GitHub

Working with Genie apps (projects)

Working with Genie in an interactive environment can be useful – but usually we want to persist the application and reuse it between sessions. One way to achieve this is to save it as an IJulia notebook and rerun the cells.

However, you can get the best of Genie by working with Genie apps. A Genie app is a MVC (Model-View-Controller) web application which promotes the convention-over-configuration principle. By working with a few predefined files, within the Genie app structure, the framework can lift a lot of weight and massively improve development productivity. By following Genie's workflow, one instantly gets, out of the box, features like automatic module loading and reloading, dedicated configuration files, logging, support for environments, code generators, caching, support for Genie plugins, and much more.

In order to create a new Genie app, we need to run Genie.Generator.newapp($app_name):

julia> Genie.Generator.newapp("MyGenieApp")

Upon executing the command, Genie will:

  • make a new dir called MyGenieApp and cd() into it,
  • install all the app's dependencies,
  • create a new Julia project (adding the Project.toml and Manifest.toml files),
  • activate the project,
  • automatically load the new app's environment into the REPL,
  • start the web server on the default Genie port (port 8000) and host (127.0.0.1 – aka localhost).

At this point you can confirm that everything worked as expected by visiting http://127.0.0.1:8000 in your favourite web browser. You should see Genie's welcome page.

Next, let's add a new route. Routes are used to map request URLs to Julia functions. These functions (also called "handler" functions) provide the response that will be sent back to the client. Routes are meant to be defined in the dedicated routes.jl file (but there is no restriction enforcing this rule). Open MyGenieApp/routes.jl in your editor or run the following command (making sure that you are in the app's directory):

julia> edit("routes.jl")

Append this at the bottom of the routes.jl file and save it:

# routes.jl
route("/hello") do
  "Welcome to Genie!"
end

We are using the route method, passing in the "/hello" URL and an anonymous function which returns the string "Welcome to Genie!". What this means is that for each request to the "/hello" URL, our app will invoke the route handler function and will respond with the welcome message.

Visit http://127.0.0.1:8000/hello for a warm welcome!

Working with resources

Adding our code to the routes.jl file works great for small projects, where you want to quickly publish features on the web. But for larger projects we're better off using Genie's MVC structure (MVC stands for Model-View-Controller). By employing the Model-View-Controller design pattern we can break our code into modules with clear responsibilities: the Model is used for data access, the View renders the response to the client, and the Controller orchestrates the interactions between Models and Views and handles requests. Modular code is easier to write, test and maintain.

A Genie app can be architected around the concept of "resources". A resource represents a business entity (something like a user, or a product, or an account) and maps to a bundle of files (controller, model, views, etc). Resources live under the app/resources/ folder and each resource has its own dedicated folder, where all of its files are hosted. For example, if we have a web app about "books", a "books" folder would be found at app/resources/books and will contain all the files for publishing books on the web (usually called BooksController.jl for the controller, Books.jl for the model, BooksValidator.jl for the model validator – as well as a views folder for hosting all the view files necessary for rendering books data).

HEADS UP

When creating a default Genie app, the app/ folder might be missing. It will be automatically created the first time you add a resource via Genie's generators.

Using Controllers

Controllers are used to orchestrate interactions between client requests, Models (which handle data access), and Views (which are responsible for rendering the responses which will be sent to the clients' web browsers). In a standard workflow, a route points to a method in the controller – which is charged with building and sending the response over the network, back to the client.

Let's add a "books" controller. Genie comes with handy generators and one of them is for creating new controllers:

Generate the Controller

Let's generate our BooksController:

julia> Genie.Generator.newcontroller("Books")
[info]: New controller created at ./app/resources/books/BooksController.jl

Great! Let's edit BooksController.jl (julia> edit("./app/resources/books/BooksController.jl")) and add something to it. For example, a function which returns some of Bill Gates' recommended books would be nice. Make sure that BooksController.jl looks like this:

