Forem: Julien Dephix

How no IDE & a single monitor paved the way to my PHP expert certification

Julien Dephix — Fri, 22 Dec 2023 16:23:10 +0000

Hello, coders! 💻

Today I'd like to talk about how a frugal setup helped me learn core PHP 20 years which made it a lot easier to get my PHP 5 Expert certification later on.
Yes, PHP 5, the most recent version back then. :D
Context

Back in 2002 I was still a student. In one assignment we had to develop a complete product. I decided to have a go at coding a forum engine.
We're talking login/registration forms, forgotten password, My Account page with some settings, threads and replies with creation dates, last updated at and all that you'd expect from a decent forum engine.

Tools used

I decided to not use any framework and start from scratch to better learn PHP.

Since I fancied a challenge I also decided to use Windows Notepad (this was before I switched to Ubuntu).
With Notepad you get:

no auto completion
no jump to definition
no hints

On the other hand you get such a low memory footprint! :D

Set up: a single screen.

This one is probably the most important factor that helped me remember a ton of functions and their parameters.
You go to php.net and look a function up.
You come back to your editor and code.
You switch back to the doc because you forgot param orders.
Repeat.
Rince.
Now you remember it all because that switching is annoying!

RTFM

How many times have you heard that? It's really important to know how to read a documentation.
And not just the method you're interested in but the whole family too.
Are you looking at array_merge? Why not check array_column as well and notice it takes a third param which can help you refactor foreach loops you've been using to index elements by a specific key?

Reading comments users added on the official PHP doc is also a great source of information. Take time to read them (even though some are from older versions).

Certification

To pass the certification you need to know a lot of functions, what they do, their params etc.
Take a look: https://www.testpreptraining.com/tutorial/zend-certified-engineer-sample-questions/

I also had a 5 day training dedicated to preparing developers for the PHP 5 certification. I've learned more about PHP internals (pointers, memory allocation etc).

And then, I passed the exam!

Conclusion

Even though you might not want to be an official expert it's very important to (know how to) read the documentation.

So read your language official doc, then the one of your framework/tool of choice.
Read about aspects of the language you do not use or are not familiar enough with.

Most importantly: have fun and keep on getting better!

PS

It's 2024 and I still use a single monitor (wmctrl FTW)!

[Postgres] More fun with JSON

Julien Dephix — Fri, 08 Dec 2023 12:10:26 +0000

Hello, coders! 💻

Today we're going to have a quick look at JSON support in Postgres.

What we want to do

Let's say our users can tweak some settings and we want to save them to database in JSONB format.

We'll be using a JSONB column named, you guessed it, settings.

If a user changed their landing page preference then the stored JSON would be {"landingPage":"dashboard"}.
If they changed their language to japanese then setttings would be {"landingPage": "dashboard", "language": "jp"}.

Let's see how we can tackle that. Let's define our table first.

Users table

create table users(
    id SERIAL,
    name text,
    settings jsonb default '{}'
);

Simple enough.

Let's add a user:

insert into users(name) VALUES ('Jools');

Save a setting

First thing that might come to mind would be to

run a query to retrieve a user's settings
decode JSON
modify JSON
run an UPDATE query with the new settings value.

That would work but let's see how we can do that with just one query.

Postgres has a cool function called jsonb_build_object which allows us to, as its name states, build JSON objects.

In our case we'd need to run:

jsonb_build_object('landingPage', 'dashboard')

which would return {"landingPage": "dashboard"}.

The whole update query:

update users
set settings = jsonb_build_object('landingPage', 'dashboard')
where id=1;

yeah but what if there's already a setting, this would overwrite it, right?

Yes! Postgres has a concatenate operator which we will use.

update users
set settings = settings || jsonb_build_object('language', 'jp')
where id=1;

What's this does is build a JSON object {"language":"jp"} and concatenate it with any existing settings which results in:

{"language": "jp", "landingPage": "dashboard"}

Note that existing JSON keys will be overwritten.

select '{"age": "31"}'::jsonb || '{"age": "12"}'::jsonb;

results in:
{"age": "12"}

If you want to preserve existing data then just flip ||, which in our case becomes:

update users
set settings = jsonb_build_object('language', 'jp') || settings
where id=1;

Hope you've learned something.

Happy coding! ⌨️

Calling all beginners: simple SQL challenge

Julien Dephix — Thu, 30 Nov 2023 10:45:36 +0000

Hello, coders! 💻

Here's a little challenge for beginners.

