Miniter[05]Making an API Using Django without REST framework I

Even though I finished a tutorial on djangoproject.com, it was still a struggle to even start an assignment/project on back-end because I have never done it before. One of my colleagues who had some experience with back-end development went ahead of me. Looking at his endproduct gave me an idea of what I am suppsoed to do.

#1. Virtual Environment with Miniconda I used Miniconda to create a virtual environment to do my back-end project. Detailed instruction how to install Miniconda is here

Virtual environment helps to keep dependencies required by different projects especially you have to use different versions of python or django for each project.

#create a virtual environment
conda create -n environment_name python=python_version anaconda

#activate a virtual environment
source activate environment_name

#deactivate the virtual environment
source deactivate

#remove the virtual environment if no longer needed
conda remove -n environment_name -all

#2. Understanding different “.py” files I got to understan what each .py file does by working on this project. manage.py: a command-line utility. you can run different command lines such as makemigrations, migrate, shell, and runserver (these are pretty much all I had to use until I ran into a problem, which I will explain later in this post)
settings.py: settings/configurations for this Django project
urls.py: the URL declaration for the project
models.py: define each model that this Django project needs. model is like class in python which helps with creating objects
views.py: takes a web request and returns a web response

#3. Making objects You can use either python manage.py shell command or create an object direclty from the admin page. Objects are created according to how class is defined in models.pyI tried both ways since I am only working on the back-end side right now.

## models.py

def user_list(request):
    user_list = []
    for user in  User.objects.all():
        user_list.append({
            'user_text': user.user_text,
            'name': user.name,
            'date': user.date,
            'content': user.content,
            'password': user.password,
        })
    return JsonResponse(user_list, safe=False)

## views.py

def user_list(request):
    user_list = []
    for user in  User.objects.all():
        user_list.append({
            'user_text': user.user_text,
            'name': user.name,
            'date': user.date,
            'content': user.content,
            'password': user.password,
        })
    return JsonResponse(user_list, safe=False)

python manage.py shell

from api.models import User  # importing the class created/written in models.py

user = User(user_text="jkang14", name="jason", ..., content="first tweet", password="firstpassword")    # assign attributes as needed

user.save()      # save above information as an object, which gets appended to a list

#4. Changing attributes I ran into a problem when I added a password attribute to the User class. I created an object before adding the attribute, so the table already had some data. When I tried to migrate after changing my model, I kept running into an error that the password attribute needs a default value. even after I set a default value(default=‘password’), I kept getting the error message.

I Googled—like any other developers would do—and found out that if I delete the migration diretory, I can easily migrate it again. It looked like I had succeeded, but I couldn’t create a new object using either method that I described in part 3.

And that was because it was still checked as migrated even though I deleted the file. I had to clear the migration story using zsh in order to migrate again by using the command below

$ python manage.py migrate --fake api zero

This cleared the history, so I was able to migrate the new model and then create objects.


This is what I can see in my screen right now. Will work on it more tomorrow and see how it works.