# app/resources/books/BooksController.jl
module BooksController

struct Book
  title::String
  author::String
end

const BillGatesBooks = Book[
  Book("The Best We Could Do", "Thi Bui"),
  Book("Evicted: Poverty and Profit in the American City", "Matthew Desmond"),
  Book("Believe Me: A Memoir of Love, Death, and Jazz Chickens", "Eddie Izzard"),
  Book("The Sympathizer", "Viet Thanh Nguyen"),
  Book("Energy and Civilization, A History", "Vaclav Smil")
]

function billgatesbooks()
  "
  <h1>Bill Gates' list of recommended books</h1>
  <ul>
    $(["<li>$(book.title) by $(book.author)</li>" for book in BillGatesBooks]...)
  </ul>
  "
end

end

Our controller is just a plain Julia module where we define a Book struct and set up an array of book objects. We then define a function, billgatesbooks, which returns an HTML string, with a H1 heading and an unordered list of all the books. We used an array comprehension to iterate over each book and render it in a <li> element. The elements of the array are then concatenated using the splat ... operator.

The plan is to map this function to a route and expose it on the internet.

Checkpoint

Before exposing it on the web, we can test the function in the REPL:

julia> using MyGenieApp.BooksController

julia> BooksController.billgatesbooks()

The output of the function call should be a HTML string which looks pretty much like this:

"<h1>Bill Gates' list of recommended books</h1><ul><li>The Best We Could Do by Thi Bui</li><li>Evicted:
Poverty and Profit in the American City by Matthew Desmond</li><li>Believe Me: A Memoir of Love, Death, and Jazz
Chickens by Eddie Izzard</li><li>The Sympathizer by Viet Thanh Nguyen</li><li>Energy and Civilization, A History by
Vaclav Smil</li></ul>"

Please make sure that it works as expected.

Setup the route

Now, let's expose our billgatesbooks method on the web. We need to add a new route which points to it. Add these to the routes.jl file:

# routes.jl
using Genie
using MyGenieApp.BooksController

route("/bgbooks", BooksController.billgatesbooks)

In the snippet we declared that we're using MyGenieApp.BooksController. Do notice that Genie automatically included the module as there was no need to explicitly include the file – and in addition Genie will reload the source code every time we change it. Next we defined a route mapping the /bgbooks URL and the BooksController.billgatesbooks function (we say that the BooksController.billgatesbooks is the route handler for the /bgbooks URL or endpoint).

That's all! If you now visit http://localhost:8000/bgbooks you'll see Bill Gates' list of recommended books (well, at least some of them, the man reads a lot!).

PRO TIP

If you would rather work with Julia instead of wrangling HTML strings, you can use Genie's Renderer.Html API. It provides functions which map every standard HTML element. For instance, the BooksController.billgatesbooks function can be written as follows, as an array of HTML elements:

using Genie.Renderer.Html

function billgatesbooks()
  [
    h1() do
      "Bill Gates' list of recommended books"
    end
    ul() do
      for_each(BillGatesBooks) do book
        li() do
          book.title * " by " * book.author
        end
      end
    end
  ]
end

The for_each function iterates over a collection of elements and concatenates the output of each loop into the result of the loop. We'll talk about it more soon.

Adding views

However, putting HTML into the controllers is a bad idea: HTML should stay in the dedicated view files and contain as little logic as possible. Let's refactor our code to use views instead.

The views used for rendering a resource should be placed inside the views/ folder, within that resource's own folder structure. So in our case, we will add an app/resources/books/views/ folder. Just go ahead and do it, Genie does not provide a generator for this task:

julia> mkdir(joinpath("app", "resources", "books", "views"))

We created the views/ folder in app/resources/books/. We provided the full path as our REPL is running in the the root folder of the app. Also, we use the joinpath function so that Julia creates the path in a cross-platform way.

Naming views

Usually each controller method will have its own rendering logic – hence, its own view file. Thus, it's a good practice to name the view files just like the methods, so that we can keep track of where they're used.

At the moment, Genie supports HTML and Markdown view files, as well as plain Julia. Their type is identified by file extension so that's an important part. The HTML views use a .jl.html extension while the Markdown files go with .jl.md and the Julia ones by .jl.

