s() {
    return j();
  }

  function j() {
    return x ** x;
  }
})(3);
  • A: undefined
  • B: 18
  • C: 81
  • D: 12
Answer

Answer: C

The function js() can be automatically executed without calling it and known as IIFE (Immediately Invoked Function Expression). Noted the parameter x of the function js is actuallly passed with the value 3.

The value return of the function is y(s())), meaning calling three other functions y(), s() and j() because the function s() returns j().

j() returns 3^3 = 27 so that s() returns 27.

y(s()) means y(27) which returns 27*3 = 81.

Note that we can call declare function BEFORE the function is actually declared but not with expression function.


5. What's the output?
var tip = 100;

(function () {
  console.log("I have $" + husband());

  function wife() {
    return tip * 2;
  }

  function husband() {
    return wife() / 2;
  }

  var tip = 10;
})();
  • A: "I have $10";
  • B: "I have $100";
  • C: "I have $50";
  • D: "I have $NaN";
Answer

Answer: D

We have here an IIFE (Immediately Invoked Function Expression). It means we do not have to call it but it will be excuted automatically when declared. The flow is as: husband() returns wife()/2 and wife() returns tip*2.

We might think that tip = 100 because it is a global variable when declaring with var keyword. However, it is actually undefined because we also have var tip = 10 INSIDE the function. As the variable tip is hoisted with default value undefined, the final result would be D. We know that undefined returns NaN when we try to divide to 2 or multiple with 2.

If we do not re-declare var tip = 10; at the end of the function, we will definately get B.

JS is fun, right?


6. What's the output?
const js = { language: "loosely type", label: "difficult" };

const edu = { ...js, level: "PhD" };

const newbie = edu;

delete edu.language;

console.log(Object.keys(newbie).length);
  • A: 2;
  • B: 3;
  • C: 4;
  • D: 5;
Answer

Answer: A

This challenge revises the ES6's feature regarding spread operator ... Spread operator is quite useful for retrieving parameter in function, to unite or combine object and array in JavaScript. PHP also has this feature.

In the variable edu, we use ...js (spread operator here) to combine both objects into one. It works in the same way with array.

Then we declare another variable named newbie. IMPORTANT note: By declaring the variable like that, both variables point to the SAME POSITION in the memory. We may have known something like $a = &$b in PHP, which let both varibles work in the same way. We might have known about pass by reference in the case.

Then we have 2 as edu.language is deleted. Both objects now have only two elements.

Now is time to think about coping an object in JS either shallow or deep one.


7. What's the output?
var candidate = {
  name: "Vuong",
  age: 30,
};

var job = {
  frontend: "Vuejs or Reactjs",
  backend: "PHP and Laravel",
  city: "Auckland",
};

class Combine {
  static get() {
    return Object.assign(candidate, job);
  }

  static count() {
    return Object.keys(this.get()).length;
  }
}

console.log(Combine.count());
  • A: 5;
  • B: 6;
  • C: 7;
  • D: 8;
Answer

Answer: A

The buit-in method Object.assign(candidate, job) merges the two objects candidate and job into one object. Then the method Object.keys counts the number of key in the object.

Note that two methods get() and count() are defined as static, so they need to be called statically using Class.staticmethod() syntax. Then the final object get 5 elements.


8. What's the output?
var x = 1;

(() => {
  x += 1;
  ++x;
})();
((y) => {
  x += y;
  x = x % y;
})(2);
(() => (x += x))();
(() => (x *= x))();

console.log(x);
  • A: 4;
  • B: 50;
  • C: 2;
  • D: 10;
Answer

Answer: A

Initially x is declared with the value 1. In the first IIFE function, there are two operations. First x becomes 2 and then 3.

In the second IIFE function, x = x + y then the current value is 5. In the second operation, it returns only 1 as it undergoes 5%2.

In the third and fouth IIFE functions, we get 2 x = x + x and then 4 x = x * x. It is more than simple.


9. What's the output?
$var = 10;

$f = function($let) use ($var) {
    return ++$let + $var;
};

$var = 15;
echo $f(10);
var x = 10;

const f = (l) => ++l + x;

x = 15;
console.log(f(10));
  • A: 26 and 26;
  • B: 21 and 21;
  • C: 21 and 26;
  • D: 26 and 21;
Answer

Answer: C

This question illustrates the diffences between PHP and JavaScript when handling closure. In the first snippet, we declare a closure with the keyword use. Closure in PHP is simply an anonymous function and the data is passed to the function using the keyword use. Otherwise, it is called as lambda when we do not use the keyword use. You can check the result of the snippet here https://3v4l.org/PSeMY. PHP closure only accepts the value of the variable BEFORE the closure is defined, no matter where it is called. As such, $var is 10 rather than 15.

On the contrary, JavaScript treats the variable a bit different when it is passed to anonymous function. We do not have to use the keyword use here to pass variable to the closure. The variable x in the second snippet is updated before the closure is called, then we get 26.

Note that in PHP 7.4, we have arrow function and we then do not have to use the keyword use to pass the variable to function. Another way to call a global ariable inside a function in PHP is to use the keyword global or employ the built-in GLOBAL variable $GLOBALS.


10. What's the output?
let x = {};
let y = {};
let z = x;

console.log(x == y);
console.log(x === y);
console.log(x == z);
console.log(x === z);
  • A: true true true true;
  • B: false false false false;
  • C: true true false false;
  • D: false false true true;
Answer

Answer: D

Technically, x and y have the same value. Both are empty objects. However, we do not use the value to compare objects.

z is x are two objects referring to the same memory position. In JavaScript, array and object are passed by reference. x and z therefore return true when being compared.


11. What's the output?
console.log("hello");

setTimeout(() => console.log("world"), 0);

console.log("hi");
  • A: "hello" -> "world" -> "hi"
  • B: "hello" -> "hi" -> "world"
  • C: "hi" -> "world" -> "hello"
  • D: "hi" -> "hello" -> "world"
Answer

Answer: B

Given that the function setTimeout() will be kept in the task queue before jumping back to stack, "hello" and "hi" will be printed first, then A is incorrect. That is also the case of the answers C and D.

No matter how many seconds you set to the setTimeout() function, it will run after synchronous code. So we will get "hello" first as it is put into the call stack first. Though the setTimeout() is then being put into the call stack, it will subsequently offload to web API (or Node API) and then being called when other synchronous codes are cleared. It means we then get "hi" and finally "world".

So B is the correct answer.

Credit: @kaitoubg (voz) for your suggestion regarding the timeout throttled by which I have decided to alter the question slightly. It will ensure that readers will not get confused as the previous code might bring out different results when tested on other browsers or environments. The main point of the question is about the discrepancy between the synchronous code and asynchronous code when using setTimeout..


12. What's the output?
String.prototype.lengthy = () => {
  console.log("hello");
};

let x = { name: "Vuong" };

delete x;

x.name.lengthy();
  • A: "Vuong";
  • B: "hello";
  • C: "undefined"
  • D: "ReferenceError"
Answer

Answer: B

String.prototype.someThing = function () {} is the common way to define a new built-in method for String. We can do the same thing with Array, Object or FunctionName where FunctionName is the function designed by ourself.

That is not challenging to realise that "string".lengthy() always returns hello. Yet, the tricky part lies in the delete object where we might think that this expression will entirely delete the object. That is not the case as delete is used to delete the property of the object only. It does not delete the object. Then we get hello rather than ReferenceError.

Note that if we declare object without let, const or var, we then have a global object. delete objectName then return true. Otherwise, it always returns false.


13. What's the output?
let x = {};

x.__proto__.hi = 10;

Object.prototype.hi = ++x.hi;

console.log(x.hi + Object.keys(x).length);
  • A: 10
  • B: 11
  • C: 12
  • D: NaN
Answer

Answer: C

First we have an empty object x, then we add another property hi for x with x.__proto__.hi. Note this is equivalent to Object.prototype.hi = 10 and we are adding to the father object Object the property hi. It means every single object will inherit this propety. The property hi becomes a shared one. Say now we declare a new object such as let y = {}, y now has a propery hi inherited from the father Object. Put it simply x.__proto__ === Object.prototype returns true.

Then we overwrite the property hi with a new value 11. Last we have 11 + 1 = 12. x has one property and x.hi returns 11.

Updated (July 27th 2021). If you write Object.prototype.hi = 11; instead of Object.prototype.hi = ++x.hi; as written in the code above, then Object.keys(x) will return an empty array as Object.keys(object) only returns the property of the object itself, not the inherited ones. It means the final result will be 11 rather than 12. For some reason, the code ``Object.prototype.hi = ++x.hi;will create a property for the objectx` itself and then `Object.keys(x)` gives us the array `["hi"]`.

Yet, if you run console.log(x.hasOwnProperty("hi")) it still returns false. By the way, when you deliberately add a property for x such as x.test = "testing", then console.log(x.hasOwnProperty("test")) returns true.


14. What's the output?
const array = (a) => {
  let length = a.length;
  delete a[length - 1];
  return a.length;
};

console.log(array([1, 2, 3, 4]));

const object = (obj) => {
  let key = Object.keys(obj);
  let length = key.length;
  delete obj[key[length - 1]];

  return Object.keys(obj).length;
};

console.log(object({ 1: 2, 2: 3, 3: 4, 4: 5 }));

const setPropNull = (obj) => {
  let key = Object.keys(obj);
  let length = key.length;
  obj[key[length - 1]] = null;

  return Object.keys(obj).length;
};

console.log(setPropNull({ 1: 2, 2: 3, 3: 4, 4: 5 }));
  • A: 333
  • B: 444
  • C: 434
  • D: 343
Answer

Answer: C

This question examines how the delete operator works in JavaScript. In short, it does nothing when we write delete someObject or delete someArray. It nonetheless completely deletes and removes a property of an object when writing something like delete someObject.someProperty. In the case of array, when we write delete someArray[keyNumber], it only removes the value of the index, keep the index intact and the new value is now set to undefined. For that reason, in the code first snippet, we get (the length) 4 elements as in the original array but only 3 properties left in the object passed when the function object() is called, as in the second snippet.

The third snippet gives us 4 as declaring an object's propery to either null or undefined does not completely remove the property. The key is intact. So the length of the object is immutable.

For those who are familiar with PHP, we have unset($someArray[index]) that remove the array element, both key and value. When print_r the array, we might not see the key and value that have been unset. However, when we push (using array_push($someArray, $someValue)) a new element in that array, we might see that the previous key is still kept, but no value and not being displayed. That is something you should be aware of. Have a look at https://3v4l.org/7C3Nf


15. What's the output?
var a = [1, 2, 3];
var b = [1, 2, 3];

var c = [1, 2, 3];
var d = c;

var e = [1, 2, 3];
var f = e.slice();

console.log(a === b);
console.log(c === d);
console.log(e === f);
  • A: true true true
  • B: false false true
  • C: true true false
  • D: false true false
Answer

Answer: D

a and b returns false because they point to different memory location even though the values are the same. If you are coming from PHP world, then it will return true obviously when we compare either value or value + type. Check it out: https://3v4l.org/IjaOs.

In JavaScript, value is passed by reference in case of array and object. Hence in the second case, d is the copy of c but they both point to the same memory position. Everything changes in c will result in the change in d. In PHP, we might have $a = &$b;, working in the similar way.

The third one gives us a hint to copy an array in JavaScript using slice() method. Now we have f, which is the copy of e but they point to different memory locations, thus they have different "life". We get false accordingly when they are being compared.


16. What's the output?
var languages = {
  name: ["elixir", "golang", "js", "php", { name: "feature" }],
  feature: "awesome",
};

let flag = languages.hasOwnProperty(Object.values(languages)[0][4].name);

(() => {
  if (flag !== false) {
    console.log(
      Object.getOwnPropertyNames(languages)[0].length <<
        Object.keys(languages)[0].length
    );
  } else {
    console.log(
      Object.getOwnPropertyNames(languages)[1].length <<
        Object.keys(languages)[1].length
    );
  }
})();
  • A: 8
  • B: NaN
  • C: 64
  • D: 12
Answer

Answer: 64

The code snippet is quite tricky as it has a couple of different built-in methods handling object in JavaScript. For example, both Object.keys and Object.getOwnPropertyNames are used even thought they are quite similar except that the latter can return non-enumerable properties. You might want to have a look at this thoroughly written reference https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Object/getOwnPropertyNames

Object.values and Object.keys return the property value and property name of the object, respectively. That is nothing new. object.hasOwnProperty('propertyName') returns a boolean confirming whether a property exists or not.

We have flag true because Object.values(languages)[0][4].name returns feature, which is also the name of the property.

Then we have 4 << 4 in the if-else flow that returns the bitwise value, equivalent to 4*2^4 ~ 4*16 ~ 64.


