Introduction to Django 3.0 - Building, Authenticating, and Deploying

Introduction to Django 3.0 - Building, Authenticating, and Deploying


Introduction - Why Python Web Frameworks

Python is a simple, powerful, and open-source language. It allows us to focus more on the problem at hand rather than struggling with the syntax and semantics of the language. There are many Python packages we can use based on our requirements. For example, if we are working on a machine learning problem, we can use PyTorch or TensorFlow; if we are building web applications, we can use Flask, Tornado, or Django. Python provides tools that allow us to integrate different solutions into our application with ease.

In this article, we’ll learn about Django 3.0. This article will equip us with the right amount of information to get the hang of building websites. However, a little knowledge of Python and the web will help us understand it even better. Let’s get started by understanding what Django is all about.

"Python is a simple, powerful, and open-source language."

Tweet This
About Django

Django can build scalable, deployable websites in little time. Once we master Django, it lays down the foundation for us to transform our ideas into applications with ease.

Various top-notch applications and websites such as Instagram, NASA, Pinterest, NatGeo, Dubsmash, and DoorDash are written in Django. It is one of the most-used web frameworks at top companies such as Google, Facebook, Amazon, Oracle, Adobe, and many more. All the documentation and information related to Django can be found on the Django homepage.

Features of Django

The main goal behind developing Django was to quicken the pace of building websites while also satisfying all the requirements of developers by providing the right amount of tools and plugins. Once the initial configuration is done, all we need to concentrate on is the workflow that our website should operate under. It uses an MVT model, which stands for models, views, and templates. We’ll be learning about MVT in future sections. But for now, let’s see what features are in Django:

It’s Superfast: Django takes care of all the files and settings. This allows us to build applications swiftly, right from development to production.
Highly Secure: We tend to constantly worry about the security of our website. Questions like, is the website safe or is it hackable, seem to penetrate our minds. However, in Django, when it comes to security, just sit back and relax. SQL injection, cross-site scripting, cross-site request forgery, and clickjacking are all taken care of.
Easily Scalable: Django can handle huge traffic with zero latency. We can use any cloud platform like GCP, AWS, or Heroku to deploy our Django application. One other advantage is we can integrate machine learning algorithms using Python libraries and still be able to handle it.
Easy Templating: Django is a back-end web framework. To interact with the front end, Django uses Jinja, a simple templating engine that allows us to pass variables from back end to front end. The syntax is quite easy; however, we can also configure our Django project with other templating engines.

There are many more features to motivate us to choose Django as our preferred web framework, but we’ll slowly understand them once we get started with the code!

"Django can handle huge traffic with zero latency!"

Tweet This
What’s new in Django 3.0?

Django 3.0 was released recently in December of 2019). There are only a few main changes as listed below:

Fully Asynchronous: Django 3.0 is now fully async-capable with its new ASGI: an improvement on WSGI.

WSGI succeeded in allowing more freedom and innovation in the Python webspace, and ASGI’s goal is to continue this into the land of asynchronous Python.

If we aren’t sure what synchronous and asynchronous are, with synchronous state, we can run one file at a time, while in the asynchronous state, we can run multiple files at a time.

Python Compatibility: Django 3.0 supports Python versions that are later than 3.5. The Django 2.2.x series is the final version to support Python 3.5.

MariaDB: MariaDB is now supported! In the new 3.0 release, we can now configure our application with MariaDB.

New Model Fields: In Django 3.0, we have more options to classify our models — basically, our database schemas.

These are the core features included in Django 3.0. If we are interested in reading more, please go to the release notes.

Alright, this is a basic introduction for us to understand the capabilities of Django. In the next section, we’ll learn about the installation and configuration of our local machines to get started with code.

Setting Up

In this section, we’ll install Python and Django on our machines.

