TrackFinder Backend

Backend for managing a music database with MongoDB Atlas, using FastAPI.

Features

1. Search Songs: Flexible filtering based on various fields.
2. List Collections: Retrieve a list of all collections in the database.
3. Add Songs: Add a new song to a specified collection, with duplicate checks.

Structure

TrackFinder/
│
├── main.py               # Entry point of the application, runs the FastAPI app
├── models/
│   ├── __init__.py       # Initializes the models package
│   └── songModels.py     # Defines the SongMetadata model
├── music_folder/         # Contains JSON files with song metadata
│   └── *.json            # Example: rock.json contains metadata for rock songs
├── routers/
│   ├── __init__.py       # Initializes the routers package
│   └── general_use_router.py # Contains route definitions
├── .env                  # Environment variables for MongoDB connection
├── .gitignore            # Git ignore file
└── seeding_db.py         # Script for seeding the database with JSON files

Environment Setup

Ensure you have a .env file with the following content:

MONGO_DB_USER= ...
MONGO_DB_PASSWORD= ...

So the software can get those and use them here:
MONGO_DB_URL=mongodb+srv://<username>:<password>@cluster0.mongodb.net/<dbname>?retryWrites=true&w=majority

Seeding the Database

To seed the database with song metadata from JSON files, run:

python seeding_db.py

This will populate MongoDB Atlas with your song collections.

Running the Application

To run the application:

python main.py

This will host the API on http://127.0.0.1:8000 by default.

Endpoints

Search Songs

• Endpoint: /search_songs
• Method: POST
• Description: Search for songs based on criteria.
• Parameters: title, artist, album, release_year, genre ... (all optional).

List Collections

• Endpoint: /collections
• Method: GET
• Description: Retrieve a list of all collections in the database.

Add a New Song

• Endpoint: /add_song
• Method: POST
• Description: Add a new song to a specified collection. If the collection does not exist, it will be created.
• Parameters:
  • collection_name: The name of the collection (path parameter).
  • SongMetadata: The song data (body parameter).

SongMetadata Model

from pydantic import BaseModel
from typing import Optional

class Song_metadata(BaseModel):
    name: str
    title: str
    sub_title: str
    bitrate: int
    commentaries: str
    main_artist: str
    collaborators: str
    album_artist: str
    album: str
    year: str
    track_number: str
    genre: str
    duration: int

Documentation

Access detailed API documentation generated by FastAPI at http://127.0.0.1:8000/docs after running the application.

---

Context:

How did all this start?

I don't like to use spotify, so I rather manually select and download music.

The thing is that now I have at least 900 songs divided in multiple folders. If I want to make a new playlist I do not know anymore if I already placed a song in another playlist before. So I proceeded to:

1. Manually added the metadata for all the songs I downloaded (not funny at all)

2. I used another project of mine called MusicToMetaData to extract the metadata into json files. Like that evey folder became a single json file (rock, salsa, etc) and inside every json file there is the metadata of the songs in the json format.

3. I uploaded all the files to mondoDB atlas, making for evey file a collection and every song, a document.

4. I developed this backend software so I can check if any new song I like in the future has been already added to a collection.