17. What's the output?
var player = {
  name: "Ronaldo",
  age: 34,
  getAge: function () {
    return ++this.age - this.name.length;
  },
};

function score(greeting, year) {
  console.log(
    greeting + " " + this.name + `! You were born in  ${year - this.getAge()}`
  );
}

window.window.window.score.call(window.window.window.player, "Kiora", 2019);

score.apply(player, ["Kiora", 2009]);

const helloRonaldo = window.score.bind(window.player, "Kiora", 2029);

helloRonaldo();
  • A: "Kiora Ronaldo! You were born in 1985", "Kiora Ronaldo! You were born in 1985", "Kiora Ronaldo! You were born in 1985"
  • B: "Kiora Ronaldo! You were born in 1991", "Kiora Ronaldo! You were born in 1991", "Kiora Ronaldo! You were born in 1999"
  • C: "Kiora Ronaldo! You were born in 1991", NaN, "Kiora Ronaldo! You were born in 1980"
  • D: "Kiora Ronaldo! You were born in 1991", "Kiora Ronaldo! You were born in 1980", "Kiora Ronaldo! You were born in 1999"
Answer

Answer: D

We can use call(), apply() and bind() to apply a function to any object. At first sight, it seems that three functions do the same thing. Yet there are some situations where they are differently employed to handle respective contexts or solve particular problems.

Of the three, only bind() can be executed after binding. We can create a variable to store the result as helloRonaldo() in the code snippet above. apply() and call() will bind and execute the function at the same time. apply() hints us a ~ array where we need to pass an array as parameter. call() hints us c or comma where we pass parameters with a comma. You might want to have a look at this post https://stackoverflow.com/questions/15455009/javascript-call-apply-vs-bind

Note that window.window.window.score or window.score or simply score do the same thing. It points to the score() function in the global scope.

The correct anwser is D. The score() and getAge() functions are nothing special. The only tricky part is that this.age is incremented each time.


18. What's the output?
var ronaldo = { age: 34 };

var messi = { age: 32 };

function score(year, tr, t) {
  if (typeof tr === "function" && typeof t === "function") {
    console.log(`You score ${tr(year, t(this.age))} times`);
  }
}

const transform = (x, y) => x - y;

const title = (x) => ++x + x++;

const helloRonaldo = score.bind(ronaldo, 2029, transform, title);

helloRonaldo();

const helloMessi = score.bind(messi, 2029, transform, title);

helloMessi();
  • A: "You score 1989 times" and "You score 1963 times"
  • B: "You score 1959 times" and "You score 1989 times"
  • C: "You score 1989 times" and "You score 1953 times"
  • D: "You score 1959 times" and "You score 1963 times"
Answer

Answer: D

bind() allows us to bind a function declared with any object. Here we bind score() and both ronaldo and messi.

In score() we pass three parameters year, tr and t in which both tr and t are function. They handle simple things as defined afterwards.

When we bind score() with ronaldo and messi, we pass three parameters as declared in the score() function wherein transform and title are functions.


19. What's the output?
var person = {};

Object.defineProperties(person, {
  name: {
    value: "Vuong",
    enumerable: true,
  },
  job: {
    value: "developer",
    enumerable: true,
  },
  studying: {
    value: "PhD",
    enumerable: true,
  },
  money: {
    value: "NZD",
    enumerable: false,
  },
});

class Evaluate {
  static checkFlag(obj) {
    return Object.getOwnPropertyNames(obj) > Object.keys(obj)
      ? Object.getOwnPropertyNames(obj)
      : Object.keys(obj);
  }
}

const flag = Evaluate.checkFlag(person);

console.log(flag.length);
  • A: 1
  • B: 2
  • C: 3
  • D: 4
Answer

Answer: D

Object.keys(obj) is almost identical to Object.getOwnPropertyNames(obj) except the fact that the latter returns any type of object's property regardless of enumerable. By default enumerable is true when creating object. Using Object.defineProperties or Object.defineProperty we can manually set this option to false.

As such the object person will get 3 usingObject.keys(obj)but 4 with Object.getOwnPropertyNames(obj). In short Object.keys(obj) only returns the property setting the enumerable as true.


20. What's the output?
const id = 10;

const getID = (...id) => {
  id(id);

  function id(id) {
    console.log(typeof id);
  }
};

getID(id);
  • A: ReferenceError
  • B: 10
  • C: undefined
  • D: 'function'
Answer

Answer: D

When declaring a function inside another function, we are working with Closure in JavaScript. Note that if a function is declared as normal (rather than function expression), it is hoisted. We might see several id in the code snippet above but in fact, some of them does nothing.

The result of the code depending on the operator typeof id, which is function. So id in this operation is the id() function.


21. What's the output?
var book1 = {
  name: "Name of the rose",
  getName: function () {
    console.log(this.name);
  },
};

var book2 = {
  name: { value: "Harry Potter" },
};

var bookCollection = Object.create(book1, book2);

bookCollection.getName();
  • A: 'Harry Potter'
  • B: 'Name of the rose'
  • C: ReferenceError
  • D: Object object
Answer

Answer: A

Object.create allows us to create an object which is based on another object. If we do not pass the second parameter - book2 in this case - the name property of the object bookCollection will be Name of the rose inherited from the book1. It means we can provide additional properties when declaring object with Object.create.

bookCollection has its own property name and another one inherited from book1. In this case its own property name will show up as it has higher priority. That is why we get 'Harry Potter'.


22. What's the output?
(() => {
  const a = Object.create({});

  const b = Object.create(null);

  let f1 = a.hasOwnProperty("toString");

  let f2 = "toString" in b;

  let result =
    f1 === false && f2 === false
      ? console.log((typeof a.toString()).length)
      : console.log(b.toString());
})();
  • A: ReferenceError
  • B: undefined
  • C: 0
  • D: 6
Answer

Answer: D

The two objects a and b are created using Object.create() operator. There is a bit of difference between them as a inherits from Object prototype but b is totally empty when we pass the null paramater. Yet hasOwnProperty('toString') always returns false neither a nor b given that toString() is not defined inside these objects. The method however is still available as it is inherited from Object prototype.

Both f1 and f2 return false. Note that we use object.hasOwnProperty('key') and ('key' in object) to check the availability of a key in an object. There is a bit difference between the two as the latter also returns the key inherited. You might want to have a look here: https://stackoverflow.com/questions/455338/how-do-i-check-if-an-object-has-a-key-in-javascript

Then typeof a.toString() returns string, which gives us 6 with the .length property.

If the syntax is odd to you, you might look for 'self-invoking function' and 'arrow function' in JavaScript.


23. What's the output?
let promise = new Promise((rs, rj) => {
  setTimeout(() => rs(4), 0);

  Promise.resolve(console.log(3));

  console.log(2);
});

promise
  .then((rs) => {
    console.log(rs ? rs ** rs : rs);
    return rs;
  })
  .then((rs) => console.log(rs == 256 ? rs : rs * rs));
  • A: 3, 2, 256, 256
  • B: 3, 2, 256, 16
  • C: 256, 16, 3, 2
  • D: 16, 256, 3, 2
Answer

Answer: B

We first declare a promise-based code with let and then call it. Given that setTimeout() is an asynchronous action, it will run last even the time is set to 0 in setTimeout(() => rs(4), 0);. Although Promise.resolve(console.log(3)) also returns a promise but it is a Microtasks, then it has a higher priority than Tasks as set by setTimeout(). You might want to have a look at this post https://jakearchibald.com/2015/tasks-microtasks-queues-and-schedules/.

In .then() we chain the result so that we have 4^4 in the first then() and 4*4 in the second then(). Note that return rs returns the original value.


24. What's the output?
async function f() {
  let promise = new Promise((resolve, reject) => {
    setTimeout(() => resolve("done!"), 0);
  });

  setTimeout(() => console.log("world"), 0);

  console.log(await promise);

  console.log("hello");
}

f(setTimeout(() => console.log("kiora"), 0));
  • A: ReferenceError
  • B: done, hello, world
  • C: hello, done, world
  • D: kiora, done, hello, world
Answer

Answer: D

Though we do not declare any paramater for the function f(), we pass setTimeout(()=>console.log("kiora"),0) when call it. We therefore get 'kiora' first.

Given that the variable promise returns a solved promise and it is called with the keyword await, JavaScript will 'pause' at this line console.log(await promise); till the result is resolved. That is why we get "done" at the next result.

Why we do not get "world" or "hello" at the second ? As JavaScript "pauses" at the line with await keyword, we cannot get "hello" as usual (note that whenever we call setTimeout(), this function will run last because it is an asynchronous task operator), whereas setTimeout(()=> console.log("world"), 0); should always run last.

Here we might see a bit of difference when employing await keyword before asynchronous operator (in this case, we use setTimeout() as an example) or when call the function/operator without it.


25. What's the output?
function name() {
  return new Promise((resolve) => {
    setTimeout(() => {
      resolve("New Zealand");
    }, 10);
  });
}

function fruit() {
  return new Promise((resolve) => {
    setTimeout(() => {
      resolve("Kiwi");
    }, 20);
  });
}

(async function countryandfruit() {
  const getName = await name();
  const getFruit = await fruit();

  console.log(`Kiora: ${getName} ${getFruit}`);
})();

(async function fruitandcountry() {
  const [getName, getFruit] = await Promise.all([name(), fruit()]);

  console.log(`Hello: ${getName} ${getFruit}`);
})();
  • A: Null
  • B: Kiora
  • C: "Hello: New Zealand Kiwi" -> "Kiora: New Zealand Kiwi"
  • D: "Kiora: New Zealand Kiwi" -> "Hello: New Zealand Kiwi"
Answer

Answer: C

Both countryandfruit and fruitandcountry are self invoking functions. Both are declared with the keyword async, it means the code inside will run step by step. It helps us control the flow of data much more concise as compared to Promise-based operator or callback way.

The first function returns "Kiora: New Zealand Kiwi" and the second one ouputs "Hello: New Zealand Kiwi". We might think that the order will be the same but actually the order of the result is reversed because the function with await keyword will run step by step rather than in in parallel as Promise.all. It means fruitandcountry will run faster than countryandfruit.

You might want to have a look at the difference between the two at https://alligator.io/js/async-functions/


26. What's the output?
class MySort {
  constructor(object) {
    this.object = object;
  }

  getSort() {
    return Object.entries(this.object)[0][1].sort()[
      Object.values(this.object).length
    ];
  }
}

const object = {
  month: ["July", "September", "January", "December"],
};

const sortMe = new MySort(object);

console.log(sortMe.getSort());
  • A: July
  • B: September
  • C: January
  • D: December
Answer

Answer: C

Object.entries returns an array consisting of both key and value from an object while Object.values retuns an array of the values of object and Object.keys gives us an array of keys of the object. As such, Object.entries(object) in the code snippet above gives us a nested array with just one element in which the values are put in another nested array like that [["month", ["July", "September", "January", "December"]]].

For that reason, Object.entries(this.object)[0][1].sort() will actually sort the value array and return a new order as "December" -> "January" -> "July" -> "September". Hence, when we get the element with the index given by [Object.values(this.object).length] we get January because [Object.values(this.object).length] give us 1 (the length of the array given by Object.values);


27. What's the output?
const flag = [] !== !!!!![];

let f = () => {};

console.log((typeof f()).length + flag.toString().length);
  • A: NaN
  • B: 12
  • C: 13
  • D: 14
Answer

Answer: C

Comparing two arrays or two objects in JavaScript always return false because both are passed by reference, unlike primitive types such as string, number or boolean. That is why comparing [] and [] using either == or === returns false. The weird part is the !==!!!!! which is equivalent to !==, nothing special. So the flag is true.

In the expression function f(), we use arrow function here but and {} is a part of the function rather than an object. In case you want to return an object, you have to write as let f = () => ({}) or simply using normal way to define function. With the keyword return, we can easily catch the content of the function when using normal way to define function.

Thus, the typeof f() returns undefined rathern object. We then get the length 9 and the flag (true) becomes 'true' (a string, by using toString() function), which returns 3 with the property length. We finally get 13.


28. What's the output?
(function (a, b, c) {
  arguments[2] = (typeof arguments).length;
  c > 10 ? console.log(c) : console.log(++c);
})(1, 2, 3);
  • A: 4
  • B: 5
  • C: 6
  • D: 7
Answer

Answer: D

We have a self-invoking function with three parameters declared. Note that arguments inside a function returns an object consisting of the parameters of the function.

The key part here is that when we assign a value to that array (it is array-like, as mentioned above) (or any element), the function will use that value rather than the value from the parameter we pass to it when calling the function. Hence, c will be (typeof arguments).length; (6) rather than 3.

As c has a new value of 6, it is definitely less than 10, so we get the final result console.log(++c), which returns 7.

Note that arguments is not available on arrow functions. See more detailed here https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Functions/arguments

From ES6 onwards, it is recommended to use ...restParameter given that it is a true array. It means you can manipulate the parameter with native JavaScript functions such as map, reduce or filter.

For PHP developer, we have func_get_args() in PHP that does the same thing, but it will not override the value passed. Check it by yourself at https://3v4l.org/dMfhW


29. What's the output?
class Calculator {
  constructor(a, b) {
    this.a = a;
    this.b = b;
  }
  static getFlag() {
    return new Array(this.a).length == new Array(this.b).toString().length;
  }

  getValue() {
    return Calculator.getFlag() ? typeof this.a : typeof new Number(this.b);
  }
}

const me = new Calculator(5, 5);

console.log(me.getValue());
  • A: NaN
  • B: "string"
  • C: "object"
  • D: "number"
Answer

Answer: C

We have a class named Calculator. When declaring a new instance of the object, we pass two parameters a and b. These two parameters have the same value but new Array(this.a).length is totally different from new Array(this.b).toString().length because the latter returns a string ",,,," meaning the length 4 while the former returns the length of an array and we therefore get 5.

