DEV Community: Pau Riosa

Thank you to all my followers! (and followers to be)

Pau Riosa — Thu, 22 Feb 2024 16:14:54 +0000

I just want to say thank you to all my followers (and followers to be).

It means a lot for me.

Knowing that what I've put on every single time a write has a meaning and value for someone.

Just when I thought to give up, this growing number of people signifies that I should continue what I am doing right now and someday, make a world a better place for learning.

Thanks, guys!

Happy Coding!

How does Ecto.Schema's `has_one/3` works?

Pau Riosa — Mon, 19 Feb 2024 21:26:08 +0000

The problem

I have two schemas, coffees and orders wherein I want to get the recent completed order from each coffees

First solution

defmodule Schemas.Coffee do
    # ... more code
    has_many :completed_orders, Schemas.Order

    def coffee_query(coffee_id) do
        from c in __MODULE__,
            where: c.id == coffee_id,
            preload: [completed_orders: Orders.completed_order],
            limit: 1
    end
end

defmodule Schemas.Order do
    # ... more code

    @completed_status :completed

    def completed_order_query() do
        from o in __MODULE__,
            where: o.status in ^@completed_status,
            order_by: [desc: o.updated_at] 
    end
end

defmodule Contexts.Coffees do
    # more code

    def get_recent_completed_coffee_order(coffee_id) do
     coffee = coffee_id
              |> Schemas.Coffee.coffee_query()
              |> Repo.one

        List.first(coffee.completed_orders)
    end
end

Based from the setup, I can actually get the completed_orders of coffees but the only downside for me is that the preload will return a list
Since it will return a list , that will result into another process for me just to get the recent completed order.

Then, `has_one/3` came into my life

defmodule Schemas.Coffee do
    # ... more code
    has_one :completed_order, Schemas.Order,
        where: [status: :completed],
        preload_order: [desc: :updated_at]

    def coffee_query(coffee_id) do
        from c in __MODULE__,
            where: c.id == ^coffee_id
            preload: [:completed_order],
            limit: 1
    end
end

defmodule Contexts.Coffees do
    # more code

    def get_recent_completed_coffee_order(coffee_id) do
        coffee_id
        |> Schemas.Coffee.coffee_query()
        |> Repo.one
        |> Map.get(:completed_order, nil)
    end
end

What's the magic behind this scheme?

How does `has_one/3` works?

Indicates a one-to-one association with another schema

From the current situation, coffee belongs_to order. As a result, coffee has_many orders, but I only need one result to return coming from the orders schema to identify if the coffee order is completed already.

Using has\_one/3\ I was able to achieve what I want.

has_one/3 - indicates a one-to-one association with another schema.
where - indicates a Filtering associations to get my specific items.
preload_order - Sets the default order_by of the association, and since I am using has_one/3 Ecto set a query LIMIT set to one .

[see has_one/3 on docs](https://hexdocs.pm/ecto/Ecto.Schema.html#has_one/3)

Happy Coding

Things I Learned using Phoenix LiveView in 2024 (Part 2)

Pau Riosa — Sat, 17 Feb 2024 01:05:03 +0000

The beauty of `assign_async/3` and `async_result/1`

# LiveView 
def handle_params(params, _uri, socket) do
    {:noreply, 
         socket
         # ... more assigns here
         |> assign_async_coffees()}
end

# ... more code here

defp assign_async_coffee(socket) do
    assign_async(socket, :coffees, fn -> 
        {:ok, %{coffees: Coffee.load_more_coffees}}
    end)
end

https://hexdocs.pm/phoenix_live_view/Phoenix.LiveView.html#module-async-operations

It allows the user to get a working UI quickly while the system fetches some data in the background or talks to an external service, without blocking the render or event handling
What I liked about the beauty of using assign_async/3 is that everything is asynchronous. I get the chance to process data without blocking the render or event handling
I also have the change to handle different states like the loading, failures and the successful result

You also want to catch any errors or exits and translate it to a meaningful update in the UI rather than crashing the user experience.
And just in case there will any exit errors, I will be able to handle that in the liveview.

The built-in component

# LiveView heex file
<.async_result :let={coffees} assign={@coffees}>
    <:loading>Preparing what are the best coffees for you...</:loading>
    <:failed :let={_failure}>Please be patient, something came up...</:failed>
    <div :for={coffee <- coffees}>
        Here is your coffee <%= coffee %>
    </div>
</.async_result>

assign_async/3 has a partner that I could say a built-in partner wherein you can use it to render the different states that you want to show to your users.
You can even customize the loading and failed component slots to have an even more beautiful UI.

Users can choose nicely from the list of coffees from your application if you seamlessly implemented this kind of approach.