We have a users table and we need to know the number of users in 3 age ranges.

if age < 18 then age range is 'minor'.
if age < 65 then age range is 'adult'.
if age >= 65 then age range is 'senior'.

Table structure

CREATE TABLE users (
  id int NOT NULL AUTO_INCREMENT,
  age SMALLINT,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB;

Sample data

INSERT INTO users(age)
VALUES
(2), (4), (6),
(12), (14), (16), (21), (25),
(27), (31), (33), (37), (39),
(57), (57), (59), (61), (63), (65)
;

Expected results

| ageRange | nb     |
| -------- | ------ |
| 6        | minor  |
| 12       | adult  |
| 1        | senior |

Go ahead and give it a try!

You might want to read about the following statements: IF, CASE/WHEN and GROUP BY.

💡 Click to view my solution 💡
Note that I'm using MySQL but since it is standard SQL you should be able to port it to another DB engine.

First we need to determine age ranges and for that we're going to add a CASE statement.

CASE
    WHEN age < 18 THEN 'minor'
    WHEN age < 65 THEN 'adult'
    ELSE 'senior'
END as ageRange

Pretty straightforward, right? We could have used two IF statements but I opted for CASE/WHEN as I find it elegant. :)

Add that to a SELECT statement and you get users' age along with their age group.

Next we need to count how many users belong to each group.
It's as simple as adding a GROUP BY ageRange clause.

SELECT
    count(*),
    CASE
        WHEN age < 18 THEN 'minor'
        WHEN age < 65 THEN 'adult'
        ELSE 'senior'
    END as ageRange
FROM users
GROUP BY ageRange

And we're done!

Happy coding! ⌨️

Let's have fun with MySQL JSON functions!

Julien Dephix — Sun, 26 Nov 2023 20:53:06 +0000

Header generated with Microsoft Bing's Image Creator (AI).

Hello, coders! 💻

Intro

I'd like to talk about SQL and how we can have fun with the many functions it has to offer.
Today we'll have a look at how we can leverage JSON functions to help us build queries.

What we want to retrieve from our database

Let's say we log actions our users perform on some app. Actions can be start or stop for simplicity's sake.

You're tasked to craft a query to retrieve ID of users whose latest actions is start.

Database structure and data

I'm using MySQL for this tutorial and the table we'll be working with is as follows:

CREATE TABLE actions (
  action_name varchar(20) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NOT NULL,
  user_id tinyint NOT NULL,
  created_at datetime NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;

And some data to play with:

INSERT INTO actions (action_name, user_id, created_at) VALUES
('start', '1', '2023-11-24 10:48:00'),
('stop', '1', '2023-11-24 10:49:00'),
('stop', '1', '2023-11-24 10:49:00'),
('start', '1', '2023-11-24 10:50:00'),
('stop', '1', '2023-11-24 10:51:00'),

('start', '2', '2023-11-24 10:52:00'),
('stop', '2', '2023-11-24 10:53:00'),
('start', '2', '2023-11-24 10:54:00'),
('stop', '2', '2023-11-24 10:55:00'),
('start', '2', '2023-11-24 10:56:00'),

('start', '3', '2023-11-24 10:48:00'),
('stop', '3', '2023-11-24 10:49:00'),

('start', '4', '2023-11-24 10:50:00')
;

Expectations

Based on the above data users with ID 2 and 4 are the ones we're looking for.

Challenge / rule

I tried to do that in a single query, no joins or sub queries. Why? Ya know, for fun and as learning experience!

Query build up

GROUP_CONCAT immediately came to mind. It allows us to get a list of actions, grouped by user, ordered by created date in descending order so the latest action is the first from left to right.

SELECT user_id, GROUP_CONCAT(action_name ORDER BY created_at DESC) as list_of_actions
FROM actions
GROUP BY user_id;

Which results in:

| user_id | list_of_actions             |
| ------- | --------------------------- |
| 1       | stop,start,stop,stop,start  |
| 2       | start,stop,start,stop,start |
| 3       | stop,start                  |
| 4       | start                       |

We can easily see that users with ID 2 and 4 are the ones we want but how can we check that the leftmost value is start?

How about we convert list_of_actions into an array, grab the first element and make sure it is start?

Sounds fun, let's do this.

What does a JSON array look like?