First things first: Installing Python

Today we will be using macOS and install Python with Brew. It is a package manager, and we can install any package or software with a single command, "brew install package-name." Please use the command from the website https://brew.sh to install Brew. Once Brew is installed, open Terminal from applications and run the following command:

brew install python

At this point, we’ve installed Python 3.

Now to check if the installation is successful or not, in the terminal, type the following command:

python3

We’ll see the Python shell running, similar this:

Python 3.7.6 (default, Jan 29 2020, 13:58:13) 
[Clang 11.0.0 (clang-1100.0.33.16)] on darwin
Type "help", "copyright", "credits" or "license" for more information.

We can exit from the Python shell using the command:

exit()

This exits us from the Python shell and brings us back to the terminal.

In Mac, insert sudo at the beginning of the command to force install. This allows us to run our commands as an administrator or superuser.

Installing Django

Installing Python also installs pip, a Python package installer that installs different Python packages with a single command. Make sure both pip and Python are running on the same version of Python. To check this, use the following commands:

python3 --version
pip3 --version

We should see Python 3.7.6 and pip 19.3.1 from /usr/local/lib/python3.7/site-packages/pip (python 3.7) or something similar returned.

We will be using python3 and pip3 in today's example.

Let’s now install Django. Open the terminal or command prompt and run the following command:

pip3 install django

This command installs the Django package with a lot of other supporting packages. To check if Django was successfully installed, open the terminal followed by a Python shell, and import the Django package using the import keyword. If we see no errors after importing, then the installation is successful; otherwise, try reinstalling it. Next, use the __version__ method as shown in the below snippet. Check the version of the installed Django package. To do all of that, run the following commands:

python3
import django
print(django.__version__)
3.0
Creating a Django project

In this section, we’ll learn how to start a project in Django and understand the Django lifecycle.

Using the codeshare application, we can share code snippets for others to view (as in a blog), prior to which we should authenticate. The authentication will be handled using the Auth0, and we’ll push this application into production using Heroku in the second part of the article.

Building the CodeShare application

We’re now fully equipped with the right instruments to get started with Django. Let’s understand its basics.

Our first step is to create a new Django project. To do so, we need to use the django-admin command, which was installed with Django, followed by startproject and the name of the project. Let’s create a new Django project named codeshare. Be sure to be in the directory of choice, and run this command to create the project:

django-admin startproject codeshare

Next, go inside that particular folder using the following command:

cd codeshare

We will be working in VSCode in today's example. Open up the project in the code editor of choice.

We’ll see a folder named codeshare and a file named manage.py.

The manage.py file is the heart of the django project. To start the application or to make changes in the databases, we need to run this python file. Let’s quickly start our server and check how it works like.

To start the server in Django, use the below command:

python3 manage.py runserver

Here, we’re running manage.py with Python, and runserver is the command-line argument that starts the server. There are other arguments to manage.py, which we’ll learn about in the further sections. Let’s now see what this command does:

Watching for file changes with StatReloader
Performing system checks...
System check identified no issues (0 silenced).
You have 17 unapplied migration(s). Your project may not work properly
until you apply the migrations for app(s): admin, auth, contenttypes, sessions.

Run 'python manage.py migrate' to apply them.
Django version 3.0, using settings 'test.settings'
Starting development server at https://127.0.0.1:8000/
Quit the server with CONTROL-C.

Great, this is the message that Django displays once the server is running. Now open the browser and go to the server link mentioned or localhost:8000. We’ll see a page similar to the screenshot below:

The page rendered is a basic HTML template that comes with Django.

Thus, we’ve understood the manage.py file. Let’s understand how views, URLs, and other files are used. All these are consolidated into the Django lifecycle.