For that reason getFlag() returns false. In getValue() we get typeof new Number(this.b); which returns object. That is a bit different from typeof b, which returns number.


30. What's the output?
var name = "Auckland";

const nz = {
  name: "Kiwi",

  callMe: function () {
    return this.name;
  },
};

let me = nz.callMe;

let she = nz.callMe.bind(nz);

let result = me() === nz.callMe() ? she() : `${me()} ${she()}`;

console.log(result);
  • A: undefined
  • B: "Auckland"
  • C: "Kiwi"
  • D: "Auckland Kiwi"
Answer

Answer: D

The key point in this question involves the keyword this in JavaScript. We have a simple object that contains one method and one string property name.

First, it is important to write down is that let me = nz.callMe; and then call me() is totally different from directly calling nz.callMe(). If we assign a variable to a method delared inside an object, this in that method will behave differently (when we call the variable as a method and when dirrectly call that method). In particular, in the first case, this is the window object while in the second one, this inside the function still points to property name in the object nz. It means me() returns "Auckland" while nz.callMe returns "Kiwi".

Then result will return false and we get the final output value ${me()} ${she()}. Why she() is different from me()? You might easily guess that she still bind to the object nz rather than window object as in me().


31. What's the output?
const club = {
  name: "Juventus",
  player: ["Ronaldo"],
  showMePlayer: function () {
    this.player.map(function (thename) {
      console.log(this.name.length);
    }, this);
  },
  showMe: function () {
    this.player.forEach(
      function (thename) {
        console.log(this.name.length);
      }.bind(this)
    );
  },
  show: function () {
    const self = this;
    this.player.map(function (thename) {
      console.log(self.name.length);
    });
  },
  Me: function () {
    this.player.map(function (thename) {
      console.log(this.name.length);
    });
  },
};

club.showMePlayer();
club.showMe();
club.show();
club.Me();
  • A: 8 - 8 - 8 - 8
  • B: "Juventus" - "Juventus" - "Juventus" - "Juventus"
  • C: "Ronaldo" - "Ronaldo" - "Ronaldo" - "Ronaldo"
  • D: 8 - 8 - 8 - 0
Answer

Answer: D

The code snippet above is not a big challenge for you I guess. It simply gives you an example of this in different contexts when we declare an anonymous function inside a method of an object. The three first methods are common ways to handle this using this as second parameter in map(), by using bind(this) in forEach (or map()) or by that = thistechnique (you might see people use self = this rather than that= this).

The last method Me() will cause unexpected result because this.name does not bind to the object club. Note that you might get another result when testing the code on jsbin.com. On Chrome and Firefox, we get 0.

For further information, kindly have a look at http://speakingjs.com/es5/ch17.html#_pitfall_losing_this_when_extracting_a_method


32. What's the output?
((...a) => {
  const b = ["javascript", "new zealand"];

  const c = [...a, typeof a, ...b, "kiwi"];

  console.log(c.length + c[0].length);
})(new Array(10));
  • A: 5
  • B: 10
  • C: 15
  • D: 20
Answer

Answer: C

... can be used in two ways in JavaScript (and PHP) as either spread operator or rest parameter. You might have to check the following article about the two. They are the same as three dots, but the way they are employed vary considerably between the two. https://javascript.info/rest-parameters-spread-operator

We see both spread operator and rest parameter in the code snippet above. First the parameter (...a) in the self-invoking function is of course a rest parameter while the constant c we see the spread operator. In the former case, it simply means that you can pass to the function as many parameter as you want. Note that the typeof a in this case is object even though it is a native array in JavaScript. (I means native array because you might think about array-like if we use arguments. Please have a look at the question 28 or this link https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Functions/arguments).

Spread operator as in the constant c allows us to combine array. So ...a in the code above is rest parameter when it is used as function parameter but in this case it is the syntax of spread operator.

Finally, we get c with 5 elements (...a is a nested array, so the length is 1) but the first element has 10 child elements (when we pass to the function new Array(10)). The length of both then returns 15.


33. What's the output?
function Kiora(name, ...career) {
  this.name = name;

  return Array.isArray(career) === true && typeof career === "object" ? {} : "";
}

var student = new Kiora("Vuong");

console.log(student.name);
  • A: "Vuong"
  • B: undefined
  • C: ErrorReference
  • D: false
Answer

Answer: B

We have a function constructor Kiora (written with a capital letter, but that is optional) that can be used to create object, as the student object in the code above. In the function, we have two parameters with the second one is actually a rest parameter. The typeof operator is object but if we check with Array.isArray(array) it also returns true.

For that reason, Array.isArray(career) === true && typeof career === "object" returns true. Hence the return operator finally returns an object {}.

You might be surprised when console.log(student.name); outputs undefined given that the constructor function Kiora returns an object. Otherwise, we might simply get the value name.


34. What's the output?
class Filter {
  constructor(element) {
    this.element = element;
  }
  filter() {
    return this.type() === "object" ? this.element[0].name : "hello";
  }

  type() {
    return typeof this.element;
  }
}

let countries = [
  { name: "New Zealand", isdeveloped: true },
  { name: "Vietnam", isdeveloped: false },
];

let x = new Filter(countries);

const filter = countries.filter((item) => {
  return !item.isdeveloped;
});

console.log(x.filter().length + filter[0].name.length);
  • A: 15
  • B: 16
  • C: 17
  • D: 18
Answer

Answer: D

Apologize that the code snippet is a bit longer than usual. But actually it is not really challenging as you might think. You can easily get the correct result after spending a little of time to debug.

First we declare a class that has two methods. The first method filter() will returns the first element of the array (of the propterty element) or simply returns hello depending on the type() method. We know that typeof of array will return object so the filter() method return this.element[0].name.

Try to make you feel confused, we then call the built-in filter() method. This native method returns a new array depending on the condition we pass to the call-back function. Note that !item.isdeveloped means false. It means we get Vietnam.

Finally we get New Zealand.length and Vietnam.length, which in total returns 18.


35. What's the output?
async function abc() {
  console.log(8);

  await Promise.resolve(2).then(console.log);

  console.log(3);
}

setTimeout(() => {
  console.log(1);
}, 0);

abc();

queueMicrotask(() => {
  console.log(0);
});

Promise.resolve(4).then(console.log);

console.log(6);
  • A: 6 - 8 - 3 - 0 - 4 - 2 - 1
  • B: 8 - 2 - 3 - 0 - 4 - 6 - 1
  • C: 6 - 8 - 2 - 0 - 4 - 3 - 1
  • D: 8 - 6 - 2 - 0 - 4 - 3 - 1
Answer

Answer: D

D is correct anwser. The order of the asynchronous code's output depends on the MicroTask or MacroTask. MicroTask has a higher priority. Note that the synchronous code always be executed before asynchronous code. So in essense, we have the order as follows:

  1) synchronous code
  2) microtask code (promise, queueMicrotask)
  3) macrotask code (setTimeout, setInterval)

Be awared that in Nodejs environment, we also have process.nextTick(callback) which has the highest priority but we dont have it in this code.

So, first callback in the setTimeout() will be executed last given that this is a MacroTask. That is why we got 1 last.

Second, the function abc() is called next. Then we have 8 printed out in the console first. As the next line of code inside that function is an asynchrnous code with the keyword "await", we then console.log(6) as Promise.resolve(4).then(console.log) is an asynchrnous code. That is why we got 6.

Now is the time for Promise.resolve(2), so we get 2. At this point, you might have some sort of confusion. What will happend if we do not pass the keyword "await" before Promise.resolve(2) ?

As we have await, the code will be blocked here. Then what? We get 0 and 4 not 3. Promise and queueMicrotask are both microtask and they are already to run before console.log(3). The reason is that microtask queue need to be emptied before any other codes can be called in the callstack.

In the next step, we get 3 and the last one is 1.

What would happend if we do not have the await keyword? Then the order of the output will be 8 - 3 - 6 - 2 - 0 - 4 -1.

36. What's the output?
function myAccount(money) {
  let myMoney = money;

  return {
    status: function () {
      return `You have $ ${myMoney} in your account`;
    },
    dePoSit: function (amount) {
      myMoney = myMoney + amount;
    },
    withDraw: function (amount) {
      if (amount > myMoney) {
        return `You cannot withdraw money now`;
      }
      myMoney = myMoney - amount;
    },
  };
}

const vuong = myAccount(1000);

vuong.withDraw(500);

vuong.withDraw(200);

vuong.dePoSit(100);

vuong.withDraw(50);

console.log(vuong.status());
  • A: "You have $ 950 in your account"
  • B: "You have $ 1000 in your account"
  • C: "You have $ 550 in your account"
  • D: "You have $ 350 in your account"
Answer

Answer: D

As the "state" of the data is preserved each time we call dePoSit() or withDraw(), hence we get $350 after all.

Noted that that is a kind of "factory" function with "preload" data. You might think about another object when pass to myAccount(somedata); some other data. That is a really helpful way to create multiple objects from a factory function.

37. What's the output?
const hoccoban = {
  x: "youtube.com/hoccoban".length,
  getMe() {
    const inner = function () {
      console.log(++this.x);
    };
    inner.bind(this)();
  },
};

hoccoban.getMe();
  • A: 20
  • B: 21
  • C: 22
  • D: 23
Answer

Answer: B

We get 21. First "youtube.com/hoccoban" returns 20 as we are using the property length of the string. Then it is being added one more value in ++this.x. The question here seems trivial but it is actually not. There is a crucial note we should keep in mind is that console.log(++this.x) will not work as x is undefined when it is called outside of the object.

We can solve the problem with this in this case by using arrow function in the inner so that is can become something like const inner = () => {} as the arrow function does not actually have this. It will automatically look around and call the available object when the function is executed.

The second solution is that we can somehow "bypass" the tricky this by using that/this solution. We just need to declare a new variable const that = this inside getMe() and before declaring inner function. That is a quite common practice.

The third solution is to take advantage of call(), bind() and apply() which are native methods of function (yes, function is also an object in JavaScript). In this case, we implement bind(this) to "bind" the function and the object so that this can actually point to the object when the function is executed. Note that bind() cannot be instantlly executed so that we need to add () after we bridge the function and the object. If we replace bind() with call(), then we do not need to pass () as in the above example. So inner.bind(this)(); will become inner.call(this);. They are technically equal. In practice, we tend to create a new variable to get the result from the binding of the function and the object.

38. What's the output?
function* hocCoBan() {
  yield "js.edu.vn";
  yield "youtube.com/hoccoban";
  yield "Vuong Nguyen";
}

let data = hocCoBan();

console.log((typeof data).length + data.next().value.length);
  • A: NaN
  • B: 10
  • C: Error
  • D: 15
Answer

Answer: D

First, take a closer look at the function. It has a asterisk (*) next to the keyword "function". We do not have return keyword inside the function itself. What is going on here?

It you have already known about generator, then this code snippet is not a big deal at all. We do not use generator very often, but this native JavaScript feature is the basis for async/await function, which is supported in ES7 that allows us to handle the flow of asynchronous code much easily.

The operator typeof data will return object rather than function, which is the same case with typeof hocCoBan(). Of course, typeof hocCoBan still returns function. But it is actually a normal function. Basically, we get 6 in the operator (typeof data).length.

Then data.next() calls the the built-in method next() which will output the value in the first yield, which is declared in the function. Then we get the length 9 with the string js.edu.vn.

After all, we get 15. Not that understanding generator is quite important if you really want to understand async/await function.

39. What's the output?
const a = [1, 2, "chó", 3, 1, "chó", "mèo", 3];

const b = [...new Set(a)];

b.length = "chó".length;

console.log(b);
  • A: 4
  • B: [1, 2, "chó", 3, "mèo"]
  • C: [1, 2, "chó", "mèo"]
  • D: [1, 2, "chó"]
Answer

Answer: D

When using ... in array, it is called spread operator in JavaScript which, technically, is similar to rest parameter (using in the context of function). It provides a more elegant way to concat (combine) or copy array. In the code above, b is a copy of a. However, as we pass a in to a Set, it will return the unique value only in a. It means, now we have `[1, 2, "chó", 3, "mèo"] in b.

However, we then set the length for b as 3. Note that "chó".length returns 3 but in PHP, strlen("chó") returns 4, just in case you are coming from PHP world.

As we set the length for the array b, we also cut down the array itselt. That is the reason why we get [1, 2, "chó"] printing out in the console.

40. What's the output?
const mot = function (m) {
  return arguments[0];
};

const hai = function (...m) {
  return arguments[arguments[0]];
};

const a = [mot(123), hai(1, 2, 3)];

console.log(typeof a !== "object" ? a[0] : a[1]);
  • A: 1
  • B: 2
  • C: 3
  • D: 123
Answer

Answer: B

First, it should be noted that arguments cannot be used in an arrow function, so in order to take advantage of this feature, we have to write the function in the casual form. arguments returns an array-like object that contains any parameter we pass into the function when executing it.