It's a list of comma separated values, each surrounded by double quotes.
Example: ["first_value","second_value"]

Values, double quotes and commas

Let's start by surrounding each value with double quotes.

GROUP_CONCAT's default separator is , but we can pass any string we want.
What would we get if we joined with "," instead?

SELECT user_id, GROUP_CONCAT(action_name ORDER BY created_at DESC SEPARATOR '","') as list_of_actions
FROM actions
GROUP BY user_id;

which gives us:

| user_id | list_of_actions                     |
| ------- | ----------------------------------- |
| 1       | stop","start","stop","stop","start  |
| 2       | start","stop","start","stop","start |
| 3       | stop","start                        |
| 4       | start                               |

Almost! We're just missing a " both at the start and end of list_of_actions.
Let's fix this and, while we're at it, let's also add [ at the start and ] at the end using CONCAT!

SELECT user_id, CONCAT('["', GROUP_CONCAT(action_name ORDER BY created_at DESC SEPARATOR '","'), '"]') as list_of_actions
FROM actions
GROUP BY user_id;

which results in:

| user_id | list_of_actions                         |
| ------- | --------------------------------------- |
| 1       | ["stop","start","stop","stop","start"]  |
| 2       | ["start","stop","start","stop","start"] |
| 3       | ["stop","start"]                        |
| 4       | ["start"]                               |

Looking good!
list_of_actions is now a valid JSON string.

Getting the first element of each array

Now we just need to tell MySQL to consider list_of_actions as JSON and then take the first element.
Converting is done with CAST(value) AS JSON.

And to extract the first value of an array we use JSON_EXTRACT.

Our query is now:

SELECT user_id, JSON_EXTRACT( CAST( CONCAT( '["', GROUP_CONCAT( action_name ORDER BY created_at DESC SEPARATOR '","' ), '"]' ) AS JSON ), '$[0]' ) as latest_action
FROM actions
GROUP BY user_id;

which results in:

| user_id | latest_action |
| ------- | ------------- |
| 1       | "stop"        |
| 2       | "start"       |
| 3       | "stop"        |
| 4       | "start"       |

Excellent!

Final query

Since we said we just wanted users whose latest action is start then we can leave user_id in the SELECT clause and move our JSON manipulating functions to an HAVING clause:

SELECT user_id
FROM actions
GROUP BY user_id
HAVING JSON_EXTRACT( CAST( CONCAT( '["', GROUP_CONCAT( action_name ORDER BY created_at DESC SEPARATOR '","' ), '"]' ) AS JSON ), '$[0]' ) = 'start';

and boom:

| user_id |
| ------- |
| 2       |
| 4       |

Thank you very much indeed.

Conclusion

I hope you've learned a thing or two while using functions such as GROUP_CONCAT, CONCAT, CAST or JSON_EXTRACT.

You probably have better alternatives so please don't hesitate to share your methods! That's how we get better.

Would you be interested in other articles where we experiment more with JSON (or other) functions? Let me know in the comments.

Till next time.

Happy coding! ⌨️

Calling all beginners - part III - fun with strings

Julien Dephix — Tue, 03 Jan 2023 07:43:51 +0000

Photo by Steve Johnson on Unsplash

Hello, coders! 💻

How about a quick little challenge?

How would you left pad zeros to month and day parts of this stringified date 2022-5-3 so that we get 2022-05-03?

👇
Click for a solution using PHP

$input = '2022-5-3';
$output = vsprintf('%04d-%02d-%02d', explode('-', $input));

vsprintf takes an array as input so we do that using explode and then we specify the output format.

More info on the official doc

Happy coding! ⌨️

Celebrities mashup: show us your photo editing skills!

Julien Dephix — Sun, 24 Jul 2022 22:16:43 +0000

Photo by Igor Lypnytskyi on Unsplash

Hello, coders! 💻

Code is everything but there's room for more!
When I first stumbled upon coucoutesqui I knew I had to give it a try!

Rule is simple:

pick two celebrities where one's last name is the other's first name and combine the two to give birth to a new celebrity.

I'll get started: Stevie Wonder Woman.

Fire up your favorite image editor and post fun images in the comments!