Django Lifecycle

Now that we’ve learned about the Django directory structure and getting the server running, let’s understand what constitutes the Django lifecycle:

Models: Models are like classes; these define the database schema. It can be defined as an object (an object has properties and methods) that needs to be saved in the database.
Views: Views are the functions. It’s where we write our business logic. It takes a web request and returns a web response. Here, we can access the models and define our own functionalities.
Templates: Templates are the basic HTML files; here, we pass our models using views.
URLs: URLs define the paths to pages. For example, we can have a "home" URL pointing to our home page, an "about" URL pointing to an about page, and a "contact" URL to display the contact information in an HTML page.

Django stacks all these into one lifecycle. It defines the possible interactions among them, depicted in the below image:

Imagine a user accessing our website. Firstly, our Django server needs to be up and running. As soon as the user enters a URL, the Django server looks at the view that corresponds to that particular URL. If there is no view against that URL, it returns a 500 Internal server error, else it processes the code written in our view. If there is any model to be used as specified in the view, it requests the model and fetches the data, and later renders that data onto the templates, else it simply returns the template as an HTML page.

This is what happens repeatedly during our website access. It’s as simple as that.

Creating Core app within CodeShare app

That’s not it! To keep the code clean, let’s create an app inside the project. Any number of applications can be created inside a project. In our case, one should suffice. Let’s first change our directory path from the parent folder to the codeshare folder and create our application in there. Execute the command below at the base folder of the project:

python3 manage.py startapp core

We should now have two main folders within our project:

codeshare/codeshare
codeshare/core

After we create an app named core, it’s our responsibility to inform Django about the same. To do so, we need to include it in the codeshare/settings.py file. So in the settings.py file, append 'core' to the INSTALLED_APPS list; below is how it looks like:

# codeshare/settings.py

INSTALLED_APPS = [
    'django.contrib.admin',
    'django.contrib.auth',
    'django.contrib.contenttypes',
    'django.contrib.sessions',
    'django.contrib.messages',
    'django.contrib.staticfiles',
    'core',
]

We’ve got our directory structure ready. However, there are various other things to work on. But what to choose first - Models or Views or URLs or Templates? We will use the following workflow: Models -> Views -> Templates -> URLs

Setting up the database - PostgreSQL

To store all the posts that are to be rendered into the application, we need a database. In our application, any number of users can log in and post their set of blog posts. To keep track of what content is to be rendered, we require a database. How about the type of database that suits our application? There are structured and unstructured databases. We store posts that usually follow the same pattern, say, title, content, author, etc. It seems structured and neat. Therefore, a structured database is the one we shall opt for.

MySQL and PostgreSQL are among the most used structured databases; which is a better option? PostgreSQL is the most preferred because of the ease of integrating it with Django, and the creators of Django have themselves recommended it (If you're not tied to any legacy system and have the freedom to choose a database back end, we recommend PostgreSQL, which achieves a fine balance between cost, features, speed, and stability). So, let’s employ the same in our application.

Make sure to install the PostgreSQL database on our local machine; below are the links to proceed with the installation:

macOS: https://postgresapp.com/

Windows/Linux: https://www.postgresql.org/

Let's go open our codeshare/settings.py file. In the DATABASE section, we'll find the following code:

DATABASES = {
    'default': {
        'ENGINE': 'django.db.backends.sqlite3',
        'NAME': os.path.join(BASE_DIR, 'db.sqlite3'),
    }
}

By default, Django uses "sqlite3" as its database. Let’s modify it to embed PostgreSQL into our codeshare folder. Follow the below code snippet to modify the code:

DATABASES = {
    "default": {
        "ENGINE": "django.db.backends.postgresql_psycopg2",
        "NAME": "codeshare",
        "USER": "",
        "PASSWORD": "",
        "HOST": "",
    }
}

Here we’ve set codeshare as the DB name, but make sure to create a PostgreSQL DB with the same name locally. We can use the below command if we’re using a Mac or Linux machine:

createdb codeshare

Note: Be sure Postgresql is running locally

If we’re using a Windows machine, we need to connect to PSQL first, then type the command:

CREATE DATABASE codeshare;

It looks like:

postgres=# CREATE DATABASE codeshare;

In the next step, let’s create tables in the codeshare database.