... is a rest operator. We use this feature in function and array. Noted that in the context of array, it is called spread operator and it behaves differently. When declaring a function with ..., we can pass as many parameters into the function itselt when executing it as we want.

Note that in the function hai, we return arguments[arguments[0]] which means hai(1, 2, 3) will return 2 rathern than 1 because arguments[0] return 1 and then arguments[1] returns 2.

The last thing we have to take note is that the typeof operator of an array will return object, here the trick seems more daunting. The final anwser is 2 as we got it in a[1], or hai(1, 2, 3).

41. What's the output?
class Component {
  constructor(age) {
    this.age = age + `${typeof Coder}`.length;
  }

  getAge() {
    return ++this.age;
  }
}

class Coder extends Component {
  constructor(age) {
    super(age);
    this.age = age - `${typeof Coder}`.length;
  }
}

const a = new Coder(16);

console.log(a.getAge());
  • A: 7
  • B: 8
  • C: 9
  • D: 10
Answer

Answer: C

We have two simple classes in which Coder extends Component. Nothing fancy. As typeof ClassName returns function rather than class, we then get 8 in the operator "function".length.

Though we implement super(age) in the Coder class, we actually overwrite the contructor of the parent class Component in the child class Coder. Therefore, when initiating the object a, the following code is automatically triggered this.age = age -${typeof Coder}.length;. The difference between the child and parent 's constructor is minus (-) and plus (+) in the above code.

As such, we have 16 - 8 rather than 16 + 8, which returns 8. The function getAge() returns 9, so the corrent answer is C.

Bear in mind that JavaSCript is not a "real" OOP programming language even though we can now implement class and object as in other languages.

42. What's the output?
class RemoveFalse {
  constructor(element) {
    this.element = element;

    this.length = this.removeFalse().length;
  }

  removeFalse() {
    this.element = this.element.filter(Boolean);

    return this.element;
  }
}

const theArray = [true, false, 1, 0, NaN, undefined, "", null, "js.edu.vn"];

const a = new RemoveFalse(theArray);

console.log(a.length);
  • A: false
  • B: true
  • C: 2
  • D: 3
Answer

Answer: D

The key message that can be taken away in the code snippet above is filer(Boolean) which can be taken into consideration in case you want to eliminate falsy values in an array. We can use filter(callback) or filter(Boolean) in particular in this case to do that. Note that we have to pass into the filter function a callback and in this case Boolean is actually a function. You can check typeof Boolean to see it.

Similar to map or reduce function, filter always returns a new array from the exisiting one. [true, false, 1, 0, NaN, undefined, "", null, "js.edu.vn"].filter(Boolean); will return [true, 1, "js.edu.vn"];, hence calling the function removeFalse() gives us 3. So the correct answer is 3.

43. What's the output?
const coderfarm = [1, [], {}, [], 2, 3];

const converted = Number(coderfarm instanceof Array);

const result = coderfarm.indexOf(converted + true);

console.log(result);
  • A: []
  • B: {}
  • C: 2
  • D: 4
Answer

Answer: D

We have a simple array in the code snippet above that includes some digits, two other arrays and one object. Using the built-in function Number, we can convert any value passing to the function into digit. As coderfarm instanceof Array returns true, then converted get 1. Noted that you can use another way to check the type of an array is Array.isArrray(arrayToBeChecked) which return a boolean value. Suprisingly, the operator typeof [] returns object rather than array.

The built-in function indexOf will return the index of the element that is being checked. So as converted + true return 2, we are going to check the index of the element with the value 2 in the array coderfarm.

We get 4 in the console.log and the correct answer is D.

44. What's the output?
const converter = (arrayInput) => {
  return { ...arrayInput };
};

const content = ["function", "object", "decorator"];

const checking = content[Number(false)];

const result = typeof converter(content) === content[1];

console.log(checking ? (result ? (typeof converter).length : false) : false);
  • A: 6
  • B: NaN
  • C: true
  • D: 8
Answer

Answer: D

The operator ... in JavaScript is very handy. The function converter is quite trivial, it takes advantege of ... (rest operator || spread operator) to turn an array into an object.

First we have the constant checking with the value function given that Number(false) gives us 0 and that is the first index in the array content.

Second, the constant result gives us the value true as the typeof converter(content) is function, which is also the value of content[1].

Then in the final code, we have checking = true, and then result = true as well, so the final result is (typeof converter).length which is equivalent to "function".length because the typeof of converter is simply function. We get 8 after all and the correct answer is D.

So the key message here is that we can take advantate of the spread operator (or ...) to turn an array to an object. For example: const a = ["hello", 2], then we can have a go with const b = {...a} and b is now an object with the following value: {0: "hello", 1: 2}. The key of the object is actually the index of the original array.

45. What's the output?
function* js(length) {
  for (let i = length.length; i > 0; --i) {
    yield i;
  }
}

let getJS = js(typeof js);

let result = getJS.next().value;

console.log(result + getJS.next().value);
  • A: 10
  • B: 14
  • C: 15
  • D: 16
Answer

Answer: C

We have a generator function in the code snippet above, which is defined with the *. Noted that we can "store" as many result as we want in a generator thanks to the keyword yield.

As the typeof js is function, so the length of the string function is 8. So when calling getJS.next().value;, we get 8. However, in the next calling, it returns 7, and in the following calling after that, we get 6. That is why generator can "store" and "release" (or return) as many value as we want.

So the answer is C, which is 8 (first execution of the generator) + 7 (second execution of the generator).

46. What's the output?
var ages = [10, 15, 20, 25];

let response = [];

ages.some(function (currentValue, index, ages) {
  if (currentValue > ages[ages.length - index])
    response.push(currentValue + ages.length);
});

console.log(response);
  • A: [20]
  • B: [20, 25]
  • C: [25, 29]
  • D: [29]
Answer

Answer: D

Array.prototype.some() is a built-in function facilitating us to iterate the array using a callback. As in the code snippet above, there are three parameters in the callback, namely currentValue (the value of the current element that is being checked), index (the index of the element in the array that is being checked/evaluated) and ages (the array itself).

The function some() returns a boolean value. The code currentValue > ages[ages.length - index] returns true only one time, which is the last element. Let 's examine the code when it runs through each element:

  1. 10 > ages[4 - 0]. As ages[4] returns undefined, and 10 > undefined returns false, it stops.

  2. 15 > ages[4 - 1]. As ages[3] returns 25, it breaks as the operator returns false.

  3. 20 > ages[4 - 2]. As ages[2] returns 20, it breaks as the operator returns false.

  4. 25 > ages[4 - 3]. As ages[1] returns 10, it returns true. Only this value is being pushed to the array response.

So response.push(currentValue + ages.length) will add the value 25 + 4 to the array response, D is the correct answer.

47. What's the output?
const getSTring = (string, method = false) => {
  if (method === true) {
    return string.slice(1, 4).length;
  }

  return string.substr(1, 4).length;
};

console.log(getSTring("hello", true) + getSTring("hello"));
  • A: 6
  • B: 7
  • C: 8
  • D: 9
Answer

Answer: B

getString() is an arrow function with two parameters. As you can see that the parameter method has the default value false, then if you do not pass any value to it when executing the function, the default value will be used.

The key thing to take note from the code above is the difference betweet slice(1, 4) (which returns 3 characters) and substr(1, 4) (which returns 4 ones).

Finally console.log(getSTring("hello", true) + getSTring("hello")) returns 7 because slice and substr are both used.

48. What's the output?
(function (a, b, c) {
  console.log(Boolean([...arguments].slice(2, 3)[0].slice(3, 4)));
})("hello", "world", "new zealand");
  • A: "new"
  • B: true
  • C: "land"
  • D: false
Answer

Answer: B

The code above is a self-executing function. It runs when it is being declared. We have three parameters and three arguments passed are "hello", "world" and "new zealand".

First, arguments returns an object consisting of arguments passed to the function when executing it. However, using spread operator ..., we then convert the object to an array. We can also do it by using Array.from(object).

Second, slice(2, 3) extracts the element from the index 2 to the index 3, which returns "new zealand". It is still an array. We then extract the element with the index [0] and we get the string "new zealand" rather than an array.

Third, "new zealand".slice(3, 4) gives us an empty string (with a space between) " ". The Boolean(" ") gives us true. Noted that if there is no space in the empty string, we get false instead.

So the correct answer is B.

49. What's the output?
class HocCoBan {
  name = "hello world";

  getSlice(slice) {
    return this.getName(slice).slice(true, this.name.length);
  }

  getName(space) {
    return this.name.split(space);
  }
}

HocCoBan.prototype.split = function (argument) {
  return this.getSlice(argument);
};

const a = new HocCoBan();

console.log(a.split("").length);
  • A: NaN
  • B: true
  • C: 10
  • D: 11
Answer

Answer: C

The code above is nothing much special. However it is written in a complicated way on purpose. First, we have a class named "HocCoBan" with two methods and one property. Then we add another method split using the tradional way (via prototype). Note that class in JavaScript is simply a syntactic sugar of function given that typeof ClassName return function.

When we call the method split, we pass the an empty string to it. This method then call other methods. The flow is as follows:

split("") ==> this.getSlice("") ==> this.getName("") ==> this.name.split(""). Here split is a built-in function that convert a string to an array.

Noted that in getSlice(), we also use .slice(true, this.name.length) to slice (cut) the array from the index 1 to 11. So the length is 10.

So the final answer is C.

This code might help us master the concept function prototype in JavaScript and the understand the difference between the built in function String.prototype.split and the function we declare by ourself HocCoBan.prototype.split.

50. What's the output?
function javaScript(node) {
  let mot = node.includes("I") ? "love" : "you";

  return function (deno = mot) {
    let hai = node.replace(deno, "done");

    return function (done = hai) {
      return (node + deno + done).length;
    };
  };
}

console.log(javaScript("I love you")()());
  • A: 18
  • B: 24
  • C: 20
  • D: 25
Answer

Answer: B

Apart from learning some built-in functions to handle string such as replace and inclues, we are reviving the concept of currying function in JavaScript. Say you want to declare a function with three parameters, you may consider refactoring the code by declaring 3 nested functions, each with one parameter you wish to pass to. Basically, both of them work in the same way. However, noted that only the outerest (the main) function has the name as javaScript in the code above. Both nested (inner) functions are declared without the name. We also use three return keywords in the code.

When executing the function, you then have three () as in the javaScript("I love you")()(). We do not pass any argument into the second and third functions (both are inner/nested functions without the name) and these functions will take the default value we have alreaded declared when being executing.

All in all, we have the final operator return (node + deno + done).length; in which node is "I love you", deno is "love" and done is "I done you". The length of these strings is 24, which you can calculate by yourself the concatenated string I love youyou I done you. Be aware of the empty space, which is also taken into account.

51. What's the output?
const www = ["hello", "coranovirus", "kiora", "world", "new zealand"];

const found = www.find(function (world) {
  return world > "victory";
});

const result = found[1] < www[0][0] ? www[false ? 1 : 0] : www[true ? 0 : 1];

console.log(result);
  • A: "hello"
  • B: "world"
  • C: "victory"
  • D: "w"
Answer

Answer: A

The key information in the question above is about the method Array.prototype.find(). It returns the first element in the array that meets the condition declared in the callback function, which is passed to the function. The array is being iterated to check every single element. In the code above, we might easily see that the element world is the first element in the array that has a larger value than victory. Remember that "w" > "v" return trues if the two letters are compared. When two words are being compared, only the first letter in each word is being utilised to compare.

As the result, found is now world and thus found[1] returns the letter w whereas www[0][0] gives us the letter h in the element hello. It means found[1] < www[0][0] returns false.

So the final result is www[true ? 0: 1] or www[0], which is hello. And the correct answer is A.

52. What's the output?
(function (flag) {
  let age = Boolean(NaN === NaN ? false : flag);

  console.log(age.toString()[Number(flag)]);
})([]);
  • A: "f"
  • B: "t"
  • C: true
  • D: false
Answer

Answer: B

We have a self-executing function with the parameter/argument is an empty array. Noted that NaN === NaN returns false, then age gets the value flag, which is an empty array. However, the boolean value is true when we call Boolean([]).

The function toString() returns the string true and the Number([]) returns 0. Then we get "t" in the console.log. The correct answer is B.

Keep in mind that Boolean([]) ==> true but Number([]) ==> 0. And sadly NaN === NaN returns false.

53. What's the output?
1) console.log(Boolean([]));
2) console.log(Number([]));
3) console.log(Number(Boolean([])));
4) console.log(Boolean(Number([])));

5) console.log(Boolean({}));
6) console.log(Number({}));
7) console.log(Number(Boolean({})));
8) console.log(Boolean(Number({})));

9) console.log(Boolean(new Boolean(false)));
  • A: true - 0 - 1 - false - true - 1 - 1 - false - false
  • B: true - 0 - 1 - false - false - NaN - 1 - false - true
  • C: true - 0 - 1 - false - false - false - 1 - false - false
  • D: true - 0 - 1 - false - true - NaN - 1 - false - true
Answer

Answer: D

JavaScript is sometimes tedious to deal with given that it is a loosely type language. The data type of a variable can be changed depending on the value. An unexpected behaviour might unfortunately occur when you change/convert the original value to another one.