Calling all beginners: solution to part II's homework

Julien Dephix — Wed, 20 Jul 2022 07:38:53 +0000

Hello, coders! 💻

In the previous episode of this beginners series I gave you an exercise based on the following initial data:

const purchases = [
    {item: 'socks', amount: 25.98, date: '2022-05-07 09:31:29'},
    {item: 'jacket', amount: 145.98, date: '2022-05-17 17:01:09'},
    {item: 'cap', amount: 12.99, date: '2022-05-17 17:03:12'},
    {item: 'socks', amount: 43.99, date: '2022-06-11 18:24:16'},
    {item: 'cap', amount: 13.99, date: '2022-06-11 18:24:16'},
    {item: 'jacket', amount: 132.99, date: '2022-06-21 11:21:35'},
    {item: 'socks', amount: 13.99, date: '2022-06-23 09:43:28'},
    {item: 'socks', amount: 12.99, date: '2022-07-01 13:14:56'},
    {item: 'jacket', amount: 132.99, date: '2022-07-03 15:24:24'},
    {item: 'cap', amount: 32.99, date: '2022-07-05 11:14:15'},
    {item: 'socks', amount: 13.99, date: '2022-07-07 11:14:15'},
];

Create a list of best and worst selling items of each month.
Best and worst are defined by the revenue they generated.

Expected result:

const topFlops = {
  "2022-05": {
    top: 'jacket',
    flop: 'cap'
  },
  "2022-06": {
    top: 'jacket',
    flop: 'cap'
  },
  "2022-07": {
    top: 'jacket',
    flop: 'socks'
  }
}

Let's see how we could go about implementing a solution.

Step by step solution

First, let's write in plain English what we need to do:

group purchases by month and items
get first and last item of each month and make them top and flop

Sounds good. Let's write our code outline.

const purchases = [ /* omitted for readability */];
const itemsMonthlyRevenue = getItemsMonthlyRevenue(purchases);
const topFlop = getTopFlopFromMonthlyRevenue(itemsMonthlyRevenue);

`getItemsMonthlyRevenue`

getItemsMonthlyRevenue needs to add one level compared to what we had in Part II's first exercise to store items' monthly revenue.

const getItemsMonthlyRevenue = (purchases) => {
    let monthlyRevenue = {}
    purchases.forEach(function (purchase) {
        const purchaseDate = purchase.date.substring(0, 7);
        if (monthlyRevenue[purchaseDate] === undefined) {
            monthlyRevenue[purchaseDate] = {}
        }
        monthlyRevenue[purchaseDate][purchase.item] = purchase.amount + (monthlyRevenue[purchaseDate][purchase.item] || 0);
    })

    return monthlyRevenue;
}

const purchases = [ /* omitted for readability */];
const itemsMonthlyRevenue = getItemsMonthlyRevenue(purchases);
console.log(JSON.stringify(itemsMonthlyRevenue, null, 2));

The main difference with Part II's first exercise is that monthlyRevenue[purchaseDate] is now an object so we need to make sure it's initialized before accessing it.

Let's run it and see what we get in the console.

Output:

{
  "2022-05": {
    "socks": 25.98,
    "jacket": 145.98,
    "cap": 12.99
  },
  "2022-06": {
    "socks": 57.980000000000004,
    "cap": 13.99,
    "jacket": 132.99
  },
  "2022-07": {
    "socks": 26.98,
    "jacket": 132.99,
    "cap": 32.99
  }
}

Excellent. Onto getTopFlopFromMonthlyRevenue!

`getTopFlopFromMonthlyRevenue`

const getTopFlopFromMonthlyRevenue = (monthlyRevenue) => {
    let topFlop = {};
    const NAME_IDX = 0;
    const REVENUE_IDX = 1;
    for (let month in monthlyRevenue) {
        // step 1: get an array of item/monthlyRevenue
        const productsSales = Object.entries(monthlyRevenue[month]);

        // step 2: sort by revenue
        productsSales.sort(function (p1, p2) {
            return p2[REVENUE_IDX] < p1[REVENUE_IDX] ? 1 : -1;
        })

        // step 3: fill the topFlop array object
        topFlop[month] = ({
            flop: productsSales[0][NAME_IDX],
            top: productsSales[productsSales.length - 1][NAME_IDX]
        })
    }

    return topFlop;
}

What's going on here?
Each month holds a list of items as keys and revenue as values and that's what we want to sort.
I make use of Object.entries which returns an array of [key, value] tuples. Example:

console.log(monthlyRevenue['2022-06']);
// outputs
[
  [ "socks", 57.980000000000004 ],
  [ "cap", 13.99 ],
  [ "jacket", 132.99 ],
]

We can now sort that array and get the first and last element which will be out flop and top respectively.