But I wonder does this one have any drawbacks?

That's another thing to find out.

Happy coding

Things I Learned using Phoenix LiveView in 2024

Pau Riosa — Fri, 16 Feb 2024 11:25:12 +0000

Use handle_params/3 more often rather than mount/3 for assigning.

def mount(_params, _session, socket), do: {:ok, socket}
def handle_params(params, _uri, socket) do
   {:noreply,
        socket
        |> assign_here_1(params)
        |> assign_here_2(params)}
end

Use pipelines, it adds more visibility and readability to keep track of what you have added and need more to add.
You can also do pattern matching at handle_params/3. So whenever you need to load or assign special use-cases depending on the parameters, that would be a great thing
And since you are using private functions to assign, you can reuse them whenever you want in different handle_params/3

LiveView as The Source of Truth

# LiveView module
def handle_params(params, _uri, socket) do
  {:noreply,
        socket
        |> assign_here_1(params)
        |> assign_here_2(params)}
end

# functional component
def component(assigns) do
    coffee_for_me = Coffee.get_coffee_for_me(assigns.id)
    # ...some code here
end

Try to create functional components that depends on assigns coming from the liveview.
So instead of doing the code above, try to create something like this

def handle_params(params, _uri, socket) do
  {:noreply,
        socket
        # ...more assign here
        |> assign_coffee(params)}
end

defp assign_coffee(socket, %{"coffee" => coffee} = _params} do
   coffee = Coffee.get_coffee(coffee)
   assign(socket, coffee: coffee)
end

attr(:coffee, Coffee, required: true)
def component(assigns) do
    # ...you can use your coffee here
    <p><%= coffee.title %><p>
end

As much as possible, try to avoid database functions inside your functional components and make Liveview as the source of truth

Happy Coding

What is your win today?

Pau Riosa — Mon, 04 Jul 2022 16:22:07 +0000

Post your daily wins here via comment!

Elixir Today: Create a Left Triangle Pattern using Elixir

Pau Riosa — Wed, 11 May 2022 04:38:07 +0000

Process

create a file named left_triangle.ex
write the code

left_triangle = fn n ->
  for i <- 0..n do
    for _ <- 0..i do
      "*"
    end
  end
  |> Enum.into("", fn string ->
    string = Enum.join(string)
    "#{string}\n"
  end)
end

IO.puts(left_triangle.(10))
IO.puts(left_triangle.(5))

run elixir left_triangle.ex

Result

*
**
***
****
*****
******
*******
********
*********
**********
***********

*
**
***
****
*****
******

Happy Coding!

Elixir Today: Create a Right Triangle Using Elixir

Pau Riosa — Wed, 04 May 2022 07:11:42 +0000

Process

create a file named right_triangle.ex
write the code

right_triangle = fn number ->
  for i when i <= number <- 0..number do
    # for printing spaces
    spaces =
      for j when j < number - i <- 0..number do
        " "
      end

    # for printing star
    star =
      for _k <- 0..i do
        "*"
      end

    spaces ++ star
  end
  |> Enum.into("", fn string ->
    string = Enum.join(string)
    "#{string}\n"
  end)
end

IO.puts(right_triangle.(10))
IO.puts(right_triangle.(5))

run elixir right_triangle.ex

Result

          *
         **
        ***
       ****
      *****
     ******
    *******
   ********
  *********
 **********
***********

     *
    **
   ***
  ****
 *****
******

Happy Coding!

Elixir Today: Create a Hollow Square using Elixir

Pau Riosa — Fri, 29 Apr 2022 06:17:20 +0000

Process

create a file named "hollow.ex"
write the code

hollow = fn number ->
  for i when i < number <- 0..number do
    for j when j < number <- 0..number do
      if i === 0 || i === number - 1 do
        "*"
      else
        if j === 0 || j === number - 1 do
          "*"
        else
          " "
        end
      end
    end
  end
  |> Enum.into("", fn string ->
    string = Enum.join(string)
    "#{string}\n"
  end)
end

IO.puts(hollow.(10))
IO.puts(hollow.(5))

run elixir hollow.ex on your terminal.

Result

**********
*        *
*        *
*        *
*        *
*        *
*        *
*        *
*        *
**********

*****
*   *
*   *
*   *
*****

Happy Coding!