HTML views

All right then, let's add our first view file for the BooksController.billgatesbooks method. Let's create an HTML view file. With Julia:

julia> touch(joinpath("app", "resources", "books", "views", "billgatesbooks.jl.html"))

Genie supports a special type of dynamic HTML view, where we can embed Julia code. These are high performance compiled views. They are not parsed as strings: instead, the HTML is converted to a Julia function that uses native Julia rendering code and is cached to the file system and loaded like any other Julia file. Hence, the first time you load a view, or after you change one, you might notice a certain delay – it's the time needed to generate, compile and load the view. On next runs (especially in production) it's going to be blazing fast!

In case you're curious, the auto-generated Julia view functions are stored by default in the build/ folder in the app. Feel free to take a look. The build/ folder can be safely deleted, Genie will create it back as needed. The location of the build/ folder is configurable and can be changed from Genie's config.

Now all we need to do is to move the HTML code out of the controller and into the view, improving it a bit to also show a count of the number of books. Edit the view file as follows (julia> edit("app/resources/books/views/billgatesbooks.jl.html")):

<!-- billgatesbooks.jl.html -->
<h1>Bill Gates' top $(length(books)) recommended books</h1>
<ul>
  <% for_each(books) do book %>
    <li>$(book.title) by $(book.author)</li>
  <% end %>
</ul>

As you can see, it's just plain HTML with embedded Julia. We can add Julia code by using the <% ... %> code block tags – these should be used for more complex, multiline expressions. Or by using plain Julia string interpolation with $(...) – for simple values outputting.

To make HTML generation more efficient, Genie provides a series of helpers, like the above for_each macro which allows iterating over a collection, passing the current item into the processing function.

Rendering views

We now need to refactor our controller to use the view, passing in the expected variables. We will use the html method which renders and outputs the response as HTML. Update the definition of the billgatesbooks function to be as follows:

# BooksController.jl
using Genie.Renderer.Html

function billgatesbooks()
  html(:books, :billgatesbooks, books = BillGatesBooks)
end

First, notice that we needed to add Genie.Renderer.Html as a dependency, to get access to the html method. As for the html method itself, it takes as its arguments the name of the resource, the name of the view file, and a list of keyword arguments representing view variables:

  • :books is the name of the resource (which effectively indicates in which views folder Genie should look for the

view file – in our case app/resources/books/views);

  • :billgatesbooks is the name of the view file. We don't need to pass the extension, Genie will figure it out since

there's only one file with this name;

  • and finally, we pass the values we want to expose in the view, as keyword arguments.

That's it – our refactored app should be ready! You can try it out for yourself at http://localhost:8000/bgbooks

Markdown views

Markdown views work similar to HTML views – employing the same embedded Julia functionality. Here is how you can add a Markdown view for our billgatesbooks function.

First, create the corresponding view file, using the .jl.md extension. Maybe with:

julia> touch(joinpath("app", "resources", "books", "views", "billgatesbooks.jl.md"))

Now edit the file and make sure it looks like this:

<!-- app/resources/books/views/billgatesbooks.jl.md -->
# Bill Gates' $(length(books)) recommended books

$(
  for_each(books) do book
    "* $(book.title) by $(book.author) \n"
  end
)

Notice that Markdown views do not support Genie's HTML embedded Julia tags <% ... %>. Only string interpolation $(...) is accepted, but it works across multiple lines.

If you reload the page now, however, Genie will still load the HTML view. The reason is that, if we have only one view file, Genie will manage. But if there's more than one, the framework won't know which one to pick. It won't error out but will pick the default one, which is the HTML version.

It's a simple change in the BookiesController: we have to explicitly tell Genie which file to load, extension and all:

# BooksController.jl
function billgatesbooks()
  html(:books, "billgatesbooks.jl.md", books = BillGatesBooks)
end

Taking advantage of layouts

Genie's views are rendered within a layout file. Layouts are meant to render the theme of the website, or the "frame" around the view – the elements which are common on all the pages. The layout file can include visible elements, like the main menu or the footer. But also maybe the <head> tag or the assets tags (<link> and <script> tags for loading CSS and JavaScript files in all the pages).