Putting it all together

const getItemsMonthlyRevenue = (purchases) => {
    let monthlyRevenue = {}
    purchases.forEach(function (purchase) {
        const purchaseDate = purchase.date.substring(0, 7);
        if (monthlyRevenue[purchaseDate] === undefined) {
            monthlyRevenue[purchaseDate] = {}
        }
        monthlyRevenue[purchaseDate][purchase.item] = purchase.amount + (monthlyRevenue[purchaseDate][purchase.item] || 0);
    })

    return monthlyRevenue;
}

const getTopFlopFromMonthlyRevenue = (monthlyRevenue) => {
    let topFlop = {};
    const NAME_IDX = 0;
    const REVENUE_IDX = 1;
    for (let month in monthlyRevenue) {
        const productsSales = Object.entries(monthlyRevenue[month]);

        productsSales.sort(function (p1, p2) {
            return p2[REVENUE_IDX] < p1[REVENUE_IDX] ? 1 : -1;
        })

        topFlop[month] = ({
            flop: productsSales[0][NAME_IDX],
            top: productsSales[productsSales.length - 1][NAME_IDX]
        })
    }

    return topFlop;
}


const purchases = [
    {item: 'socks', amount: 25.98, date: '2022-05-07 09:31:29'},
    {item: 'jacket', amount: 145.98, date: '2022-05-17 17:01:09'},
    {item: 'cap', amount: 12.99, date: '2022-05-17 17:03:12'},
    {item: 'socks', amount: 43.99, date: '2022-06-11 18:24:16'},
    {item: 'cap', amount: 13.99, date: '2022-06-11 18:24:16'},
    {item: 'jacket', amount: 132.99, date: '2022-06-21 11:21:35'},
    {item: 'socks', amount: 13.99, date: '2022-06-23 09:43:28'},
    {item: 'socks', amount: 12.99, date: '2022-07-01 13:14:56'},
    {item: 'jacket', amount: 132.99, date: '2022-07-03 15:24:24'},
    {item: 'cap', amount: 32.99, date: '2022-07-05 11:14:15'},
    {item: 'socks', amount: 13.99, date: '2022-07-07 11:14:15'},
];
const itemsMonthlyRevenue = getItemsMonthlyRevenue(purchases);
const topFlop = getTopFlopFromMonthlyRevenue(itemsMonthlyRevenue);
console.log(JSON.stringify(topFlop, null, 2));

Hope it helps you, as a beginner, to see how to tackle a task.

Happy coding! ⌨️

Share your coding mistake(s)!

Julien Dephix — Sat, 16 Jul 2022 01:24:35 +0000

Hello, coders! 💻

I’ll get the ball rolling.

Back in 2006 we were building an e-commerce from scratch and in a hurry as deadline was very tight!
I was in charge of orders from shopping carts to order modifications.
Users were able to edit their orders until the same day at midnight.
Some users complained their orders were missing a DVD or a book. It was a bit random and hard to track.
Upon investigation my colleague realised I had forgotten the order ID clause when deleting products from orders...
So any users who removed a DVD from their own order would remove it from all orders! D’oh!

What are some oops moments you can share?

Calling all beginners - part II: Return Of The Arrays!

Julien Dephix — Wed, 13 Jul 2022 09:40:51 +0000

Cover photo by Faris Mohammed

Hello, coders! 💻

In this second post aimed at beginners we'll continue manipulating arrays because:

arrays are fun
you will have to deal with arrays A LOT in your career so you might as well feel comfortable working with them

Context

You're given a list of purchases customers have made over the past few months. Given the quantity it's fair to say it's not a very popular shop...

const purchases = [
    {item: 'socks', amount: 25.98, date: '2022-05-07 09:31:29'},
    {item: 'jacket', amount: 145.98, date: '2022-05-17 17:01:09'},
    {item: 'cap', amount: 12.99, date: '2022-05-17 17:03:12'},
    {item: 'socks', amount: 43.99, date: '2022-06-11 18:24:16'},
    {item: 'cap', amount: 13.99, date: '2022-06-11 18:24:16'},
    {item: 'jacket', amount: 132.99, date: '2022-06-21 11:21:35'},
    {item: 'socks', amount: 13.99, date: '2022-06-23 09:43:28'},
    {item: 'socks', amount: 12.99, date: '2022-07-01 13:14:56'},
    {item: 'jacket', amount: 132.99, date: '2022-07-03 15:24:24'},
    {item: 'cap', amount: 32.99, date: '2022-07-05 11:14:15'},
    {item: 'socks', amount: 13.99, date: '2022-07-07 11:14:15'},
];

Exercise 1

Your task is to compute the total amount of purchases per month and store that in an object named monthlyRevenue.