Unapplied migration?

Django has a few built-in models: "admin", "auth", "contenttypes", and "sessions". These comprise the predefined table schemas. So to eliminate the error "Unapplied Migrations", which we’ve seen previously, we need to create a database and add these models into tables. Since we’re already created the database, let’s now create tables for these models. To do that, use the following command in the terminal or command prompt:

python3 manage.py migrate

We’ll see a few messages similar to ones below:

**Operations to perform:**

**Apply all migrations:** admin, auth, contenttypes, sessions
**Running migrations:**
… OK
… OK

Now run the server again using the command:

python3 manage.py runserver

We won't come across the "Unapplied Migrations" error. This is how we migrate tables using the migrate argument, and this establishes a connection from our Django environment to PostgreSQL.

Defining the Models

We’re now left with defining our database schema. Let’s quickly go to the codeshare/core/models.py file and create new models in there:

# codeshare/core/models.py

from django.db import models

# Create our models here.
class TimeAuditModel(models.Model):
    """ To path when the record was created and last modified """
    created = models.DateTimeField(auto_now_add=True, verbose_name="Created At",)
    updated = models.DateTimeField(auto_now=True, verbose_name="Last Modified At")

    class Meta:
        abstract = True


class Post(TimeAuditModel):
    post_title = models.CharField(max_length=255)
    post_description = models.TextField()
    post_description_markdown = models.TextField(blank=True)
    post_author = models.CharField(max_length=255)

    def __str__(self):
        return self.post_title

Perfect! This is how our models.py file should look. Let’s discuss in detail the code in it. Initially, we create a new class named TimeAuditModel (an abstract class as mentioned in class 'Meta' - It cannot be used as a normal Django model, but has to be inherited), and this should include the subclass models.Model as it has all the fields and data types required to create Django models. We then initialize two variables inside the TimeAuditModel: the first is created, which corresponds to a DateTimeField. Noticed models.DataTimeField here? This is why we’ve imported models from django.db in the first place. Moving on, we add a variable called updated, which is a DateTimeField as well. These two variables help in acquiring information about the date and time the objects being loaded into the model were created and updated; look at the arguments they take in and we’ll understand how they work.

Next, we create a class named Post which inherits the TimeAuditModel. This inheritance says that the two fields, created and updated, are now a part of the Post model. In this class, we declare all the variables that our codesharing post should consist of. Usually, the code snippet has a title, description, markdown (to store code in markdown format), tags (to add specific tags that relate to the post), feedback, and author. Hence, we initialize them using their corresponding model types. The __str__() method is to define the whole object with a single value being returned in it. Here, post_title defines an object being created using the Post model.

After defining our model, we’re left with the task of adding that in our defined database codeshare. To do so, first, we need to tell Django about the new additions made to our models. For the same, type the below command:

python3 manage.py makemigrations core

This prints the following output on the terminal or command prompt:

Migrations for 'core':
  core/migrations/0001_initial.py
    - Create model Post

This creates a migration file in the folder codeshare/core/migrations. We then migrate the defined models using the below command:

python3 manage.py migrate core

The output then looks like:

This adds our models Post and TimeAuditModel to the database. To check the credibility of these commands, let’s look into Django Admin.

Django Admin

Django Admin provides a user-friendly UI to access the models on the fly. Data is fetched from the models we create and is displayed to the admin. Let’s peek into the admin site now!

Prior to accessing the site, we first need to allow the Django admin to have access to the models we created. To do that, modify the codeshare/core/admin.py file in the core folder.

# codeshare/core/admin.py

from django.contrib import admin
from .models import Post
admin.site.register(Post)

This registers the Post model with the admin. We now have access to view the schema of it. In addition, we can add, delete, or modify the posts. To understand this, let’s see what constitutes the admin GUI.