Every Genie app has a main layout file which is used by default – it can be found in app/layouts/ and is called app.jl.html. It looks like this:

<!-- app/layouts/app.jl.html -->
<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="utf-8" />
    <title>Genie :: The highly productive Julia web framework</title>
    <!-- link rel="stylesheet" href="/css/application.css" / -->
  </head>
  <body>
    <%
      @yield
    %>
    <!-- script src="/js/application.js"></script -->
  </body>
</html>

We can edit it. For example, add this right under the opening <body> tag, just above the <% tag:

<h1>Welcome to top books</h1>

If you reload the page at http://localhost:8000/bgbooks you will see the new heading.

But we don't have to stick to the default; we can add additional layouts. Let's suppose that we have, for example, an admin area which should have a completely different theme. We can add a dedicated layout for that:

julia> touch(joinpath("app", "layouts", "admin.jl.html"))

Now edit it (julia> edit("app/layouts/admin.jl.html")) and make it look like this:

<!-- app/layouts/admin.jl.html -->
<!DOCTYPE html>
<html lang="en">
  <head>
    <title>Genie Admin</title>
  </head>
  <body>
    <h1>Books admin</h1>
    <%
      @yield
    %>
  </body>
</html>

If we want to apply it, we must instruct our BooksController to use it. The html function takes a keyword argument named layout, for the layout file. Update the billgatesbooks function to look like this:

# BooksController.jl
function billgatesbooks()
  html(:books, :billgatesbooks, books = BillGatesBooks, layout = :admin)
end

Reload the page and you'll see the new heading.

The @yield instruction

There is a special instruction in the layouts: @yield. It outputs the contents of the view as rendered through the controller. So where this macro is present, Genie will output the HTML resulting from rendering the view by executing the route handler function within the controller.

Using view paths

For very simple applications the MVC and the resource-centric approaches might involve too much boilerplate. In such cases, we can simplify the code by referencing the view (and layout) by file path, ex:

# BooksController.jl
using Genie.Renderer

function billgatesbooks()
  html(path"app/resources/books/views/billgatesbooks.jl.html", books = BillGatesBooks, layout = path"app/layouts/app.jl.html")
end

Rendering JSON views

A common use case for web apps is to serve as backends for RESTful APIs. For such cases, JSON is the preferred data format. You'll be happy to hear that Genie has built-in support for JSON responses. Let's add an endpoint for our API – which will render Bill Gates' books as JSON.

We can start in the routes.jl file, by appending this

route("/api/v1/bgbooks", BooksController.API.billgatesbooks)

Next, in BooksController.jl, append the extra logic at the end of the file, before the closing end. The whole file should look like this:

# BooksController.jl
module BooksController

using Genie.Renderer.Html

struct Book
  title::String
  author::String
end

const BillGatesBooks = Book[
  Book("The Best We Could Do", "Thi Bui"),
  Book("Evicted: Poverty and Profit in the American City", "Matthew Desmond"),
  Book("Believe Me: A Memoir of Love, Death, and Jazz Chickens", "Eddie Izzard"),
  Book("The Sympathizer!", "Viet Thanh Nguyen"),
  Book("Energy and Civilization, A History", "Vaclav Smil")
]

function billgatesbooks()
  html(:books, :billgatesbooks, layout = :admin, books = BillGatesBooks)
end


module API

using ..BooksController
using Genie.Renderer.Json

function billgatesbooks()
  json(BooksController.BillGatesBooks)
end

end

end

We nested an API module within the BooksController module, where we defined another billgatesbooks function which outputs a JSON.

If you go to http://localhost:8000/api/v1/bgbooks it should already work as expected.

JSON views

However, we have just committed one of the cardinal sins of API development. We have just forever coupled our internal data structure to its external representation. This will make future refactoring very complicated and error prone as any changes in the data will break the client's integrations. The solution is to, again, use views, to fully control how we render our data – and decouple the data structure from its rendering on the web.

Genie has support for JSON views – these are plain Julia files which have the ".json.jl" extension. Let's add one in our views/ folder:

julia> touch(joinpath("app", "resources", "books", "views", "billgatesbooks.json.jl"))

We can now create a proper response. Put this in the view file:

# app/resources/books/views/billgatesbooks.json.jl
"Bill Gates' list of recommended books" => books

Final step, instructing BooksController to render the view. Simply replace the existing billgatesbooks function within the API sub-module with the following:

function billgatesbooks()
  json(:books, :billgatesbooks, books = BooksController.BillGatesBooks)
end

This should hold no surprises – the json function is similar to the html one we've seen before. So now we're rendering a custom JSON response. That's all – everything should work!