Expected object

⚠️ Amounts have 2 decimal places.

let monthlyRevenue = {
  '2022-05': 184.95,
  '2022-06': 204.96,
  '2022-07': 192.96
}

Solution to exercise 1

👇
Click for a step by step guide

Let's see how we can break down what we need to do.

loop through purchases
for each purchase we need to do some sort of date manipulation to get the year/month part
then we add the amount to monthlyRevenue for the date found in our previous step.
then we'll see what's left to do.

Sounds good. Let's code! ⌨️

// purchases omitted for readability
let monthlyRevenue = {}
// first step: loop!
purchases.forEach(function (purchase) {
    // do something
})

Easy.
Now, how do we get the year/month part from each purchase.date?
Well, purchase.date is a string and the 7 first characters are year and month which is what we need.
So how about using substring?

purchase.date.substring(0, 7)

Let's try the following:

// purchases omitted for readability
let monthlyRevenue = {}
// first step: loop!
purchases.forEach(function (purchase) {
    // second step: year/month part
    const purchaseDate = purchase.date.substring(0, 7);
    console.log(purchaseDate);
})

which outputs:

All good!

Onto step 3: sum of amounts.

Now that we have the year/month part we can store amounts in monthlyRevenue[purchaseDate].

Let's see what happens when we innocently add each purchase.amount to monthlyRevenue[purchaseDate] to get a sum.

// purchases omitted for readability
let monthlyRevenue = {}
// first step: loop!
purchases.forEach(function (purchase) {
    // second step: year/month part
    const purchaseDate = purchase.date.substring(0, 7);
    // step 3: sum
    monthlyRevenue[purchaseDate] += purchase.amount;
})
console.log(monthlyRevenue);

outputs:

Object {
  "2022-05": NaN,
  "2022-06": NaN,
  "2022-07": NaN
}

Ouch! NaN? 🍞?
It means Not a Number and is due to the fact that we're trying to add a number to monthlyRevenue[purchaseDate] which is initially undefined.
We need to make sure it's 0 before adding anything to it.

monthlyRevenue[purchaseDate] = purchase.amount + (monthlyRevenue[purchaseDate] || 0);

should do the trick!
|| return the rightmost element if the first one is falsy.
So when we first access monthlyRevenue[purchaseDate] it will return undefined which is falsy so || returns 0.
From now on monthlyRevenue[purchaseDate] will no longer be falsy.

Let's update our code:

// purchases omitted for readability
let monthlyRevenue = {}
// first step: loop!
purchases.forEach(function (purchase) {
    // second step: year/month part
    const purchaseDate = purchase.date.substring(0, 7);
    // step 3: sum
    monthlyRevenue[purchaseDate] = purchase.amount + (monthlyRevenue[purchaseDate] || 0);
})
console.log(monthlyRevenue);

and see the output:

Object {
  "2022-05": 184.95,
  "2022-06": 204.96000000000004,
  "2022-07": 192.96000000000004
}

Hooray! 🚀

Wait, Why so many decimal places?! It has to do with floating points. It's a topic in its own right and I won't go into details here. What I can tell you though, is that there we're not done yet!

Looking back at the exercise's description I see a ⚠️ sign to let us know numbers must have 2 decimal places.

And that's step 4 which I mentioned earlier. ^^

Let's round things after the loop that populates monthlyRevenue. We'll use toFixed:

// step 4: rounding numbers
for (let yearMonth in monthlyRevenue) {
    monthlyRevenue[yearMonth] = monthlyRevenue[yearMonth].toFixed(2);
}
console.log(monthlyRevenue);

which outputs

Object {
  "2022-05": "184.95",
  "2022-06": "204.96",
  "2022-07": "192.96"
}

Almost there! toFixed returns a string and we want a number so we call Number to parse strings and voilà:

// step 4: rounding numbers
for (let yearMonth in monthlyRevenue) {
    monthlyRevenue[yearMonth] = Number(monthlyRevenue[yearMonth].toFixed(2));
}
console.log(monthlyRevenue);

What does the output look like?

