Projektowanie schematów baz danych

Imagine building a house. Before you lay a single brick, you need a blueprint! In databases, that blueprint is called a schema.

A database schema defines the structure of your data. It tells you:

• What tables exist
• What columns each table has
• How different tables are connected

A good schema is key for a well-organized, efficient, and scalable Supabase project.

First, think about the 'things' you want to store. These are your entities. For a social app, entities might be Users, Posts, or Comments.

Next, what details does each entity have? These are its attributes.

• User: name, email, profile picture URL
• Post: title, content, creation date

Listing these helps you imagine your tables.

Every entity needs a way to be uniquely identified. That's where a Primary Key comes in!

A primary key is a column (or set of columns) that contains a unique value for each row in a table. Think of it like a student ID number – no two students have the same one.

The most common primary key is an id column, often an auto-incrementing integer or a UUID (Universally Unique Identifier).

Your entities don't live in isolation; they interact! These interactions are called relationships.

The most common type is One-to-Many. For example:

• One User can write Many Posts.
• One Album can have Many Songs.

Understanding these connections is crucial for linking your tables.

To establish a One-to-Many relationship, we use a Foreign Key.

A foreign key is a column in one table that refers to the primary key in another table. It's how we say, 'This post belongs to this specific user.'

If a posts table needs to know which user created it, the posts table would have a user_id column that points to the id of a row in the users table.

When defining your columns, you need to pick a data type. This tells the database what kind of information that column will hold (text, numbers, dates, etc.).

Common PostgreSQL (Supabase's database) data types include:

• text: For names, descriptions
• integer: For whole numbers (e.g., age, count)
• boolean: For true/false values (e.g., is_active)
• timestamp: For dates and times (e.g., created_at)

Choosing the right type saves space and improves performance.

Normalization is a process to organize your database to reduce data redundancy and improve data integrity.

Imagine if every post stored the user's full name and email. If the user changes their email, you'd have to update it in many places! Normalization helps avoid this.

By splitting data into separate, related tables (e.g., users and posts), you store common information only once.

While normalization is great for data integrity, sometimes it can slow down data retrieval because you need to 'join' many tables.

Denormalization is the intentional introduction of redundancy to improve read performance. For example, storing a user's name directly in a posts table, even though it's also in the users table.

This is a trade-off: faster reads, but more complex updates and potential for inconsistencies. Use it wisely!

Let's design a simple schema for users and their to-do tasks. Each user can have many tasks.

Here's how we might define the tables based on our design principles:

You're planning a database for a blog. You need to store Authors and their Articles. Each author can write many articles.

Which of these statements correctly apply good schema design principles for this scenario?

You've just taken your first steps into database schema design!

We covered:

• Identifying entities and attributes
• The role of primary and foreign keys
• Understanding relationships (especially One-to-Many)
• The importance of selecting proper data types
• Briefly touching on normalization and denormalization

A well-designed schema is the foundation of any robust application. In the next lesson, we'll put this knowledge into practice by creating tables!