HEADS UP

Why JSON views have the extension ending in .jl but HTML and Markdown views do not?

Good question! The extension of the views is chosen in order to preserve correct syntax highlighting in the IDE/code editor.

Since practically HTML and Markdown views are HTML and Markdown files with some embedded Julia code, we want to use the HTML or Markdown syntax highlighting. For JSON views, we use pure Julia, so we want Julia syntax highlighting.

Accessing databases with SearchLight models

You can get the most out of Genie by pairing it with its seamless ORM layer, SearchLight. SearchLight, a native Julia ORM, provides excellent support for working with relational databases. The Genie + SearchLight combo can be used to productively develop CRUD (Create-Read-Update-Delete) apps.

HEADS UP

CRUD stands for Create-Read-Update-Delete and describes the data workflow in many web apps, where resources are created, read (listed), updated, and deleted.

SearchLight represents the "M" part in Genie's MVC architecture (thus, the Model layer).

Let's begin by adding SearchLight to our Genie app. All Genie apps manage their dependencies in their own Julia environment, through their Project.toml and Manifest.toml files.

So we need to make sure that we're in pkg> mode first (which is entered by typing ] in julian mode, ie: julia>]). The cursor should change to (MyGenieApp) pkg>.

Next, we add SearchLight:

(MyGenieApp) pkg> add SearchLight

Adding a database adapter

SearchLight provides a database agnostic API for working with various backends (at the moment, MySQL, SQLite, Postgres and Oracle). Thus, we also need to add the specific adapter. To keep things simple, let's use SQLite for our app. Hence, we'll need the SearchLightSQLite package:

(MyGenieApp) pkg> add SearchLightSQLite

Setup the database connection

Genie is designed to seamlessly integrate with SearchLight and provides access to various database oriented generators. First we need to tell Genie/SearchLight how to connect to the database. Let's use them to set up our database support. Run this in the Genie/Julia REPL:

julia> Genie.Generator.db_support()

The command will add a db/ folder within the root of the app. What we're looking for is the db/connection.yml file which tells SearchLight how to connect to the database. Let's edit it. Make the file look like this:

env: ENV["GENIE_ENV"]

dev:
  adapter: SQLite
  database: db/books.sqlite
  config:

This instructs SearchLight to run in the environment of the current Genie app (by default dev), using SQLite for the adapter (backend) and a database stored at db/books.sqlite (the database will be created automatically if it does not exist). We could pass extra configuration options in the config object, but for now we don't need anything else.

HEADS UP

If you are using a different adapter, make sure that the database configured already exists and that the configured user can successfully access it – SearchLight will not attempt to create the database.

Now we can ask SearchLight to load it up:

julia> using SearchLight

julia> SearchLight.Configuration.load()
Dict{String,Any} with 4 entries:
  "options"  => Dict{String,String}()
  "config"   => nothing
  "database" => "db/books.sqlite"
  "adapter"  => "SQLite"

Let's just go ahead and try it out by connecting to the DB:

julia> using SearchLightSQLite

julia> SearchLight.Configuration.load() |> SearchLight.connect

This should return a SQLite.DB("db/books.sqlite") object. If that is the case, the connection succeeded and we got back a SQLite database handle.

PRO TIP

Each database adapter exposes a CONNECTIONS collection where we can access the connection:

julia> SearchLightSQLite.CONNECTIONS
1-element Array{SQLite.DB,1}:
 SQLite.DB("db/books.sqlite")

Awesome! If all went well you should have a books.sqlite database in the db/ folder.

shell> tree db
db
├── books.sqlite
├── connection.yml
├── migrations
└── seeds

Managing the database schema with SearchLight migrations

Database migrations provide a way to reliably, consistently and repeatedly apply (and undo) changes to the structure of your database (known as "schema transformations"). They are specialised scripts for adding, removing and altering DB tables – these scripts are placed under version control and are managed by a dedicated system which knows which scripts have been run and which not, and is able to run them in the correct order.