Elixir Today: Creating a Square Pattern using Elixir

Pau Riosa — Wed, 27 Apr 2022 07:21:56 +0000

Process

create an elixir file named square.ex
write the code

square = fn n ->
  string =
    for x when x < n <- 0..n do
      for y when y < n <- 0..n do
        " *"
      end
    end

  result =
    Enum.into(string, "", fn f ->
      new_string = Enum.join(f)
      "#{new_string}\n"
    end)

  result
end

IO.puts(square.(10))
IO.puts(square.(5))

run elixir square.ex

Result

 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *
 * * * * * * * * * *

 * * * * *
 * * * * *
 * * * * *
 * * * * *
 * * * * *

Update

square = fn n ->
  string =
    for _x <- 0..n do
      for _y <- 0..n do
        " *"
      end
    end

  result =
    Enum.into(string, "", fn f ->
      new_string = Enum.join(f)
      "#{new_string}\n"
    end)

  result
end

Happy Coding!

Elixir Today: Combining a List of Map via Specific Field

Pau Riosa — Fri, 08 Apr 2022 08:11:11 +0000

References

references

Intro

Recently, I've been working on a similar question about How to combine a list of map via specific field.

I looked at the elixir forum and gladly I found something. I saw that one of the answers there is using Enum.group_by + Enum.map.

Given my curiosity, How can I come up with a solution using Enum.reduce?

Curiosity Kicks In

So I started working out using Enum.reduce.


 @list [
    %{id: 1, list: [1]},
    %{id: 1, list: [2]},
    %{id: 2, list: [1]},
    %{id: 3, list: [1]},
    %{id: 1, list: [3, 4, 5]},
    %{id: 2, list: [2, 3]},
    %{id: 3, list: [2, 3, 4, 5]},
    %{id: 4, list: [2, 3, 4, 5]},
    %{id: 4, list: [1]}
  ]

    @list
    |> Enum.reverse()
    |> Enum.reduce(%{}, fn %{id: id, list: _list} = f, acc ->
      case acc do
        %{^id => existing} -> Map.put(acc, id, [f | existing])
        %{} -> Map.put(acc, id, [f])
      end
    end)
    |> Enum.map(fn {id, list} -> %{id: id, list: Enum.flat_map(list, & &1.list)} end)

In this way, I created a sample list with specific id and their list.
I reverse the order of the list
I used Enum.reduce to sort out and combine the list of the ids accordingly.
Finally, used Enum.map to transform the data.

Benchmarking!

I started to compare the result between using group_by and reduce. In order to that, I used Benchee.

To install, add benchee to your dependencies.

{:benchee, "~> 1.0", only: :dev}

Given the setup

 @list [
    %{id: 1, list: [1]},
    %{id: 1, list: [2]},
    %{id: 2, list: [1]},
    %{id: 3, list: [1]},
    %{id: 1, list: [3, 4, 5]},
    %{id: 2, list: [2, 3]},
    %{id: 3, list: [2, 3, 4, 5]},
    %{id: 4, list: [2, 3, 4, 5]},
    %{id: 4, list: [1]}
  ]

def groupby(list) do
    Enum.group_by(list, & &1.id)
    |> Enum.map(fn {id, list} -> %{id: id, list: Enum.flat_map(list, & &1.list)} end)
end

def reduce(list) do
    list
    |> Enum.reverse()
    |> Enum.reduce(%{}, fn %{id: id, list: _list} = f, acc ->
      case acc do
        %{^id => existing} -> Map.put(acc, id, [f | existing])
        %{} -> Map.put(acc, id, [f])
      end
    end)
    |> Enum.map(fn {id, list} -> %{id: id, list: Enum.flat_map(list, & &1.list)} end)
end

Then run

Benchee.run(
   %{
        "groupby_enum_map" => fn -> PracticeElixir.groupby(@list) end,
        "reduce_enum_map" => fn -> PracticeElixir.reduce(@list) end
      },
      print: [fast_warning: false]
)

Benchmark Result

Given my dev environment, here is the result.

available_memory: "16 GB",
cpu_speed: "Intel(R) Core(TM) i9-9880H CPU @ 2.30GHz",
elixir: "1.12.3",
erlang: "24.0.6",
num_cores: 16,
os: :macOS

Benchmark suite executing with the following configuration:
warmup: 2 s
time: 5 s
memory time: 0 ns
reduction time: 0 ns
parallel: 1
inputs: none specified
Estimated total run time: 14 s

Benchmarking groupby_enum_map ...
Benchmarking reduce_enum_map ...

Name                       ips        average  deviation         median         99th %
reduce_enum_map         1.08 M        0.93 μs   ±208.21%        0.80 μs           5 μs
groupby_enum_map        0.97 M        1.03 μs   ±251.03%        0.90 μs           4 μs

Comparison:
reduce_enum_map         1.08 M
groupby_enum_map        0.97 M - 1.12x slower +0.109 μs

You can view the full implementation here!

Happy Coding!

