What's Toogether app?

An app where users can create a group with their friends, match with other groups and hang out

This is just the backend of the app, if you're looking for the frontend: https://github.com/damianstone/toogether-mobile

Technologies

• 🐍 Django REST framework
• 🔥 Websockets (channels)
• 📍 Geolocation with Gdal, Geos and Postgis
• 🐘 PostgreSQL
• 🏓 Redis
• 🐬 Daphne
• 🖼️ Amazon S3
• 🧹 Frake8 and black
• 🔒 Token authentication (JWT)
• 👥 Faker
• 📸 Pillow
• ☁️ Heroku

Featues

👀 Basic

• Login and registration with auth token
• Create profile
• Report profiles
• Block profiles
• Password recovery with email verification

👤 Your profile

• Update your personal information
• Add photos

💃 Swipe single and group profiles

• Support single and group profiles
• Like
• Unlike
• Undo

🕺🏼 Create a group profile with your friends

• Create a groups
• Invite friends using unique link
• If admin (creator of the group): remove and add members
• If not admin, join to group using the link and leave

💬 Group chat

• Group chat generated autmatically when joining our creating a group
• For now the chat just support text messages

💬 Matches and chats

• Chat with your matches
• Delete matches
• Report and block profiles

🔗 Matchmaking algorithm

List based on

• Your location
• Age
• Gender
• Preferences
• Group sizes
• More below in this documentation!

🕵️‍♂️ Internal endpoints

• Create fake data for the whole app using Faker
• Admin restricted endpoints for the models

mailto: damianstonedev@gmail.com

Initialization

Create a virtual environment

For IOS / Linux

python3 -m venv venv-people
source venv-people/bin/activate 

For Windows

python -m venv venv-people
venv-people\Scripts\activate.bat

Install requirements.txt

pip install -r requirements.txt

Create .env for local PostgreSQL database

LOCAL_DB_NAME=name-of-the-database
LOCAL_DB_USER=db-user-name
LOCAL_DB_PASSWORD=db-password
LOCAL_DB_HOST=host-you-want-to-use
LOCAL_DB_PORT=post-you-want-to-use

Migrations folder

Inside api create a new folder called migrations and inside add the following file:

__init__.py

Installing Geospatial libraries

Depending on your operating systems the installation can be quite different therefore we recommend you to follow the official documentation for this:

https://docs.djangoproject.com/en/4.1/ref/contrib/gis/install/geolibs/

Migrate models

python manage.py makemigrations
python manage.py migrate

Create a super user account

Creating a superuser will give you administrative privileges, and most important, access to our local internal endpoints for development purposes

When running the command, choose a memorable email and password

To create a superuser account, use the following command:

python manage.py createsuperuser

Run

python manage.py runserver

Install Redis for WebSockets connectios

In the following link you can see the different installations for different operating systems https://redis.io/docs/getting-started/installation/

Run Redis

redis-server

Stop Redis

killall redis-server

Deployment with Heroku

Buildpakcs information

Geolocation: GDAL, Geos and PROJ: https://github.com/heroku/heroku-geo-buildpack

Install Heroku CLI

Using the following link: https://devcenter.heroku.com/articles/heroku-cli

Login and check the apps

heroku login
heroku apps

Before deploy

From the heroku branch run the following

python mangage.py makemigrations
python manage.py migrate
python manage.py collectstatic

In order to use the last builpacks of GDAL and Geos, make sure you do not have set BUILD_WITH_GEO_LIBRARIES

If you do, run the following command

heroku config:unset BUILD_WITH_GEO_LIBRARIES --app toogether-api

Besides, as we are collecting then static files manually, we need to disable the auto coollect static

heroku config:set DISABLE_COLLECTSTATIC=1 --app toogether-api

Push the latest changes

git push heroku

Manual deploy in Heroku

Heroku website panel -> Deploy -> Manual branch deploy

After deployment

Delete all the folders and files inside the static, this is because when the code is deployed, the static files are automatically collected, but, in order to keep the repository clean, those files must be deleted once the deployment is successful.

Troubleshooting

To check the logs

heroku logs --tail --app toogether-api

Style Standards

To format the code in the project, simply run the following command in the root directory of the project:

black .

This command will automatically format all .py files in the project according to the black style guide, which adheres to the PEP 8 style guide

Make sure you run this command before any pull request

Pull Requests

Before any merge to develop or rocket, it will be necessary to make a Pull Request and a code review.

Basic PR structure:

your-branch -> feature-branch -> develop -> rocket

Steps for a Pull Request

1. Push your branch to the remote repository: git push
2. Navigate to the GitHub website
3. Create the pull request (PR) manually by selecting the correct feature-branch you are working on and clicking on the "New pull request" button
4. Notify the team about your PR through our communication channel: Discord

Project structure

Profile views

In Together, there is a single data model for users known as the profile model. The distinction between a user and a profile is that a user may log into the app but has not yet created a profile.

User views These views manage the basic functionality of the user model, such as creating and deleting a user.

Profile Modelview This set of views manages all the actions that a user can perform on their profile, such as adding photos, updating information, and blocking other users.

Grup Views

These endpoints manage all the actions related to groups, such as creating a group with friends and removing members.

Swipe Views

The swipe views manage the "liking" functionality between users, as well as the matchmaking algorithm. This includes all the processes and calculations involved in determining the profiles that a user is matched with. These views are an integral part of the application as they enable the core user interaction and facilitate the potential formation of relationships.

Matchmaking Algorithm

The matchmaking algorithm in Together is responsible for determining the matches between profiles and groups. This process is initiated when a user "likes" another profile or group.

The algorithm is implemented in the like function and uses several checks to determine the nature of the like being given. There are four possible scenarios for a like:

1. One profile to one profile
2. One profile to a group
3. A group to one profile
4. A group to another group

Each scenario is handled by a different function that performs the necessary operations to determine the match.

Inputs

• request: a Django request object
• current_profile or current_group: the profile or group making the like
• liked_profile or liked_group : the profile or group being liked

Outputs

A Django response object containing:

• details: a string indicating the status of the like, either LIKE, ALREADY_MATCHED, NEW_MATCH or SAME_MATCH
• group_match: a string indicating whether the match was between groups: NEITHER, BOTH, LIKED or CURRENT
• match_data: the match serialization in case a new match was created, otherwise this field will not be present.

Summary of limitations

• The Toogether name and brand are registered; you cannot use the name, designs, or colors.
• You can use the app for any learning purpose as long as it runs in your local environment. In other words, you cannot deploy our code.
• You can use the code for learning purposes and references. For example, if you are creating a similar app, you can use our code to learn how to build features like a swipe function and get inspiration from our implementation.
• You cannot sell this code in any form.
• You cannot conduct business with this code in any form

More information can be found in the LICENSE file in the root of the project. If you know of someone who is not respecting these limitations, please contact the owner at damianstonedev@gmail.com. The open-source world thrives on our respectful behavior ❤️