Object {
  "2022-05": 184.95,
  "2022-06": 204.96,
  "2022-07": 192.96
}

🚀 Now we're talking. It outputs exactly what was asked.

Putting it all together

Here's all the code:

const purchases = [
    {item: 'socks', amount: 25.98, date: '2022-05-07 09:31:29'},
    {item: 'jacket', amount: 145.98, date: '2022-05-17 17:01:09'},
    {item: 'cap', amount: 12.99, date: '2022-05-17 17:03:12'},
    {item: 'socks', amount: 43.99, date: '2022-06-11 18:24:16'},
    {item: 'cap', amount: 13.99, date: '2022-06-11 18:24:16'},
    {item: 'jacket', amount: 132.99, date: '2022-06-21 11:21:35'},
    {item: 'socks', amount: 13.99, date: '2022-06-23 09:43:28'},
    {item: 'socks', amount: 12.99, date: '2022-07-01 13:14:56'},
    {item: 'jacket', amount: 132.99, date: '2022-07-03 15:24:24'},
    {item: 'cap', amount: 32.99, date: '2022-07-05 11:14:15'},
    {item: 'socks', amount: 13.99, date: '2022-07-07 11:14:15'},
];

let monthlyRevenue = {}
// first step: loop!
purchases.forEach(function (purchase) {
    // second step: year/month part
    const purchaseDate = purchase.date.substring(0, 7);
    // step 3: sum
    monthlyRevenue[purchaseDate] = purchase.amount + (monthlyRevenue[purchaseDate] || 0);
})
// step 4: rounding numbers
for (let yearMonth in monthlyRevenue) {
    monthlyRevenue[yearMonth] = Number(monthlyRevenue[yearMonth].toFixed(2));
}
console.log(monthlyRevenue);

😅 That was a long ass guide!

Too long? Too detailed? Let me know in the comments! 💬

Homework 🏡

Now that arrays and objects have no secrets for you, here's your assignment.

Create a list of best and worst selling items of each month.
Best and worst are defined by the revenue they generated.

Expected result:

let topFlops = {
  "2022-05": {
    top: 'jacket',
    flop: 'cap'
  },
  "2022-06": {
    top: 'jacket',
    flop: 'cap'
  },
  "2022-07": {
    top: 'jacket',
    flop: 'socks'
  }
}

Happy coding! 💻

Calling all beginners - part I: fun with arrays!

Julien Dephix — Mon, 11 Jul 2022 09:27:40 +0000

Cover photo by Faris Mohammed

Hello, coders! 💻

In this series we're going to work with arrays as it is a concept beginners often struggle with.

Language of choice

JavaScript seems to be quite popular on dev.to so that's what I'll use.
If you'd like to see solutions in your language of choice don't hesitate to post a reply and hopefully someone will post a solution.

About solutions shown here

Bear in mind that solutions are just my take on exercises. There are various ways to implement a solution, some more optimized than others.
If you're reading this and want to offer an alternate solution you're more than welcome to do so in the comments! 💬

Exercise 1

You are given an array of user ages. Your task is to write code to calculate users' average age.

const userAges = [12, 67, 45, 32, 71];

const averageAge = // your code here

console.log(`Average age is ${averageAge}`);

Solution to exercise 1

Click to view my solution

const userAges = [12, 67, 45, 32, 71];

const getAverageAge = (ages) => {
    if (!Array.isArray(ages) || ages.length <= 0) {
        return 0; // or anything you see fit
    }

    return ages.reduce( (total, age) => total + age, 0) / ages.length;
}

const averageAge = getAverageAge(userAges);

console.log(`Average age is ${averageAge}`);

What's going on?

First we add a condition to make sure we are dealing with an array and that it's not empty.
Then we can use reduce to sum all ages.
We divide the result by the number of elements in the array. We can safely do that because we made sure the array was not empty, which means we're not dividing by zero.