Elixir Today: Remove duplicate rows in PostgreSQL using Ecto

Pau Riosa — Wed, 06 Apr 2022 08:18:57 +0000

Intro

In this tutorial, I want to share with you how to remove duplicate rows in your schema using Ecto.

Preparation

In your schema migration, let say you have this schema setup.

CREATE TABLE(:student_teacher, primary_key: false) do

     add :id, :uuid, primary_key: true
     add :student_id, references(:students, type: :uuid)
     add :teacher_id, references(:teachers, type: :uuid)

end

And then let's insert a couple of duplicate record. (Supposed you are using Ex.Machina for your data mock ups )

student = insert(:student)
teacher = insert(:teacher)
insert_list(100, :student_teacher, student: student, teacher: teacher)

Check for Duplicate Rows using IEx

Run iex -S mix in your console and check for duplicate rows.

iex) query = SELECT COUNT(*), student_id, teacher_id FROM 
student_teacher group by student_id, teacher_id having count(*) > 1;

iex) Ecto.Adapter.SQL.query!(Repo, query)

iex) %Postgrex.Result{
  columns: ["count", "student_id", "teacher_id"],
  command: :select
  num_rows: 1,
  rows: [
    [
      100,
      <<student_id>>,
      <<teacher_id>>
    ],

  ]
}

Run Query to delete Duplicate Rows

Repo.transaction(
      fn ->
        query = """
        DELETE FROM student_teacher s1
        USING student_teacher s2
        where s1.id < s2.id
        AND s1.student_id = s2.student_id
        AND s1.teacher_id = s2.teacher_id
        """

        Ecto.Adapters.SQL.query!(Repo, query)
      end,
      timeout: :infinity
    )

Happy Coding!

Elixir Regex: How to put an underscore between a letter and a number?

Pau Riosa — Tue, 05 Apr 2022 11:44:47 +0000

Intro

Today I've been struggling on transforming a camelCase word combined with a number to become a snake_case atom

"helloWorld30" -> :hello_world_30

I tried using Macro.underscore/1 before and found out that I should not use it to manipulate strings to add underscore.

...Do not use it as a general mechanism for underscoring strings as it does not support Unicode or characters that are not valid in Elixir identifiers.

elixir docs

That's why I used Inflex: An Elixir library for handling word inflections

iex) "helloWorld30" 
|> Inflex.underscore()

iex) "hello_world30"

Someone Helped Me Please!

But still using Inflex alone cannot show the output that I want.

Luckily I asked to the community and thank God someone helped me.

Shout out to Karl Seguin 🚀.

He come up on suggesting using RegEx module.

Here is the RegEx that he gave.

iex) Regex.replace(~r/([A-Za-z])(\d)/, "Hello World9000", "\\1_\\2") 

iex) "Hello World_9000"

My Own Version

With another version using String module.


iex) "helloWorld30    
|> String.replace(~r/(\D)(\d)/, "\\1_\\2")

iex) "helloWorld_30"

Overall output

To transform a camelCase word together with a number to become a snake_case atom:

iex) "helloWorld30" 
|> Inflex.underscore() 
|> String.replace(~r/(\D)(\d)/, "\\1_\\2")
|> String.to_atom()


iex) "hello_world_30"

Happy coding!

DEV Community: Pau Riosa

Thank you to all my followers! (and followers to be)

I just want to say thank you to all my followers (and followers to be).

How does Ecto.Schema's `has_one/3` works?

The problem

First solution

Then, has_one/3 came into my life

How does has_one/3 works?

Things I Learned using Phoenix LiveView in 2024 (Part 2)

The beauty of assign_async/3 and async_result/1

The built-in component

Things I Learned using Phoenix LiveView in 2024

Use handle_params/3 more often rather than mount/3 for assigning.

LiveView as The Source of Truth

What is your win today?

Elixir Today: Create a Left Triangle Pattern using Elixir

Process

Result

Elixir Today: Create a Right Triangle Using Elixir

Process

Result

Elixir Today: Create a Hollow Square using Elixir

Process

Result

Elixir Today: Creating a Square Pattern using Elixir

Process

Result

Update

Elixir Today: Combining a List of Map via Specific Field

References

Intro

Curiosity Kicks In

Benchmarking!

Benchmark Result

Elixir Today: Remove duplicate rows in PostgreSQL using Ecto

Intro

Preparation

Check for Duplicate Rows using IEx

Run Query to delete Duplicate Rows

Elixir Regex: How to put an underscore between a letter and a number?

Intro

Someone Helped Me Please!

My Own Version

Overall output

Then, `has_one/3` came into my life

How does `has_one/3` works?

The beauty of `assign_async/3` and `async_result/1`