SearchLight needs its own DB table to keep track of the state of the migrations so let's set it up:

julia> SearchLight.Migrations.init()
[ Info: Created table schema_migrations

This command sets up our database with the needed table in order to manage migrations.

PRO TIP

You can use the SearchLight API to execute random queries against the database backend. For example we can confirm that the table is really there:

julia> SearchLight.query("SELECT name FROM sqlite_master WHERE type ='table' AND name NOT LIKE 'sqlite_%'")
┌ Info: SELECT name FROM sqlite_master WHERE type ='table' AND name NOT LIKE 'sqlite_%'
└

1×1 DataFrames.DataFrame
│ Row │ name              │
│     │ String⍰           │
├─────┼───────────────────┤
│ 1   │ schema_migrations │

The result is a DataFrame object.

Creating our Book model

SearchLight, just like Genie, uses the convention-over-configuration design pattern. It prefers for things to be setup in a certain way and provides sensible defaults, versus having to define everything in extensive configuration files. And fortunately, we don't even have to remember what these conventions are, as SearchLight also comes with an extensive set of generators.

Lets ask SearchLight to create a new model:

julia> SearchLight.Generator.newresource("Book")

[ Info: New model created at /Users/adrian/Dropbox/Projects/MyGenieApp/app/resources/books/Books.jl
[ Info: New table migration created at /Users/adrian/Dropbox/Projects/MyGenieApp/db/migrations/2020020909574048_create_table_books.jl
[ Info: New validator created at /Users/adrian/Dropbox/Projects/MyGenieApp/app/resources/books/BooksValidator.jl
[ Info: New unit test created at /Users/adrian/Dropbox/Projects/MyGenieApp/test/books_test.jl

SearchLight has created the Books.jl model, the *_create_table_books.jl migration file, the BooksValidator.jl model validator and the books_test.jl test file.

HEADS UP

The first part of the migration file will be different for you!

The *_create_table_books.jl file will be named differently as the first part of the name is the file creation timestamp. This timestamp part guarantees that names are unique and file name clashes are avoided (for example when working as a team a creating similar migration files).

Writing the table migration

Lets begin by writing the migration to create our books table. SearchLight provides a powerful DSL for writing migrations. Each migration file needs to define two methods: up which applies the changes – and down which undoes the effects of the up method. So in our up method we want to create the table – and in down we want to drop the table.

The naming convention for tables in SearchLight is that the table name should be pluralized (books) – because a table contains multiple books (and each row represents an object, a single book). But don't worry, the migration file should already be pre-populated with the correct table name.

Edit the db/migrations/*_create_table_books.jl file and make it look like this:

module CreateTableBooks

import SearchLight.Migrations: create_table, column, primary_key, add_index, drop_table

function up()
  create_table(:books) do
    [
      primary_key()
      column(:title, :string, limit = 100)
      column(:author, :string, limit = 100)
    ]
  end

  add_index(:books, :title)
  add_index(:books, :author)
end

function down()
  drop_table(:books)
end

end

The code is pretty easy to follow: in the up function we call create_table and pass an array of columns: a primary key, a title column and an author column (both strings have a max length of 100). We also add two indices (one on the title and the other on the author columns). As for the down method, it invokes the drop_table function to remove the table.

Running the migration

We can see what SearchLight knows about our migrations with the SearchLight.Migrations.status command:

julia> SearchLight.Migrations.status()
|   | Module name & status                   |
|   | File name                              |
|---|----------------------------------------|
|   |                 CreateTableBooks: DOWN |
| 1 | 2020020909574048_create_table_books.jl |

So our migration is in the down state – meaning that its up method has not been run. We can easily fix this:

julia> SearchLight.Migrations.last_up()
[ Info: Executed migration CreateTableBooks up

If we recheck the status, the migration is up:

julia> SearchLight.Migrations.status()
|   | Module name & status                   |
|   | File name                              |
|---|----------------------------------------|
|   |                   CreateTableBooks: UP |
| 1 | 2020020909574048_create_table_books.jl |

Our table is ready!