For example, the code 2 Number([]) returns 0 and 6 (Number({})) returns NaN, although both (Boolean([])) and (Boolean({})) return true.

In the code 9 Boolean(new Boolean(false)), we get true even though we pass into the function constructor Boolean() a false (as the) parameter. However, if we do not use the keyword new, then false will return. It seems that in Boolean(new Boolean(false)), we have a valid opreration, so it is true. However, in the Boolean(Boolean(false))) where we do not use the keyword new, we then get false because now a false value is being evaluated rather than an operation.

So, the correct answer is D.

Credit: @tiepphan, Vietnamese Angular Facebook group.

54. What's the output?
const myYoutube = {
  name: "hoccoban",
  address: "youtube.com/hoccoban",
  getInfo() {
    return this;
  },
  content: () => (this === window ? myYoutube.getInfo() : this),
};

console.log(myYoutube.content().name);
  • A: "hoccoban"
  • B: window (object)
  • C: NaN
  • D: undefined
Answer

Answer: A

To answer the tricky question above, you might want to have a look at the concept of this in JavaScript (on browser environment). By default, this refers to window object. Note that Window (written in capital) is the Function constructor of the window object. In this regard, console.log(this === window) return true but console.log(this === Window) returns false.

As content() is an arrow function, this declared inside this function points to window, so myYoutube.content() returns myYoutube.getInfo(). Noted that we have to explicitly write myYoutube.getInfo() to make sure the code will run correctly as this in this case does not work as it does not refer to the currect object. In the function getInfo(), however, this actually refers to the currect object instead of window object because we use a normal function here.

Then we have the property name with the value "hoccoban". So the correct answer is A.

Credit: Thanks https://github.com/phanvigiaii for fixing the typo. Please make a pull request when you have time bro. Cheer.

55. What's the output?
const myArray = [1, 2, 3];

myArray.someProperty = this;

Array.prototype.someOtherProperty = "hello";

let result = [];

for (let key in myArray) {
  result.push(key);
}

for (let key in myArray) {
  if (myArray.hasOwnProperty(key)) {
    result.push(key);
  }
}

console.log(result.length);
  • A: 10
  • B: NaN
  • C: 9
  • D: 7
Answer

Answer: C

We have a simple array that consists of 3 elements. If checking the type of the array with the operator typeof, we will have object. (Hint, you can make use of Array.isArray(array)) or array instanceof Array to check its type).

When declaring myArray.someProperty, we now add a new property to that array and when declaring Array.prototype.someOtherProperty = "hello", we add a new property to every single array.

As a result, the for... in loop will iterate through the array in question and return its key/property and the inherited property as well. However, in the second iteration, we take advantage of the method hasOwnProperty(key) to check whether a particular key/property actually belongs to the array in question rather than the inherited one.

In short, in the first iteration, we get 5 (3 original ones, 1 property that is directly added to the array, 1 inherited from the Array.prototype. In the second one, we only get 4 as the inherited property is not taken into consideration.

Keep in mind that, we use for... of to loop through an array or the traditional for loop. It is not a good practice to use for ... in to loop through an array. It is often used to loop through an object.

56. What's the output?
const coderfarm = [1, 2, 3, 4, 5];

const [top, ...bottom] = (function (a) {
  let result = a;

  a.unshift(new Array(3));

  return result;
})(coderfarm);

console.log(top.length + bottom.length);
  • A: 8
  • B: 9
  • C: 10
  • D: 11
Answer

Answer: A

We are using destructure array (or object) technique to extract element of an array (or object). We also take advantage of ... (spread parameter) here.

The array we are destructuring is returned from a self-executing function. First we pass the parameter coderfarm, which is the parameter a when declaring the function. Then we update this array with some additional value (an array with three undefined elements using new Array(3)) on the top of the array (using unshift). The array is updated now as [[undefined, undefined, undefined], 1, 2, 3, 4, 5].

So top is the first element of the array or [undefined, undefined, undefined], which returns 3 when we check the length.

The bottom returns the rest of the array in question, which is 5 when using length property.

The final number is 8 and thus the correct answer is A.

57. What's the output?
let age = { number: 10 };

const getAge = (flag) => {
  flag ? delete age.number : delete age;
  return age.number++;
};

console.log(getAge(false));

console.log(age.number);

console.log(getAge(true));

console.log(age.number);
  • A: 10 - 10 - NaN - NaN
  • B: 10 - 10 - undefined - undefined
  • C: 10 - 11 - undefined - undefined
  • D: 10 - 11 - NaN - NaN
Answer

Answer: D

The operator delete only works on the property of an object, not the object itself. In the code snippet above, we have a simple function getAge with the parameter flag. When the flag is true, we trigger delete age.number and if it is false, we will use the operator delete upon the whole object.

As this operator does not work on an object, if we can say that, it turns out that delete age actually does nothing. As such, console.log(getAge(false)) returns 10 and simultanously increases the value of age.number to 11. The value is now being kept in the memory. As such, console.log(age.number) will return 11.

When we pass the argument flag as true in the console.log(getAge(true)), we will trigger delete age.number which removes the value and the property age.number itself. It means age.number is now undefined. However, because we also attempt to increase the value of this undefined property using ++ operator, it returns NaN.

console.log(age.number) also returns NaN as well. So the correct answer is D.

58. What's the output?
const youtube = { name: "hoccoban" };

const copy = Object.create(youtube);

const cloneA = Object.assign({}, copy);

const cloneB = Object.assign({}, youtube);

console.log(cloneA.name);

console.log(cloneB.name);

console.log(copy.name);
  • A: undefined - "hoccoban" - "hoccoban"
  • B: "hoccoban" - "hoccoban" - "hoccoban"
  • C: "hoccoban" - "hoccoban" - "undefined"
  • D: undefined - "undefined" - "hoccoban"
Answer

Answer: A

We have three outputs in the code snippet above.

First console.log(cloneA.name); will return undefined but why? We use Object.assign to clone a new object from an empty and from the object copy. The object copy itself is actually created from the original object youtube using Object.create. Noted that because we use Object.create, copy will inherit the data from the original one but it is still an empty object itself.

Second, both console.log(cloneB.name) and console.log(copy.name) return "hoccoban" because cloneB.name will have the actual property name. On the contrary, copy.name outputs the property name inherited from the youtube.

So the correct answer is A.

59. What's the output?
((x) => {
  const data = !Array.isArray(x) ? x : x.entries();

  console.log(data.next().value[1]);
})(["hello", "world", "vuong"]);
  • A: NaN
  • B: "hello"
  • C: "world"
  • D: "vuong"
Answer

Answer: B

We have a self-invoking function here and we pass an array to it when the function is executed. Note that Array.isArray(x) return true but actually we use ! before Array.isArray(x). It means data will return x.entries().

The method array.entries(), as you might have already known, returns a gererator. Here we will call next() to iterate through each element. Note that if you only call next() once, it will only return the first element instead of the whole iterator.

Then when we call value, it returns an array with the index and the value of the iterator. So what will we get if we call console.log(data.next().value[0]). Sure, it returns 0 as 0 is the index.

So the correct answer is B.

60. What's the output?
let x = Symbol();

let y = Symbol();

console.log(x === y ? `${typeof x}`[1] : `${typeof x}`[2]);
  • A: NaN
  • B: "object"
  • C: "y"
  • D: "m"
Answer

Answer: D

As x and y are both instances of symbol, they are unique in our codebase; therefore, the === comparison will return false as expected. In the simple code snippet above, we get the else operation.

It should be noted that the typeof x operation gives us symbol, and since a string in JavaScript is iterable, we get m as we pass in the index 2.

So the correct answer is D.

61. What's the output?
const frameworks = ["react", "angular", "vue"];

const iterator = frameworks[Symbol.iterator]();
const i = frameworks.entries();

iterator.next();
i.next();

console.log(iterator.next().value[1]);
console.log(i.next().value[1]);
  • A: "react" - "angular"
  • B: "react" - "react"
  • C: "angular" - "angular"
  • D: "n" - "angular"
Answer

Answer: D

As frameworks is an array, it has a built-in method named Symbol.iterator. You can hence iterate through the whole array using commonly used methods such as for... of, normal for loop, forEach or map, among others. That is relatively trivial, I suppose.

This code challenge above is written to help us understand the concept of iteration better. First, we use the built-in method called entries() to create a new iteration. So does Symbol.iterator. Both seem to do the same thing.

Each time we call next() method, the iteration will output one element. We then can call value() to get the value. The difference between iterator and i is that the former shows the value itself while the latter outputs an array consisting of the index and the value. It means that in the code above, iterator.next().value returns angular and i.next().value gives us [1, angular].

So the correct answer is D.

62. What's the output?
class React {
  theName = "Not React";
}

class Author extends React {
  static theName = "Real React";

  render() {
    return this.theName;
  }

  static render() {
    return this.theName;
  }
}

const me = new Author();

console.log(me.render());

console.log(Author.render());
  • A: "Not React" - "Real React"
  • B: "Not React" - Error
  • C: Error - Error
  • D: Error - "Real React"
Answer

Answer: A

We have two classes in the code snippet above. It sounds we are imitating React. The React class has only one property named theName, and no method is declared here. Providing that Author extends the React class, it inherits that property, surely. However, we have also declared another property with the same name in the Author classs. The difference is that the property declared in the child class is given the keyword static.

The Author class also has two methods with the same name render(), one as regular methods and another with static keyword. Will that work in JavaScript?

It turns out that JavaScript is quite flexible. It supports both property and method if they are declared with the same name as long as they are either regular property (or method) or static property (or method).

The last thing you should be aware of is that the method static render() only calls the static property, here is static theName = "Real React"; So does the regular method render(). As such, the code does not run into any issues.

So the correct answer is A.

63. What's the output?
class js {
  say = "hello";
}

js.prototype.say = "goodbye";
console.log(new js().say);

js.prototype.thename = "google";
console.log(new js().thename);
  • A: Error - Error
  • B: "hello" - "google"
  • C: "goodbye" - "google"
  • D: Error - "google"
Answer

Answer: B

js is a standard class declared in the code snippet above that has only one property with the name say. Then we again declare another property with the same name say for it. You might think that the property say has been overwritten with a new value goodbye.

That is not the case as we will get hello when we run console.log(new js().say);. It is clear that the JavaScript engine prioritizes the property declared inside the class more than the property declared later using the prototype mechanism.

If the property has not been declared inside the class itself, we can then add a new one with the help of prototype as in thename. Without the doubt, the code console.log(new js().thename); gives us google as expected.

So the correct answer is B.

64. What's the output?
const App = ([y, x, z]) => {
  return () => {
    ++x;
    return () => {
      return x++;
    };
  };
};

console.log(App([10, 20, 30, 40])()());
  • A: 10
  • B: 32
  • C: 21
  • D: 22
Answer

Answer: C

To answer the question raised on the above code snippet, you might want to revisit two concepts, currying function and destructing array or object.

First, currying function means we convert a function with multiple parameters into multiple functions with a SINGLE parameter. Then you can easily manipulate the flow of the data. Noted that currying function is relevant to higher-order function, you might want to have a look.

destructing array or object means we attempt to extract a complex array or object more conveniently. For example, [y, x, z] = [10, 20, 30, 40] will extract y, x and z with the value 10, 20 and 30 respectively.

The last thing is incremental operator here ++x returns 21 but x++ does not as it still returns 21.

So the correct answer is C.

65. What's the output?
const numbers = [5, 6, 7];

function callback(accumulator, currentValue) {
  return accumulator + currentValue;
}

const theCallBack = (accumulator, currentValue) => accumulator + currentValue;

const sum = numbers.reduce(
  callback,
  numbers.reduce(theCallBack, numbers.reduce(theCallBack, 7))
);

console.log(sum);
  • A: 54
  • B: 55
  • C: 60
  • D: 61
Answer

Answer: D

Array.prototype.reduce() is a bit perplexed built-in method that allows you to manipulate data in an array. It returns a single value from the array predefined as in the case with map or filter. The syntaxt of the function is arr.reduce(callback( accumulator, currentValue, [, index[, array]] )[, initialValue]), so it accepts a callback function with four arguments including accumulator, currentValue, currentIndex (optional) and array (optional).

The second argument of the reduce method, which is optional, is called initialValue that will be counted as the first element with the index 0 when reduce is executing. If initialValue is not provided, then reduce will run with the index 1 instead. reduce() sounds complicated, but truly it is not. In case you want to revise the function, kindly take a look at MDN here: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/Reduce

The above code has two callback functions named callback and thecallback, which do the same thing. The seemingly creepy code is the variable sum, which is returned by a couple of nested reduce functions. It turns out that there is only one "real" reduce function and the other ones give us initialValue only.

  • The first initialValue is 7;
  • The first nested reduce function gives us 7 (initialValue) + 5 + 6 + 7 = 25.
  • The second nested reduce has 25 as the initialValue, which we get from the above. Then it returns 25 + 5 + 6 + 7 = 43;
  • The "real" reduce function now has 43 as the initialValue, the we get the final result: 43 + 5+ 6 + 7 = 61.