Type the URL localhost:8000/admin into our browser and a page as displayed below should pop up on our screen:

There are no admin credentials created so far. To create one, type the below command into the terminal or command prompt:

python3 manage.py createsuperuser

The below screenshot shows the next screen that will be displayed:

Make sure to enter all of our details. Now we’re able to access the admin site with the credentials that we’ve just created. Type in the username and password. It will take us to the admin dashboard.

We'll find the models Groups, Users, and Posts in there. Click on Users and we'll find the admin that we just created. Click on Posts and the below snapshot gives us a peek into it.

Note: Make sure the server is running. After creating the superuser, don’t forget to turn on the server.

Click on "Posts" under "CORE" on the home page. Next click on "ADD POST +" in the upper right-hand corner.

Fill in the inputs and click on Save. This creates our codesharing post.

Cheers! We’ve created our first post. Moving on, let’s display these posts on HTML pages for others to look at. This process includes fetching the data from the Post table and rendering it on an HTML page. All this is coded using a View.

Defining the Views

Our models are up and running, and now we need to code the logic to render them onto a particular HTML screen. To do this, we need to collect all the objects that are under the class Post and parse them into an HTML template.

Let’s open the codeshare/core/views.py file and include the below code:

#codeshare/core/views.py


from django.shortcuts import render, redirect
from django.views.generic import View
from .models import Post
class PostListView(View):
    def get(self, request):
        posts = Post.objects.all()
        context = {"posts": posts}
        return render(request, "home.html", context)

Here, we create a ClassBasedView. These are fast and flexible compared to regular function-based views. Our goal is to list all the posts on our home screen. So we define a class named PostListView, which inherits the class View that we import from django.views.generic. This is to support class-based views. Now we write a get() method that takes self and request as arguments.

self is the default argument for every method, and it refers to the object instance being explicitly passed into the method.
request refers to the HTTP request data that is being passed. To be more precise, it’s the WSGI instance.

In our get() method, we create a variable post to store all the Post objects. Hence, we import the Post object from .modelsto collect all the objects of Post class; the query we use is - Post.objects.all().

The next task is to print all these objects onto an HTML page. Hence we use the render() method in Django. This takes in three arguments: request, the name of the HTML file, and a dictionary.

The render function takes in a dictionary to pass the ORM objects into the template.

ORM or Object Relational Mapper transmits information from a relational database to the application at hand. It acts as a middleware to facilitate the interactions between DB and application and eliminates the need to write SQL. The migrations feature embedded in Django is also a part of Django ORM wherein no specific commands are required to create tables along with their respective schemas.

Along the same lines, there’s another concept called Querysets. These define Django’s way of retrieving data from the database. Django ORM supports Querysets to filter the required data from the model. Post.objects.all(), which we've used before, is a Queryset to retrieve all the objects associated with the class Post.

To keep the code clean, we name our dictionary variable as context (in this case, we are sending Post ORM object into the template).

This isn’t the whole story. Haven’t we missed the home.html which we passed in the render() method? This HTML page has to be defined as a template.

Defining the Templates

To add templates to our codeshare directory, we first need to configure it in our codeshare/codeshare/settings.py file (necessary when deploying the application on Heroku). Go ahead and open the settings.py file. We'll find a list named TEMPLATES that has a dictionary inside it. Within the dictionary, we have a key named "DIRS" mapped to an empty list. Add a string called "templates" there.

TEMPLATES = [
    {
        "BACKEND": "django.template.backends.django.DjangoTemplates",
        # below is the updated line
        "DIRS": ["templates"],
        "APP_DIRS": True,
        "OPTIONS": {
            "context_processors": [
                "django.template.context_processors.debug",
                "django.template.context_processors.request",
                "django.contrib.auth.context_processors.auth",
                "django.contrib.messages.context_processors.messages",
            ],
        },
    },
]

Now create a folder named templates in the root directory. All the template files can now be defined in the codeshare/templates directory. Let’s create our home.html there. Go into the newly created file, codeshare/templates/home.html.