Defining the model

Now it's time to edit our model file at app/resources/books/Books.jl. Another convention in SearchLight is that we're using the pluralized name (Books) for the module – because it's for managing multiple books. And within it we define a type (a mutable struct), called Book – which represents an item (a single book) which maps to a row in the underlying database.

Edit the Books.jl file to make it look like this:

# Books.jl
module Books

import SearchLight: AbstractModel, DbId, save!
import Base: @kwdef

export Book

@kwdef mutable struct Book <: AbstractModel
  id::DbId = DbId()
  title::String = ""
  author::String = ""
end

end

We defined a mutable struct which matches our previous Book type by using the @kwdef macro, in order to also define a keyword constructor, as SearchLight needs it.

Using our model

To make things more interesting, we should import our current books into the database. Add this function to the Books.jl module, under the Book() constructor definition (just above the module's closing end):

# Books.jl
function seed()
  BillGatesBooks = [
    ("The Best We Could Do", "Thi Bui"),
    ("Evicted: Poverty and Profit in the American City", "Matthew Desmond"),
    ("Believe Me: A Memoir of Love, Death, and Jazz Chickens", "Eddie Izzard"),
    ("The Sympathizer!", "Viet Thanh Nguyen"),
    ("Energy and Civilization, A History", "Vaclav Smil")
  ]

  for b in BillGatesBooks
    Book(title = b[1], author = b[2]) |> save!
  end
end

Auto-loading the DB configuration

Now, to try things out. Genie takes care of loading all our resource files for us when we load the app. To do this, Genie comes with a special file called an initializer, which automatically loads the database configuration and sets up SearchLight. Check config/initializers/searchlight.jl to see how this is done.

Heads up!

All the *.jl files placed into the config/initializers/ folder are automatically included by Genie upon starting the Genie app. They are included early (upon initialisation), before the controllers, models, views, are loaded.

Trying it out

Now it's time to restart our REPL session and test our app. Close the Julia REPL session to exit to the OS command line and run:

$ bin/repl

On Windows you will need to run bin/repl.bat instead.

The repl executable script placed within the app's bin/ folder starts a new Julia REPL session and loads the applications' environment. Everything should be automatically loaded now, DB configuration included - so we can invoke the previously defined seed function to insert the books:

julia> using MyGenieApp.Books

julia> Books.seed()

There should be a list of queries showing how the data is inserted in the DB:

julia> Books.seed()
[ Info: INSERT  INTO books ("title", "author") VALUES ('The Best We Could Do', 'Thi Bui')
[ Info: INSERT  INTO books ("title", "author") VALUES ('Evicted: Poverty and Profit in the American City', 'Matthew Desmond')
# output truncated

If you want to make sure all went right (although trust me, it did, otherwise SearchLight would've thrown an Exception!), just ask SearchLight to retrieve the books we inserted:

julia> using SearchLight

julia> all(Book)
[ Info: 2020-02-09 13:29:32 SELECT "books"."id" AS "books_id", "books"."title" AS "books_title", "books"."author" AS "books_author" FROM "books" ORDER BY books.id ASC

5-element Array{Book,1}:
 Book
| KEY            | VALUE                |
|----------------|----------------------|
| author::String | Thi Bui              |
| id::DbId       | 1                    |
| title::String  | The Best We Could Do |

 Book
| KEY            | VALUE                                            |
|----------------|--------------------------------------------------|
| author::String | Matthew Desmond                                  |
| id::DbId       | 2                                                |
| title::String  | Evicted: Poverty and Profit in the American City |

# output truncated

The SearchLight.all method returns all the Book items from the database.

All good!

The next thing we need to do is to update our controller to use the model. Make sure that app/resources/books/BooksController.jl reads like this:

# BooksController.jl
module BooksController

using Genie.Renderer.Html
using SearchLight
using MyGenieApp.Books

function billgatesbooks()
  html(:books, :billgatesbooks, books = all(Book))
end

module API

using ..BooksController
using Genie.Renderer.Json
using SearchLight
using MyGenieApp.Books

function billgatesbooks()
  json(:books, :billgatesbooks, books = all(Book))
end

end

end

Our JSON view needs a bit of tweaking too:

# app/resources/books/views/billgatesbooks.json.jl
"Bill's Gates list of recommended books" => [Dict("author" => b.author, "title" => b.title) for b in books]

Now if we just start the server we'll be able to see the list of books served from the database:

# Start the server
julia> up()

The up method starts up the web server and takes us back to the interactive Julia REPL prompt.