So the correct answer is D.

66. What's the output?
const a = { name: "hoccoban.com" };
const b = { name: "youtube.com/hoccoban" };

const first = { ...a }.name.length;
const second = { ...a, ...b }.name.length;
const third = { ...a, ...b, name: "hello" }.name.length;

console.log(first + second + third);
  • A: 12
  • B: 37
  • C: 5
  • D: 20
Answer

Answer: B

The code snippet above is relatively trivial. What we can learn from it is all about the spread operator (three-dot ...). Sometimes it is also used as a rest operator to extract data from an object or array.

We have two simple objects which both have the same key name but different values. The constant first gives us the length of the string value of the keyword name that is copied from a. So, first is now 12.

The constant second merges a and b into one object. However, as b has the same key name with a, the object created by merging two objects will have the value of b. It means the constant second gives us the length of youtube.com/hoccoban, which is 20.

third does the same thing with first and second as it merges two objects into one. However, it also adds another key-value to the object. Coincidently, the key now is name, which is the same with the key attained from a and b. Hence, this key and value will take over the merged object. That means third is the length of the string hello, which is 5.

In total, we have 12 + 20 + 5, and the final result is 37.

So the correct answer is B.

67. What's the output?
const hocCoBan = {};

Object.defineProperty(hocCoBan, "domain", {
  value: "hoccoban.com",
});

async function App({ year, age }) {
  return year - age + hocCoBan.domain.length;
}

App({ year: 2021, age: 30 }).then((r) => console.log(r));
  • A: 2051
  • B: 2001
  • C: 30
  • D: 2003
Answer

Answer: D

The code snippet above seems complicated regarding how we take advantage of Object.defineProperty to add key and value to the object hocCoBan. In fact, Object.defineProperty has a couple of handy features that allow us to control the behavior of the object in some situations where we want to make sure that the object created is mutable or not, whether it is iterable (using for..in) and so for. For example, if we set configurable: false when we declare an object with Object.defineProperty, we cannot use delete operator to delete the object's property. We cannot change the value of that property as well.

The second "take away" message when reading the code above is the unpacking object technique, or a more frequent term is the destructing object. Say you have an object with two keys called year and age, then you can get them by using the destructing object technique as follows: {year, age} = theOBject;. In the code above, when declaring the function App, we also use destructing object technique to get the key from the object and use them as the parameters.

If you are familiar with asynchronous code in JavaScript when using the keyword async, it is not a big deal to understand why we need to use then to get the function App being called. It fact, async always returns a promise, so we need to use then method to get the data we want.

The flow of the code is: 2021 - 30 + "hoccoban.com".length (which is 12).

The final result is 2003. So the correct answer is D.

68. What's the output?
class hoccoban {
  #thisyear = 2021;
  constructor(covidTheFirstYear) {
    this.covidTheFirstYear = covidTheFirstYear;
  }

  getThisYear() {
    return this.#thisyear;
  }

  getCovidFirstYear() {
    return this.covidTheFirstYear;
  }
}

const message = new hoccoban(2019);

const result =
  hoccoban.hello ?? message.getThisYear() - message.getCovidFirstYear();

console.log(result);
  • A: NaN
  • B: 2019
  • C: undefined
  • D: 2
Answer

Answer: D

This challenge partly illustrates the newest features of JavaScript detailed in ECMAScript 2020 or ES11.

Now you can declare a private property in a class thanks to the symbol #. Like other languages, a private property in JavaScript can only be accessed from inside the class. It will trigger an error when you attempt to call it outside the class, surely.

The second feature you might see on the code snippet above is the nullish coalescing operator or ??. When declaring some variable such as let myVariable = number ?? 7, if the variable number is either undefined or null, the variable myVariable will be assigned the value 7.

So hoccoban.hello means undefined because we have not added any value yet. Then by using nullish coalescing operator with ?? the variable result simply returns 2 as message.getThisYear() gives us 2020 and message.getCovidFirstYear() gives us 2019. Note that we can access the private property outside of the class via a method, as in the method getThisYear().

So the correct answer is D.

69. What's the output?
const keyWords = "hello world";

const entries = keyWords.split(" ");

const collections = [];

entries.forEach((entry, index) => {
  collections.push([entry, index]);
});

const objectResult = Object.fromEntries(collections);

const { world } = objectResult;

console.log(world);
  • A: 0
  • B: true
  • C: 1
  • D: "hello"
Answer

Answer: C

The code snippet above is not challenging for those who have had decent experience working with ES6 I suppose. First we turn keywords into an array using split() function. Then we create a variable named collection, which initially is an empty array.

Take a closer look at the forEach function, which allows us to run a for loop through the whole array entries, you might realize that push([entry, index]); add an array to collections rather than an element.

The next step is by taking advantage of Object.fromEntries() that converts an array with at least two elements (the form of key-value) to an object. This built-in method is the reversing version of Object.entries(), which extracts key and value from an object to an array.

const { world } = objectResult; is nothing special as we unpack the object using destructing object technique supported since ES6. As the object objectResult has hello and world with two respective values 0 and 1, we get 1 when printing out world, so the correct answer is C.

70. What's the output?
const target = {
  domainname: "hoccoban.com",
  author: "vuong",
};

const handler = {
  get: function (thetarget, prop, receiver) {
    if (prop === "domainname") {
      return thetarget.author.length;
    } else {
      return thetarget.domainname.length;
    }
  },
};

const proxyObject = new Proxy(target, handler);

console.log(proxyObject.domainname > proxyObject.author);
  • A: true
  • B: false
  • C: 12
  • D: 5
Answer

Answer: B

We have implemented a basic use case of Proxy in the code snippet above. Each proxyObject object has two parameters (target and handler). handler is also an object.

Apart from get() as you might see, handler also has a handful of other methods such as set, defineProperty(), has() and so forth. Sometimes, people may say a method is a trap of a proxy object.

Back to the code above, the get method allows us to modify how the proxy object will display the value of the original object. thetarget is the original object, and prop is the property of that object as you might guess. You might choose another name in the get function if you want when creating your handler.

The handler above calculates the length of the string value of the two properties. Based on the flow of if - else code, it swaps the returned value.

So proxyObject.domainname now should be understood as target.author.length which means 5 and proxyObject.author means target.domainname.length which gives us 12. So the output is false. The correct answer is B.

If you do the same thing with the original, it should be something like console.log(target.domainname.length > target.author.length) which returns true.

I believe that Proxy is worth to have a closer look. If that is the case, no place is better than MDN. So have a go at: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Proxy

71. What's the output?
const promise1 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("hello"), 5000);
});

const promise2 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("world"), 4000);
});

(async () => {
  console.time("timeleap");

  const p1 = await promise1;

  const p2 = await promise2;

  console.log(`${p1} ${p2}`);

  console.timeEnd("timeleap");
})();
  • A: Promise { } - "hello world" - timeleap: ~ 5000 ms
  • B: Promise { } - "hello world" - timeleap: ~ 9000 ms
  • C: Promise { } - "hello world" - timeleap: ~ 4000 ms
  • D: Promise { } - "hello world" - timeleap: ~ 1000 ms
Answer

Answer: A

We have already had a couple of questions regarding asynchronous code in general and handling the data flow with promise in particular. If you understand how JS works, I am sure that the code challenge above is not difficult.

We have two promises; each takes 5 or 4 seconds to complete the code and returns "hello" (in promise1) and "world" (in promise2) in the resolve methods, respectively.

Then we take advantage of the async function to chain the two promises to get the result we want. As async function returns a promise so to get the returned value from async function, we have to use then() method. As we do not do that here, then we get Promise { <pending> }.

The question is, does p2 have to wait and only run after p1 complete? It turns out that it does not. Both p1 and p2 run simultaneously in the task queue thanks to web API or nodejs API (the environments by which JavaScript engine runs). So it will not take 9 seconds to finish the code but solely around 5. It means promise1 takes 5 seconds to complete and at the same time, promise2 reaches the bottom within only 4 seconds.

That is why A is the correct answer.

Updated: What happens if promise2 takes 6 seconds instead of 4 ? Well, as promise2 runs almost at the same time with promise1, it will only take 1 second after the promise1 completes. So in total, it takes approximately 6 seconds.

72. What's the output?
const promise1 = () => {
  return new Promise((resolve, reject) => {
    setTimeout(() => resolve("hello"), 5000);
  });
};

const promise2 = () => {
  return new Promise((resolve, reject) => {
    setTimeout(() => resolve("world"), 4000);
  });
};

(async () => {
  console.time("timeleap");

  const p1 = await promise1();

  const p2 = await promise2();

  console.log(`${p1} ${p2}`);

  console.timeEnd("timeleap");
})();
  • A: Promise { } - "hello world" - timeleap: ~ 5000 ms
  • B: Promise { } - "hello world" - timeleap: ~ 9000 ms
  • C: Promise { } - "hello world" - timeleap: ~ 4000 ms
  • D: Promise { } - "hello world" - timeleap: ~ 1000 ms
Answer

Answer: B

The 72nd challenge is almost identical to the 71st. Please take a closer look.

The difference lies in the way we declare a promise. In question 71st, we use two constants, and both return promise, but in question 72, we declare functions and each returns a promise.

If you run the code, you might be surprised with the result as it takes around 9 seconds to complete the code in place of 5 seconds as in the previous question.

It means that const p1 = await promise1; and const p1 = await promise1(); are different as the latter (a function) might block the callstack and const p2 = await promise2(); can only be called after the p1 completes. The two do not run in parallel as the two promises in the previous question.

As it takes 9 seconds to finish, B is the correct answer.

73. What's the output?
let history = {
  year: 2021,
  getYear: function () {
    console.log(this.year);
  },
};

setTimeout(history.getYear, 0);
setTimeout(history.getYear.bind(history), 10);

const { year, getYear } = history;
getYear();
  • A: undefined - undefined - 2021
  • B: undefined - 2021 - 2021
  • C: 2021 - undefined - 2021
  • D: 2021 - 2021 - 2021
Answer

Answer: A

We have three outputs on the code above. First, we have a simple object with one property and one method. Noted that the method point to the property year using this keyword. The problem now happens when we attempt to extract data from the object.

Be aware of the setTimeout method, which will create a separated context that is different from the original object's context. Even though in setTimeout(history.getYear, 0); we have explicitly called the object history, setTimeout will still execute the function history.getYear withthis pointing to the global object. So it returns undefined.

getYear(); is extracted from the object we have defined in the beginning. But as this is out of the original context when executing the function, it returns undefined. This code is called last, but the output is displayed first on the console window as it is a synchronous code.

setTimeout(history.getYear.bind(history), 10); runs last and will give us 2021 as it is bound to the object history. Finally, we get undefined - undefined - 2021, and A is the correct answer.

74. What's the output?
class handleCovid {
  constructor(start) {
    this.start = start;
  }

  calculate(someValue) {
    this.start = this.start + someValue;
    return this.start;
  }

  vaccine() {
    ++this.start;
    return this;
  }

  delaying() {
    ++this.start;
    return this;
  }

  static getFinal(result) {
    return result * 2;
  }
}

const now = new handleCovid(2019);

console.log(handleCovid.getFinal(now.vaccine().delaying().calculate(2020)));
  • A: 2019
  • B: 8082
  • C: 8080
  • D: 8084
Answer

Answer: B

The code snippet above is ugly and sounds complicated at first. Yet, you might encounter a situation when some good "take away" messages might be given. The flow of the code is not hard to understand, I suppose.

First, a function in JavaScript can accept another function as its parameter. With regard to the static keyword, it means we can directly call a static method in the form of className.staticmethod without invoking the object created by the normal way new ClassName.

Besides, you might want to have a look at how we chain more than one method together. That is possible if these methods return this.

Now let break it down:

  • calculate(2020) --> 2019 + 2020 = 4039;
  • delaying().calculate(2020) --> 4040;
  • now.vaccine().delaying().calculate(2020) --> 4041;
  • handleCovid.getFinal(now.vaccine().delaying().calculate(2020) --> 4041 * 2 = 8082;

So the correct answer is B.

75. What's the output?
function HappyNewYear() {
  return "hello";
}

const year2021 = new HappyNewYear();
year2021.__proto__.greeting = "happy";
HappyNewYear.prototype.say = "new year";

console.log(year2021.__proto__ === HappyNewYear.prototype);
console.log(Object.getPrototypeOf(year2021) === HappyNewYear.prototype);
console.log(Reflect.getPrototypeOf(year2021) === HappyNewYear.prototype);

console.log(year2021.__proto__ === Object.prototype);
console.log(year2021 instanceof HappyNewYear);
console.log(year2021 instanceof Object);

const thisyear = new HappyNewYear();
console.log(`${thisyear.greeting} ${thisyear.say}`);
  • A: true - true - true - false - true - false - "happy new year"
  • B: true - true - true - false - false - true - "happy new year"
  • C: true - true - true - true - true - true - "happy new year"
  • D: true - true - true - false - true - true - "happy new year"
Answer

Answer: D

The code snippet above helps us revise the concept of prototype in JavaScript with two essential keywords: __proto__ and FunctionName.prototype. I believe that the code console.log(year2021.__proto__ === HappyNewYear.prototype); is the key to understand the difference between the two. So, in short, every single object in JavaScript has a built-in property __proto__ that gives us an overview of the built-in (internal) [[Prototype]]. They are the things (property and method) the object inherits from the "parent" function constructor or class).

