Meredith Bloom

(see voltorb markdown for part 2)

Deployment: Heroku

Database: Postgresql (the only directly Heroku compatible SQL db)

Framework: Django (I've also worked with Laravel/PHP and Express/Node.js for back ends, but found that Django/Python was my favorite of the three)

requirements.txt file has all package dependencies.

Language: Python

Before jumping into any serious coding, I took a little while to look over the data in the csv file itself. I wrote a SQL command that created a table with corresponding columns and data types that I determined based on my review of the raw data. The five final columns (genres, created_by, published_by, franchises, and regions) were the most varied, with some items having nothing listed at all, to some having multiple things listed under one category. As such, I made the data type for each of these columns an array of strings; I also knew that matching strings is a fairly easy way to execute a search request to an api, and I wanted clients to be able to filter/search for items based on genre, publisher, etc..

    CREATE TABLE media (
    id int,
    media_type varchar(10),
    name varchar(255),
    short_name varchar(255),
    long_description text,
    short_description text,
    created_at timestamp,
    updated_at timestamp,
    review_url varchar(255),
    review_score float,
    slug varchar(255),
    genres text[],
    created_by text[],
    published_by text[],
    franchises text[],
    regions text[]
);

Before deploying my api to Heroku, I ensured that the data was correctly reflected in a database I created on my local drive. Django automatically created a model based on the existing table/db that I had connected to my django project - kind of like a backwards migration. Django's literal translations and educated guesses about the data types for the columns were accurate, so I left the model mostly as-is, except for adjusting some additional parameters (i.e. Blank, null).

    class Media(models.Model):
        id = models.IntegerField(blank=True, primary_key=True)
        media_type = models.CharField(max_length=10, null=True)
        name = models.CharField(max_length=255, blank=True, null=True)
        short_name = models.CharField(max_length=255, blank=True, null=True)
        long_description = models.TextField(blank=True, null=True)
        short_description = models.TextField(blank=True, null=True)
        created_at = models.DateTimeField(blank=True, null=True)
        updated_at = models.DateTimeField(blank=True, null=True)
        review_url = models.CharField(max_length=255, blank=True, null=True)
        review_score = models.DecimalField(max_digits=100, decimal_places=2, blank=True, null=True)
        slug = models.CharField(max_length=255, blank=True, null=True)
        genres = models.TextField(blank=True, null=True)  
        created_by = models.TextField(blank=True, null=True)  
        published_by = models.TextField(blank=True, null=True) 
        franchises = models.TextField(blank=True, null=True)  
        regions = models.TextField(blank=True, null=True) 

    class Meta:
        managed = True
        db_table = 'media'
        
    def __str__(self):
        return self.name

TLDR: csv file -> sql create table -> populated table by copying csv file -> later, django auto-builds model based on existing table.

I wanted to take advantage of Django's views feature, as I've always used React as a separate front end for my Django-backed projects. As such, I created an API with routes that return pure JSON data, as well as some very simple views (virtually zero styling) for in-browser user navigation. Users can easily navigate between an index view and individual entries, with additional filters by media type (show, game, movie, comic). I also included a search bar.

Installing dependencies (for running on local drive)

1. CD to directory with requirements.txt file
2. activate virtualenv (local python environment)
3. and run: pip install -r requirements.txt

Live Browsable Link: Code Foo IGN Application

Base API URL: https://code-foo-ign.herokuapp.com/api

GET All Items (default ordered by id/A-Z)

Request path:

https://code-foo-ign.herokuapp.com/api

Additional filters

You can include multiple filters simultaneously, though only one value for genre/created_by/publisher. Strings are case insensitive (both Drama and drama will work).

GET specific item (by id)

Request path:

https://code-foo-ign.herokuapp.com/api/<id>/

GET - search for item by name

Request path:

<https://code-foo-ign.herokuapp.com/api/search>

example:

<https://code-foo-ign.herokuapp.com/api/search?search=assassin>

GET all items, custom order

Request path:

<https://code-foo-ign.herokuapp.com/api/ordered>

example - will return all items, ordered by descending review score:

<https://code-foo-ign.herokuapp.com/api/ordered/?ordering=-review_score>

GET by media type, custom order

Request path(s) by media type:

<https://code-foo-ign.herokuapp.com/api/games> OR

<https://code-foo-ign.herokuapp.com/api/movies> OR

<https://code-foo-ign.herokuapp.com/api/comics> OR

<https://code-foo-ign.herokuapp.com/api/shows> OR

Though I didn't do any kind of manual data clean-up, Django has an admin feature that has full CRUD capabilities built in. Logging into the admin portal using the below credentials would allow you to make changes to the data directly, without having to write separate POST/PATCH/DELETE routes. Of course this log in information would normally be private, but for the sake of this application I thought I'd include it, because this feature is one of the reasons I like Django so much.

<https://code-foo-ign.herokuapp.com/admin>

username: superuser

password: Pass4321