Now that we have all the objects being passed in a posts variable, we'll use Jinja templating to render post items onto the HTML. In our case, we print all our post objects using a "for loop". Add in the following code to our newly created file:

<!--codeshare/templates/home.html -->

<html>
  <body>
    {% for i in posts %}
    <h1>{{ i.post_title}}</h1>
    <h6>
      {{i.post_author}}
    </h6>
    <p>{{ i.post_description }}</p>
    <code>
      {{ i.post_description_markdown }}
    </code>
    <p>{{ i.created}}</p>
    <p>{{ i.updated}}</p>
    {% endfor %}
  </body>
</html>

We’re not done yet. We haven’t mentioned any URLs until now. So let’s quickly configure them in the next section.

Defining the URLs

We have defined all our models and views inside the codeshare/core app. So we need to add a file named urls.py in the codeshare/core app where we can manage all our URLs. Before that, in codeshare/codeshare/urls.py, let’s include our core.urls as in the below code snippet:

# codeshare/codeshare/urls.py

from django.contrib import admin
from django.urls import path, include
urlpatterns = [
    path("admin/", admin.site.urls),
    path("", include("core.urls"))
]

The empty string here defines the default route ("/"). So whenever the user opens the home route, the control goes to the codeshare/core/urls.py file. It then accesses those URLs. Now, we need to define URLs in the core app.

Our goal is to render all the posts that we added in the admin interface whenever the user hits the URL localhost:8000/. Hence we design an empty url and attach it to the PostListView that we created before. URLs typically take in two arguments: one is the path, and the other is the view. Since PostListView is a class, we call the method as_view(), which returns a callable view that takes a request and returns a response.

# codeshare/core/urls.py

from django.urls import path
from .views import PostListView
urlpatterns = [
      path("", PostListView.as_view()),
]

Aren’t we keen to look into what we’ve built so far using Django? Go to the browser and type in localhost:8000/ in the URL field. Voila! Our application indeed works awesome as shown below:


?

要查看或添加评论,请登录

Ankita Verma的更多文章

  • Project Management Tools that are used for effective Project Planning

    Project Management Tools that are used for effective Project Planning

    If you have stumbled on this page, then you’re probably wondering “What the heck are Project Management tools?” Let’s…

  • Seven Analytic Tools that Every Business Analyst Should Know

    Seven Analytic Tools that Every Business Analyst Should Know

    from supply chain management to enterprise resource planning, every business department requires relevant data for the…

  • data Analytics Tools

    data Analytics Tools

    In this blog on Best?Big Data Analytics tools, we will learn about Best Data Analytic Tools. Also, will study these…

  • Marketing Analytics Tools That Are Shaping the Industry

    Marketing Analytics Tools That Are Shaping the Industry

    Few marketers would admit to having "an excellent handle" on their analytics. But don't blame it on the marketer Marc…

  • Big Data Analytics Tools

    Big Data Analytics Tools

    Businesses can use data analytics technologies to evaluate large amounts of data in order to gain a competitive…

  • Power BI Architecture

    Power BI Architecture

    power BI is the umbrella term for data visualization, reporting, and analytics tools and services from Microsoft. Power…

  • Different Types of IT Tools

    Different Types of IT Tools

    There are numerous tools aimed at helping IT professionals operate more efficiently and deliver the most effective…

  • 7 convenient communication tools

    7 convenient communication tools

    Following are a handful of such tools. If you have a favorite and don’t see it included here, leave a comment below…

  • Data Analytics Tools

    Data Analytics Tools

    The amount of data produced by humans has exploded to unheard-of levels, with?nearly 2.5 quintillion bytes of…

  • Web Analytics Tools

    Web Analytics Tools

    Web analytics is a way of collecting and analyzing what’s happening on your website to see what your visitors are…

社区洞察

其他会员也浏览了