Now, if, for example, we navigate to http://localhost:8000/api/v1/bgbooks, the output should match the following JSON document:

{
  "Bill's Gates list of recommended books": [
    {
      "author": "Thi Bui",
      "title": "The Best We Could Do"
    },
    {
      "author": "Matthew Desmond",
      "title": "Evicted: Poverty and Profit in the American City"
    },
    {
      "author": "Eddie Izzard",
      "title": "Believe Me: A Memoir of Love, Death, and Jazz Chickens"
    },
    {
      "author": "Viet Thanh Nguyen",
      "title": "The Sympathizer!"
    },
    {
      "author": "Vaclav Smil",
      "title": "Energy and Civilization, A History"
    }
  ]
}

Let's add a new book to see how it works. We'll create a new Book item and persist it using the SearchLight.save! method:

julia> newbook = Book(title = "Leonardo da Vinci", author = "Walter Isaacson")

Book
| KEY            | VALUE             |
|----------------|-------------------|
| author::String | Walter Isaacson   |
| id::DbId       | NULL              |
| title::String  | Leonardo da Vinci |


julia> save!(newbook)

[ Info: INSERT  INTO books ("title", "author") VALUES ('Leonardo da Vinci', 'Walter Isaacson')
[ Info: ; SELECT CASE WHEN last_insert_rowid() = 0 THEN -1 ELSE last_insert_rowid() END AS id
[ Info: SELECT "books"."id" AS "books_id", "books"."title" AS "books_title", "books"."author" AS "books_author" FROM "books" WHERE "id" = 6 ORDER BY books.id ASC

Book
| KEY            | VALUE             |
|----------------|-------------------|
| author::String | Walter Isaacson   |
| id::DbId       | 6                 |
| title::String  | Leonardo da Vinci |

Calling the save! method, SearchLight has persisted the object in the database and then retrieved it and returned it (notice the updated id::DbId field).

The same save! operation can be written as a one-liner:

julia> Book(title = "Leonardo da Vinci", author = "Walter Isaacson") |> save!

HEADS UP

If you also run the one-liner save! example, it will add the same book again. No problem, but if you want to remove it, you can use the delete method:

julia> duplicated_book = find(Book, title = "Leonardo da Vinci")[end]

julia> delete(duplicated_bookd)
[ Info: DELETE FROM books WHERE id = '7'

Book
| KEY            | VALUE             |
|----------------|-------------------|
| author::String | Walter Isaacson   |
| id::DbId       | NULL              |
| title::String  | Leonardo da Vinci |

If you reload the page at http://localhost:8000/bgbooks the new book should show up.

{
  "Bill's Gates list of recommended books": [
    {
      "author": "Thi Bui",
      "title": "The Best We Could Do"
    },
    {
      "author": "Matthew Desmond",
      "title": "Evicted: Poverty and Profit in the American City"
    },
    {
      "author": "Eddie Izzard",
      "title": "Believe Me: A Memoir of Love, Death, and Jazz Chickens"
    },
    {
      "author": "Viet Thanh Nguyen",
      "title": "The Sympathizer!"
    },
    {
      "author": "Vaclav Smil",
      "title": "Energy and Civilization, A History"
    },
    {
      "author": "Walter Isaacson",
      "title": "Leonardo da Vinci"
    }
  ]
}

PRO TIP

SearchLight exposes two similar data persistence methods: save! and save. They both perform the same action (persisting the object to the database), but save will return a Bool of value true to indicate that the operation was successful or a Bool of value false to indicate that the operation has failed. While the save! variant will return the persisted object upon success or will throw an exception on failure.

Congratulations

You have successfully finished the first part of the step by step walkthrough - you have mastered the Genie basics, allowing you to set up a new app, register routes, add resources (controllers, models, and views), add database support, version the database schema with migrations, and execute basic queries with SearchLight!

In the next part we'll look at more advanced topics like handling forms and file uploads, templates rendering, interactivity and more.