For example, if you declare a literal object such as const a = {} then a.__proto__ === Object.prototype returns true because a inherits the prototype from the "parent" Object. However, if an object is created using function constructor then the "parent" prototype is function constructor itself instead of the Object. So while console.log(year2021.__proto__ === HappyNewYear.prototype); returns true, console.log(year2021.__proto__ === Object.prototype); gives us false.

Be aware of Object.getPrototypeOf (object) and Reflect.getPrototypeOf(object). The two are recommended to use as __proto__ is being deprecated.

You might want to read more about __proto__ at MDN https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Object/proto

The correct answer is D, and btw "happy new year"!

76. What's the output?
const address = {
  name: "hoccoban.com",
  author: "Vuong Nguyen",
};

const key = Reflect.has(address, "author")
  ? Reflect.ownKeys(address)[0]
  : "hello";

Reflect.set(address, "language", "JavaScript");

const totalKeys = Reflect.ownKeys(address).length;

const name = Reflect.get(address, key).length;

const language = Reflect.get(address, "language").length;

console.log(totalKeys + name + language);
  • A: 22
  • B: 10
  • C: 20
  • D: 25
Answer

The correct answer is D. Why? Now let break it down:

  • Reflect.has(address, 'author') gives us true given that the object address has the key author. Simple as it is. So the value of the variable key is now Reflect.ownKeys(address)[0], which in fact is the key name.

  • Reflect.set(address, 'language', 'JavaScript'); set another key-value to the object address.

  • Reflect.ownKeys(address).length; gives us 3 because now it has three keys, so totalKeys is now 3.

  • Reflect.get(address, key).length; gives us the length of the string hoccoban.com which is 12.

  • Reflect.get(address, 'language').length is the length of the string JavaScript, which is 10.

  • The final answer is 3 + 12 + 10 = 25.

Answer: D

77. What's the output?
const myModule = (function () {
  const covidYear = 2019;

  const year = 2021;

  function getYear() {
    return year;
  }

  function getCovidYear() {
    return covidYear;
  }

  function exposeYear() {
    return getYear();
  }

  function exposeCovidYear() {
    return getCovidYear();
  }

  return {
    nothing: undefined ?? null ?? null ?? undefined,
    exposeYear,
    exposeCovidYear,
  };
})();

const result =
  myModule.nothing ?? myModule.exposeYear() + myModule.exposeCovidYear();

console.log(result);
  • A: 2021
  • B: 2019
  • C: 4040
  • D: undefined
Answer

The challenge above will help you revise the revealing pattern and thanks to it you can declare a private variable in JavaScript. Note that we can now declare a private property in a class in modern JavaScript, so the above way of writing a private variable seems old-fashioned.

First, we have an IIFE function - immediately invoked function expressions. There are two variables and two functions as well. However, in the return, there are three key-values. We can not directly access the two variables covidYear and year except for using some already-built functions inside the IIFE.

If you feel the keyword ?? is odd, then you might want to have a look at the latest syntax supported in modern JavaScript called "Nullish Coalescing Operator". It means, if the left element is either undefined or null, the value of the right element will be assigned.

In short, we have myModule.exposeYear() (2021) and myModule.exposeCovidYear() (2019). In total, the final result is 4040. So the correct answer is C.

Answer: C

78. What's the output?
class HocCoban {
  constructor(address) {
    this.address = address;
    this.youtube = "";
  }
  get getAddress() {
    return this.address;
  }

  set setYoutube(channel) {
    this.youtube = channel;
  }

  getYoutube() {
    return this.youtube.length;
  }
}

const web = new HocCoban("hoccoban.com");

web.setYoutube = "youtube.com/hoccoban";

console.log(web.getAddress);

console.log(web.youtube);

console.log(web.getYoutube());
  • A: "hoccoban.com" - "youtube.com/hoccoban" - 20
  • B: "hoccoban.com" - function() - 20
  • C: function() - "youtube.com/hoccoban" - 20
  • D: function() - function() - 20
Answer

set and get are commonly called setter and getter. When declaring a method in a class and putting the keyword set or get before it, you can then call them without using parenthesis - (). Put another way, when using get and set, you can directly get or set the value of/for the properties. Somehow it might be convenient in some cases.

Be aware of the methods declared with a getter as we just need to call the method as we call a property (without using parenthesis).

If you know how a traditional method works in JavaScript, then the code challenge above is not difficult, I suppose. The answer is A.

Answer: A

79. What's the output?
const result = ["ronaldo", "messi", "neymar", "Ronaldo", "LuKaKUUUU"].sort();

console.log(result);
  • A: ["LuKaKUUUU", "Ronaldo", "messi", "neymar", "ronaldo"]
  • B: ["LuKaKUUUU", "messi", "neymar", "Ronaldo","ronaldo"]
  • C: ["LuKaKUUUU", "messi", "neymar", "ronaldo", "Ronaldo"]
  • D: ["messi", "neymar", "ronaldo", "Ronaldo", "LuKaKUUUU"]
Answer

In JavaScript, the built-in sort() method sorts the elements of an array. It returns a sorted array in ascending order. Note that each element will be converted to strings and then compared according to the sequences of UTF-16 code unit values. What does it mean?

It means, "banana" < "cherry" or 80 < 9 (because "80" < "9" in the Unicode order).

If you run the following code const result = [9, 11, 89].sort();, the constant result will be sorted as [11, 8, 9] rather than [9, 11, 89] because the engine will convert the number value to string.

The following codes might give you a hint about the relationship between character and number. Ultimately, as the computer can only understand 0 and 1, all characters and even decimal numbers are then converted to 1 and 0. charCodeAt() gives us the decimal value of any string evaluated.

console.log("LuKaKUUUU".charCodeAt(0)) or console.log("LuKaKUUUU".charCodeAt()) ==> 76 console.log("Ronaldo".charCodeAt(0)) or console.log("Ronaldo".charCodeAt()) ==> 82 console.log("messi".charCodeAt(0)) or console.log("messi".charCodeAt()) ==> 109 console.log("neymar".charCodeAt(0)) or console.log("neymar".charCodeAt()) ==> 110 console.log("ronaldo".charCodeAt(0)) or console.log("ronaldo".charCodeAt()) ==> 114 console.log("9".charCodeAt()) or console.log("99".charCodeAt()) ==> 57 console.log("80".charCodeAt()) or console.log("8".charCodeAt()) ==> 56

Noted that if index is not a number, it defaults to 0. The answer is A.

Answer: A

80. What's the output?
const anArray = typeof [];
const aTypeOfNull = typeof null;

const weirdFirst = null instanceof Object;
const weirdSecond = [] instanceof Object;
const weirdThird = [] instanceof Array;

console.log(anArray);
console.log(aTypeOfNull);

console.log(weirdFirst);
console.log(weirdSecond);
console.log(weirdThird);
  • A: "array" - "null" - false - true - true
  • B: "array" - "object" - false - true - true
  • C: "object" - "object" - false - false - true
  • D: "object" - "object" - false - true - true
Answer

In the 80th challenge question, we will review some fundamental "issue" or "weird" features in JavaScript relating to the typeof and instance operators. Given that the original version of the JavaScript language was designed in just 10 days, there are a bundle of inconsistent behaviors that cannot be fixed. They are permanent features existing in the modern language. If we fix it, a lot of websites might crash.

The above code shows us some of the weird features in JavaScript. For example, [] is an array but the typeof [] gives us object. Note that you might take advantage of Array.isArray([]) rather than typeof to examine whether a variable is an array or not.

typeof null; is another weird operator as it returns object. However null instanceof Object; returns false. WhatTheHell!!!

Man, [] instanceof Object; and [] instanceof Array; both return true. How inconsistent it is.

The answer is D.

Answer: D

81. What's the output?
class Dog {
  speak() {
    return this.say();
  }

  say() {
    console.log("hello world");
  }
}

class Cat {
  speak() {
    return this.say();
  }

  say() {
    console.log("kia ora");
  }
}

const animal = new Dog();
animal.speak();
Object.setPrototypeOf(animal, Cat.prototype);
animal.speak();
  • A: "hello world" - undefined
  • B: "kia ora" - "kia ora"
  • C: "hello world" - "kia ora"
  • D: "hello world" - "hello world"
Answer

The central issue/concept mentioned in the code above is the method Object.setPrototypeOf(object, prototype). It is one of the features in ES6, or ECMAScript 2015. Another way to set the prototype of an object is Object.prototype.__proto__ but the latter is controversial.

At first, animal.speak(); gives us "hello world" which is no surprise. Yet, in the second call, we get "kia ora" instead of "hello world". When checking the prototype with Object.getPrototypeOf(animal), you might see that Cat is the prototype of the object animal rather than Dog.

The answer is C.

By the way, kia ora means hello in the Māori language.

Answer: C

82. What's the output?
const js = [9, 10];

function mutate(a, b) {
  a.push(b);
}

mutate(js, 1);
console.log(js);
  • A: [9, 10]
  • B: [9, 10, 1]
  • C: [1, 9, 10]
  • D: ReferenceError
Answer

The code snippet might be pretty trivial if you have already obtained a solid understanding of the two different concepts: reference and value. In JavaScript, non-primitive type such as array and object does not store the value but the reference.

Also, as the arguments in function are passed by the reference, the function mutate will push another element into the array js. Finally, the initial variable 'js' is updated with the new value [9, 10, 1].

If the variable js is assigned a primitive value such as string or number, no matter how the function mutate, it will not mutate the initial variable 'js'. However, if the variable is an object, then it will also be mutated, as in the case of an array in the code above.

B is the correct answer.

Answer: B

83. What's the output when running the code on a browser?
console.log(this === window);
console.log(this === frames);
console.log(this === globalThis);
console.log(this === self);
console.log(this === global);
  • A: true - true - true - true - "ReferenceError"
  • B: true - false - true - true - "ReferenceError"
  • C: true - true - true - true - true
  • D: true - true - "ReferenceError" - true - "ReferenceError"
Answer

The code snippet above might output different results if running on other environments than the browser. For example, there is no self, window, or frames on Nodejs.

global plays the role of the global object in Nodejs, but that is not the case in the browser environment. In contrast, globalThis is available in both browser and Nodejs environments.

The first takeaway message is that Nodejs does have this, global, and globalThis. Browser environment has 5 ones including this, globalThis, window, frames, and self.

The second takeaway message is that Web Worker only has self as the global object.

Ultimately, A is the correct answer.

Answer: A

84. What's the output ?
class StackHocCoBan {
  constructor() {
    this.stack = [];
  }

  push(thing) {
    return this.stack.push(thing);
  }

  pop() {
    return this.stack.pop();
  }

  peek() {
    return this.stack[this.length - 1];
  }

  get length() {
    return this.stack.length;
  }

  isEmpty() {
    return this.length === 0;
  }
}

const firstThing = new StackHocCoBan();

firstThing.push(firstThing.length);
firstThing.push(firstThing.length);
!firstThing.isEmpty() ? firstThing.push(firstThing.length) : firstThing.length;
firstThing.pop();

console.log(firstThing.peek() + firstThing.length + firstThing.pop());
  • A: 3
  • B: 4
  • C: 5
  • D: 6
Answer

The code challenge above is a bit lengthy, frankly. But it might practically help you to revise the concept of stack implemented in JavaScript. Such a concept is quite crucial when it comes to the algorithm, so to say. It appears that stack and queue are blood brothers, and as a developer, you are always advised to master these concepts along with array, linked list, tree, graphs, and so on.

In my opinion, both stack and queue are simply arrays, but they are exclusively built to serve some particular jobs with strict requirements. You might see pop() or push() in the code above are standard native methods we often use when working with the array.

So now firstThing is an object initiated by the class StackHocCoBan. As the class's construct initially triggers an empty array this.stack = [];, first the code firstThing.push(firstThing.length); will actually push the number 0 into the array given that firstThing.length returns 0 as the stack, at the beginning` is empty.

Then firstThing.push(firstThing.length); (the second one) pushes number 1 to the stack as we have already had one item (the number 0).

firstThing.isEmpty() returns false because we have two things in the stack. Yet, be causious with "!" before it. As we write !firstThing.isEmpty(), the code with call firstThing.push(firstThing.length) rather than firstThing.length;. It is a simply short-hand of if-else.

So, the stack is pushed the number 2 because firstThing.length returns 2. So now the stack is as [0, 1, 2], beautiful, right?

firstThing.pop() will eliminate the number 2 and the stack is now [0, 1].

The last line of the code above is firstThing.peek() (1) + firstThing.length (2) + firstThing.pop() (1).

So B is the correct answer.

