ray-sh

defmodule Bucket do
    use GenServer

    def start_link(name) do
        GenServer.start_link(__MODULE__, %{}, name: name)
    end

    #callbacks
    def handle_call(request, from, state) do
        case request do

ray-sh

Requirements
1>Two kinds of GenServer 
2>The GenServer will share same process registry module
3>The Registry module's is used to keep the name/pip mapping with the process, so the client will talk with the Registry directly
to get correspoding PID by name. The Registry should monitor the process and keep the name/pid map up-to-date all the time.

How the following work?
1>The My.Registry must implement register_name/unregister_name/whereis_name/send
2>The client start_link function must use name: {:via, Registry, name} since the Registry will keep different kind of process,
we usually use tuple {:server,name} to distinguish

ray-sh

1>Find the max number of a list
2>Remove some elements of a Map
3>Remove some elements of a List
4>Map/reduce

defmodule MyReduce do
  @moduledoc false
  def max_number_of_list(list) do
    Enum.reduce list,List.first(list),fn(x,acc) ->
      if(x > acc) do

ray-sh

Demo how to create supervised pool worker
defmodule SupPoolWorker do
  @moduledoc false
  use Supervisor
  def start_link(name,number) do
    Supervisor.start_link(SupPoolWorker,number,name: name)
  end

  def init(number) do
    children = for i <- 1..number, do: worker(DBWorker,[String.to_atom("Worker#{i}")], id: String.to_atom("Worker#{i}"))

ray-sh

1. How to distiguish the same object?
We use tuple which consist of type and ID
def init({db_folder, pool_size}) do
    processes = for worker_id <- 1..pool_size do
      worker(
        Todo.DatabaseWorker, [db_folder, worker_id],
        id: {:database_worker, worker_id} 
      )
    end

ray-sh

http://www.zsythink.net/archives/1182

ray-sh

1.Demo how to transfor function
2.Demo how to construct function and call it
3.& capture operator will conver Module.Function/arity to lambda
4.& capture operation could be also used to shorten lambda

defmodule FunctionTransfor do
  @moduledoc false
  def display(msg), do: IO.puts msg
  def add1 do
    fn(x)-> x + 1 end

ray-sh

defmodule TupleAndListTest do
  @moduledoc false
  use ExUnit.Case
  
  test "tuple cur" do
    '''
    Tuples are most appropriate to group a small,
    fixed number of elements together. When you need a dynamically sized collection, you can use lists.
    '''
    #Bench mark for tuple and list https://aneta-bielska.github.io/blog/benchmarking-elixir-lists-and-tuples-example.html

ray-sh

#Function used in auth
#1>Save the user id in session and current_user in  conn.assign which will exposed as @current_user for all template

#Deliver pass port to user, user will carry the :user_id in next request
def deliver_passport_to_user(conn,user) do
    conn
    |>assign(:current_user, user)
    |>put_session(:user_id,user.id)
    |>configure_session(renew: true) #Renew the session to keep our safe
end

ray-sh

#Build relation is very common in DB layer, like user and it's posts

#1> We need define the relation between Module user and posts
schema "users" do
  has_many :posts, Blog.Post
end

schema "posts" do
  belongs_to(:user, Blog.User)
end