Articles Collect

https://articles.tomellm.eu

My small test project to collect articles I find interesting.

Scope

Its supposed to kill to birds with one stone:

1. By allowing me to test developing and deploying a Leptos application but also using keycloak for the auth.
2. By giving me a space to collect all of the articles I find on bluesky, reddit usw.

Architecture

Since this is a project to test a few different things out the architecture is unnessearily complicated. I have tried to create different packages for different parts of the application to reduce the dependence on specific dependencies, although there is still some cleanup needed.

./
 ┬─ projects
 │   ├─ domain       -> obiously contains the domain, has do deps
 │   └─ server       -> subdir for all server code
 │       └─ databse  -> code to work with the db
 ├─ migrations       -> contains database migrations 
 └─ web_app          -> the actual application

Deployment

The deployment very simply works by publishing a docker image (tomellm/articles_collect) to the docker hub container registry. That image is then deployed on my coolify instance simply by fetching the latest tag of the Deployed through Github actions and my own Coolify instance.

Run it Yourself

Unfortunatly I have not provided docker compose or similar to start up all of the infastructure. If you still want to give it a try you will have to install cargo-leptos toto build and run the application. The .env-template file defines the environment variables expected by the application meaning:

• keycloak
• a postgres db (migration scripts are automatically executed)

Once you have these up and running, just execute: cargo leptos watch

Development

Some notes on how to develop this app

Migrations

There are two steps to adding new Tables and creating general migrations. First of all, migrations are managed by the migrations project found in the migrations directory. The directory contains information on how to add new migrations, or run migrations to bring the database to a certain state. Exa,ple commands to create an "add_tags" migration and bring the database up to date are:

# creates migration
cargo run --manifest-path ./migration/Cargo.toml -- generate add_tags

# runs all missing migrations
cargo run --manifest-path ./migration/Cargo.toml

Once the migration is created and, for example, a new table is added, the relevant entity needs to be represented in the rust code, specifically in the ./projects/server/database/ project. Instead of creating the entity completly manually I reccomend using the sea orm cli which ist very easy to use and avoids easy to avoid mistakes. The cli looks at the databse state to create entities, meaning that the migrations need to be created and run first. A command to create a new Tag entity might look like this:

# generates entity for "tags" table
sea-orm-cli generate entity -t tags -o ./projects/server/database/src/entities