Answer: B

85. What's the output ?
class QueueHocCoBan {
  constructor() {
    this.queue = [];
  }

  enqueue(item) {
    return this.queue.unshift(item);
  }

  dequeue() {
    return this.queue.pop();
  }

  peek() {
    return this.queue[this.length - 1];
  }

  get length() {
    return this.queue.length;
  }

  isEmpty() {
    return this.queue.length === 0;
  }
}

const items = new QueueHocCoBan();

items.enqueue("constructor");
items.enqueue("QueueHocCoBan");
items.enqueue("hoccoban.com");

const item = items.peek().length - items.length;

console.log(item);
  • A: 6
  • B: 7
  • C: 8
  • D: 9
Answer

Following up the question 85th, we now pay the attention to another important concepts - queue - which is a bit different from stack. While stack leverages two native array methods push() (for adding new item) and pop()(for extracting new item), queue utilises unshift() (for adding new item) and pop() (for extracting new item). In essense, both stack and queue are array and thus the difference between them, IMO, lays in the way push() and unshift() are implemented for adding new item. While push() adds a new item/element at the end/tail of the array, unshift() adds the new one to the top/head of the array itself.

First, the object items will add three new elements into the array (initiated by the constructor) at the top one by one, thanks to the method enqueue. The array will look like this ["hoccoban.com", "QueueHocCoBan", "constructor"];

Now items.peek() gives us "constructor" and thus items.peek().length returns 11. items.length simply returns 3 and so the final result will be 8.

So C is the correct answer.

By way of summarisation, using queue, we can only take out the first element (also the tail of the array) with the method dequeue(). You might need to find another way to extract other elements of the queue.

Answer: C

86. What's the output ?
const domains = new Map();

domains.set("site", "hoccoban.com");
domains.set("youtube", "youtube.com/hoccoban");

const keys = domains.keys();
const values = domains.values();

let result = domains.has("hoccoban.com")
  ? values.next().value
  : keys.next().value;

console.log(result);
  • A: "site"
  • B: "hoccoban.com"
  • C: "youtube"
  • D: "youtube.com/hoccoban"
Answer

There are two ways hash tables are implemented in JavaScript: object and Map(). Hash tables are data structures created to store information in the form of key-value.

While the native object defined as {} is trivial for a JavaScript developer, a new way to design a hash table has been recently added into the language. When writing const domains = new Map();, we have declared an object with a couple of different features as opposed to the veteran one declared in the form of {} or new Object we all know.

You are advised to take advantage of the built-in methods such as set(), get(), has() or delete() among others to manipulate the map object.

Note that a map object can be iterated using for of. You might need to use next().value to extract the raw information written. At this point, you might want to revisit function generator to see why we need to do that.

Both keys() and values() are native methods of the class Map. You might see that they works on a map object as in a normal object. Back to the code snippet above, let result = domains.has("hoccoban.com") ? values.next().value: keys.next().value returns keys.next().value given that domains.has("hoccoban.com") returns false.

So console.log(result) gives us "site", so A is the correct answer.

Note that if you want to extract "youtube", you must run keys.next().value twice.

Answer: A

87. What's the output ?
function inner(flag) {
  hello = 10;
  if (flag) {
    return ++hello;
  }
  return --hello;
}

var hello = 12;
inner(hello > 11 ? false : true);
console.log(hello);
  • A: 9
  • B: 10
  • C: 11
  • D: 12
Answer

As a variable in JavaScript can be declared without any functional keyword such as var, let, or const standing before the variable name, as seen on the code above, this challenge sounds a bit odd as now developers tend to use const or let everywhere. What happened when we do that? The variable will have a global scope no matter where it has been written initially. So first, the variable hello will be accessed anywhere outside of the function scope whenever (after to be exact) the function inner has been executed.

Then we redeclare the hello variable with a different value.

The code inner(hello>11? false: true) is translated to inner(false), so the variable hello declared inside this function is mutated to 9.

As hello is now 9 instead of 12, A is the correct answer.

Answer: A

88. What's the output ?
const collections = ["a", [123], true, "c", { abc: "hello world" }];
const iterator = collections.values();

const a = iterator.next().value.length;
const b = iterator.next().value.length;
const c = iterator.next().value.toString().length;
const d = iterator.next().value.length;
const e = Object.values(iterator.next().value)[Number(a > b)].length;

console.log(a + b + c + d + e);
  • A: 12
  • B: 14
  • C: 16
  • D: 18
Answer

The code snippet above is frankly tedious as it does not solve any big problem but is still written in a very cryptic manner, honestly. I want you to pay a bit more attention to the native function to manipulate array values(). For your information, I also use Object.values() so that you can somehow make a comparison between the two by yourself.

In the beginning, we have a very simple array packed with different types of data such as boolean, string, array and object. The code collections.values(); returns an iterator, so you can not simply access to each element as an usual array. You might run a for of loop here to render all of the elements in this iterator. By doing that, the way iterator works is likely a typical array.

So how do we access a single element in this iterator? We need to use next().value. Each time we call it, the iterator will render the value, one by one, starting with the first element.

It means const a = iterator.next().value.length; returns 1. So a is 1. So does b. C is a bit different and we have 4 here as true, a boolean, is converted to a string. d is 1. So

The code in e is not fascinating, I suppose. Object.values gives us an array of value defined in the object { abc: "hello world" }. [Number(a > b)] give us [0]. So e is simply the length of the string "hello world", which is 11.

Finally, in the console we have 1 + 1 + 4 + 1 + 11 = 18. So D is the correct answer.

Answer: D

89. What's the output ?
const matrix = [
  [1, 2, 3],
  [4, 5, 6],
  [7, 8, 9],
];

function matrixHandle(m) {
  let total = arguments[0][0][0];
  let length = m.length;
  for (let i = 0; i < length; i++) {
    for (let j = 0; j < m[i].length; j++) {
      total += m[i][j];
    }
  }
  return total;
}

console.log(matrixHandle(matrix));
  • A: 44
  • B: 45
  • C: 46
  • D: 47
Answer

You can easily create a two-dimensional array by nesting an array inside the parent one as the variable `matrix` above. To render all of the elements in the matrix, we implement a simple code with two for-loop functions which are nested.

arguments[0][0][0]; gives us 1 because arguments has wrapped the variable matrix in an array. Note that arguments is a Array-like.

So the variable total at first is 1, and then the final value cumulated by looping through the matrix is 46.

In short, we have 1 + 1 + 2 + 3 + 4 + 5 + 6 + 7+ 8 + 9 = 46. So C is the correct answer.

Answer: C

90. What's the output ?
const container1 = {
  stack: "Docker",
  getStack: function () {
    return this.stack;
  },
};
const a = container1.getStack();

const container2 = {
  stack: "Kubernetes",
  getStack: () => this.stack,
};
const b = container2.getStack();

const container3 = {
  architect: "microservice",
  getStack: function () {
    const stack = ["K8s"];
    return stack.map(function (element) {
      return `${element} - ${this.architect}`;
    });
  },
};
const c = container3.getStack();

const container4 = {
  architect: "microservice",
  getStack: function () {
    const stack = ["K8s"];
    return stack.map((element) => `${element} - ${this.architect}`);
  },
};
const d = container4.getStack();

console.log(`${a} -> ${b} -> ${c} -> ${d}`);
  • A: "Docker -> Kubernetes -> K8s - undefined -> K8s - microservice"
  • B: "Docker -> Kubernetes -> K8s - microservice -> K8s - microservice"
  • C: "Docker -> undefined -> K8s - microservice -> K8s - undefined"
  • D: "Docker -> undefined -> K8s - undefined -> K8s - microservice"
Answer

The code above might help you revise how the arrow function works in different contexts, especially when dealing with the keyword `this` in JavaScript. There are two crucial takeaway messages you might need to keep in mind when using a function, as follows:

  • First: The arrow function does not create a new this binding when you use them. It inherits from the parent one(environment) when it is defined.

  • Second: The keyword this could be problematic when it is called in a callback function. For example when implementing setTimeout, setInterval or map, filter, reduce or some, every among others, you will need to pass a callback function. Given that the callback function changes the context, this might therefore change to global object and no longer point to the parent object.

We have 4 objects in the snippet above. Each has a simple property and a trivial method. a returns docker because this.stack exactly points to the object declared container1. However, b returns undefined because this in the arrow function points to the global one rather than container2. Why? As we mentioned above, the arrow function does not create a context for itself, so container2.getStack() is still bound to the global object. this.stack becomes undefined as a result.

Next c gives us K8s - undefined because this is called in the callback function when we use map. A new context is now created by the function map, so this will not point to the object container3. The callback function implemented with map or filter always creates a new context so that this changes.

We get K8s - microservice" in d because the arrow function helps us fix the problem caused by switching context as in the object container3. Here are some lessons learned when dealing with context, nested functions (or implementing callback function):

  • Use normal function rather than arrow function when you write a method inside an object in which the method does not have a nested function(method) or callback one. Arrow function is not recommended when creating object prototypes, classes along with object literals as well.

  • Use the arrow function when you want to access to this, especially in the case of nested method (function) or when using callback function. Otherwise, this will no longer point to the object in these cases (nested method or using callback function with map, filter). There are two other techniques (old-fashion ones) to fix that.

  • There are 3 ways to fix this issue relating to the nested method or callback function: using arrow function as mentioned above, use self = this technique or explicitly binding with call, bind or apply method.

Answer: D

91. What's the output ?
class Angular {
  vendor = "Google";
  lang = "TypeScript";
  overview() {
    let report = [];
    report.push(this.lang);
    report = report.map(function (e) {
      return e.length + this.vendor.length;
    });

    return report;
  }
}

const me = new Angular();

console.log(me.overview());
  • A: 16
  • B: 106
  • C: NaN
  • D: TypeError
Answer

The code snippet above might help you revise the concept of context in conjunction with the way this is treated in JavaScript. In short, if you implement a callback function for the map method (or in another case: nested function), then you might need to pay attention to this binding.

While report.push(this.lang); works pretty well as this points to the property declared within the class Angular, the line return e.length + this.vendor.length; does not work as this no longer points to vendor we have already declared as a property in the class. The map function creates a new context here. As vendor is undefined inside the callback of the map function, we get TypeError in the console. So D is the correct answer.

How to fix that? We can quickly fix this one with one of three techniques: (1) use arrow function for the callback passing to map, (2) temporarily create an alternative this such as let self = this before we call map and use self instead of this. (3) explicitly bind the callback for map using bind, call or apply. We can also pass this as the second parameter for the map function. It also works.

Answer: D

92. What's the output ?
class FrameWork {
  constructor(options) {
    this.options = options ?? ["Angular", "React", "Vue"];
  }

  total() {
    return this.options.length;
  }

  filter() {
    const selected = this.options.filter(function (element) {
      return element[0] === "A";
    });
    return selected[0].length + this.total();
  }
}

const app = new FrameWork();

console.log(app.filter());
  • A: 8
  • B: 2
  • C: 10
  • D: 1
Answer

The code challenge above implements a simple class with two methods. There is only one point in the syntax that you might need to pay a bit more attention to is ?? (nullish coalescing operator ) which is quite similar to || (OR).

?? returns the right-hand side value if the left-hand side is either null or undefined while || does the same thing for falsy value (false, null, undefined, 0, -0, 0n, NaN, "").length

So as we do not pass anything into the constructor when we initiate the object app, this.options takes the default value ["Angular", "React", "Vue"], then the method total() evaluates the length of the array, which is 3.

filter() gives us the length of "Angular", which is 8. So the final value is 10. The correct answer is C.

Answer: C

93. What's the output ?
const origin = [[[123], 321], 213];
const manipulated = origin.flat(origin.length);
console.log(manipulated.length + origin.length);
  • A: 2
  • B: 3
  • C: 4
  • D: 5
Answer

The challenge might hopefully help you have a grip on the native array method flat(), which is quite handy to flatten a nested array. flat() accepts a parameter that defines the level of the nested array you are going to manipulate. By default, this parameter is 1.

The method returns a manipulated array. So on the code about origin.length returns 2 given that the array origin has two elements. When flattening the original array named origin with flat(2), we then have a new array [123, 321, 213].

Finally, we have 5 in the console, and D is the correct answer.

Answer: D

94. What's the output ?
const pipe =
  (...funs) =>
  (v) => {
    funs.reduce((res, func) => {
      return func(res);
    }, v);
  };

const plusFour = (v) => v + 4;
const multiplyBySix = (v) => v * 6;
const divideByTwo = (v) => v / 2;

pipe(plusFour, multiplyBySix, divideByTwo, multiplyBySix, console.log)(1);
  • A: 80
  • B: 90
  • C: 100
  • D: 110
Answer

The pipe function can receive an unlimited number of arguments/parameters thanks to rest parameter ...funcs. These arguments/parameters turn out are function as we call the parent function pipe. In JavaScript, it is quite common to pass a function as a parameter of another function.

Please call these functions, which are passed to pipe, are child functions. They are then looped and executed one by one with reduce method, no matter how many functions you attempt to pass to pipe. v in the code is simply the argument defined in each child function.

So first we have 1, then by executing plusFour it becomes 5. When multiplyBySix is called, the output turns to 30. It becomes 15 when we call divideByTwo. Finally, it becomes 90 as we multiply 15 * 6 when the function multiplyBySix is called again.

So B is the correct answer.

Answer: B