Models
Models is an implementation for a SQL migration management tool. It supports PostgreSQL, MySQL, and SQLite.
Quick Start
install the CLI by running the following command:
$ cargo install models-cli
Now run the following command to create an environment file with the DATABASE_URL
variable set:
$ echo "DATABASE_URL=sqlite://database.db" > .env
Alternatively it can be set as a environment variable with the following command:
$ export DATABASE_URL=sqlite://database.db
We now can create the database running the following command:
$ models database create
This command will have created an SQLite file called database.db
. You can now derive the Model
trait on your structures, and models
will manage the migrations for you. For example, write at src/main.rs
:
#![allow(dead_code)]
use models::Model;
#[derive(Model)]
struct User {
#[primary_key]
id: i32,
#[unique]
email: String,
password: String,
#[default(0)]
is_admin: bool,
}
#[derive(Model)]
struct Post {
#[primary_key]
id: i32,
#[foreign_key(User.id)]
author: String,
#[default("<Untitled Post>")]
title: String,
content: String,
}
#[derive(Model)]
struct PostLike {
#[foreign_key(User.id, on_delete="cascade")]
#[primary_key(post_id)]
user_id: i32,
#[foreign_key(Post.id, on_delete="cascade")]
post_id: i32,
}
#[derive(Model)]
struct CommentLike {
#[foreign_key(User.id)]
#[primary_key(comment_id)]
user_id: i32,
#[foreign_key(Comment.id)]
comment_id: i32,
#[default(0)]
is_dislike: bool,
}
#[derive(Model)]
struct Comment {
#[primary_key]
id: i32,
#[foreign_key(User.id)]
author: i32,
#[foreign_key(Post.id)]
post: i32,
}
fn main() {}
If you now run the following command, your migrations should be automatically created.
$ models generate
The output should look like this:
Generated: migrations/1632280793452 user
Generated: migrations/1632280793459 post
Generated: migrations/1632280793465 postlike
Generated: migrations/1632280793471 comment
Generated: migrations/1632280793476 commentlike
You can check out the generated migrations at the migrations/
folder. To execute these migrations you can execute the following command:
models migrate run
The output should look like this:
Applied 1631716729974/migrate user (342.208µs)
Applied 1631716729980/migrate post (255.958µs)
Applied 1631716729986/migrate comment (287.792µs)
Applied 1631716729993/migrate postlike (349.834µs)
Applied 1631716729998/migrate commentlike (374.625µs)
If we later modify those structures in our application, we can generate new migrations to update the tables.
Reverting migration
Models can generate down migrations with the -r
flag. Note that simple and reversible migrations cannot be mixed:
$ models generate -r
In order to revert the last migration executed you can run:
$ models migrate revert
If you later want to see which migrations are yet to be applied you can also excecute:
$ models migrate info
Avaibale Attributes
primary_key
It's used to mark the primary key fo the table.
#[primary_key]
id: i32,
for tables with multicolumn primary keys, the following syntax is used:
#[primary_key(second_id)]
first_id: i32,
second_id: i32,
This is equivalent to
PRIMARY KEY (first_id, second_id),
foreign_key
It is used to mark a foreign key constraint.
#[foreign_key(User.id)]
user: i32,
It can also specify on_delete
and on_update
constraints:
#[foreign_key(User.id, on_delete="cascade")]
user_id: i32,
This is equivalent to
FOREIGN KEY (user_id) REFERENCES user (id) ON DELETE CASCADE,
default
It can be used to set a default value for a column.
#[default(false)] // when using SQLite use 0 or 1
is_admin: bool,
#[default("")]
text: String,
#[default(0)]
number: i32,
unique
It is used to mark a unique constraint.
#[unique]
email: String,
For multicolumn unique constraints the following syntax is used:
#[unique(post_id)]
user_id: String,
post_id: i32,
This is equivalent to
UNIQUE (user_id, post_id),