Exercise 2

Now instead of an array of ages you start with an array of users.
Your task is the same: write code to calculate users' average age.

const users = [            
    {id: 1, age: 12},
    {id: 2, age: 67},
    {id: 3, age: 45},
    {id: 4, age: 32},
    {id: 5, age: 71}
];

const averageAge = // your code here

console.log(`Average age is ${averageAge}`);

Solution to exercise 2

Click to view my solution

const users = [            
    {id: 1, age: 12},
    {id: 2, age: 67},
    {id: 3, age: 45},
    {id: 4, age: 32},
    {id: 5, age: 71}
];

const getAverageAge = (ages) => {
    if (!Array.isArray(ages) || ages.length <= 0) {
        return 0; // or anything you see fit
    }

    return ages.reduce( (total, age) => total + age, 0) / ages.length;
}

const averageAge = getAverageAge(users.map(user => user.age));

console.log(`Average age is ${averageAge}`);

😕 It looks almost the same as exercise 1!

Developers are lazy so we reuse code whenever we can.
Our getAverageAge function of exercise 1 does the job so why not reuse it?
We could but it needs an array of integers and we have an array of objects (users).
If we could create a new array with just the age property of each user then we'd be golden.
That's where map comes into play. We can use it to create a new array with each user's age:

userAges.map(user => user.age)

and we pass that to getAverageUser and we're done!

Feedback

Let me know in the comments what you think of these kinds of short exercises and I'll try to come up with new ones as I think repetition is important to help understand concepts.

Happy coding! ⌨️

Quiz: SQL query

Julien Dephix — Thu, 30 Jun 2022 16:30:49 +0000

Hello, coders! 💻

Today's quiz taken from a real world task is SQL related.

Database structure

There are two tables (very simplified for the sake of this quiz):

companies with two columns: id and name
users with two columns: company_id and is_pro (boolean)

Task

write a query to find all companies that only have pro users (is_pro=1).

My take on this

Click to see my solution using MySQL

SELECT
    c.name as company
FROM
    users u
JOIN companies c ON
    c.id = u.company_id
GROUP BY
    c.id
HAVING
    SUM(IF(is_pro, 0, 1)) = 0;

IF(is_pro, 0, 1) means that I give 0 points to users that are pros (counter-intuitive but bear with me!). If all users are pros then the sum of points should be 0, right? That's what HAVING IF(is_pro, 0, 1)) = 0 does.

There you have it!

How would you have done it?

📝 Edit
As pointed out by @geraldew using SUM to COUNT is hacky and as such I'd better use COUNT and CASE:

SELECT
    c.name as company,
    COUNT(
        CASE WHEN can_use_app = 1 THEN 'x'
    END
    ) isPro,
    COUNT(
        CASE WHEN can_use_app = 0 OR can_use_app IS NULL THEN 'x'
    END
    ) isNotPro
FROM
    users u
JOIN companies c ON
    c.id = u.company_id
GROUP BY
    c.id
HAVING
    isPro > 0 and isNotPro = 0;

Happy coding! ⌨️

Quiz: in-file string replacement using bash

Julien Dephix — Thu, 30 Jun 2022 10:09:46 +0000

Hello, coders! 💻

Today's quiz taken from a real world task is:

replace a string in a file using bash.

Pre-requisites

We have a dynamically generated variable newPath:
newPath=/new/path/$(date +"%Y%m%d%H%M%S")/file.json

We want to change the value of PATH_TO_FILE with the value of newPath in /app/.env.

File before replacement:

# /app/.env
# there could be lines here
PATH_TO_FILE=/path/to/file.json
# there could be lines here too

File after replacement:

# /app/.env
# there could be lines here
PATH_TO_FILE=/new/path/20220630081234/file.json
# there could be lines here too

Starter script

#!/usr/bin/env bash

newPath=/new/path/$(date +"%Y%m%d%H%M%S")/file.json

# your code here

My take on this

Click to see my solution

#!/usr/bin/env bash

newPath=/new/path/$(date +"%Y%m%d%H%M%S")/file.json

sed -iE "s|(PATH_TO_FILE).*|\1=${newPath}|g" /app/.env

-i option means in place. So any change made to /app/.env will be saved.

-E stands for extended regular expressions.

Traditionally we use / when replacing things : s/to_be_replaced/replacement/g for instance.
Since we're dealing with a file path we would need to escape slashes in newPath but there's a better way: use a different separator. I'm using | here but you can use @ if you prefer.

\1 holds what was captured by the parenthesis which is PATH_TO_FILE in our case.

And that's it!

Comments on how to improve this solution are most welcome.

Happy coding! ⌨️