My personalized news and blog aggregator
Go to file
Tyler Hallada a72bfa15bd Use HX-Target, not HX-Boost for Layout render
HX-Boost is not always sent in all AJAX requests that htmx sends, only
those initiated by an element with hx-boost enabled. It was not showing
up on requests following an HX-Redirect response.

After reading the docs more, I realized HX-Target was what I wanted. If
I can see that the request is targeting `#main-content` then I know to
only return HTML inside that element. Simple.
2023-09-27 01:28:53 -04:00
frontend Rename signup to register and center forms 2023-09-27 00:23:28 -04:00
migrations Add basic user auth 2023-09-25 01:38:18 -04:00
src Use HX-Target, not HX-Boost for Layout render 2023-09-27 01:28:53 -04:00
.eslintignore Add eslint and prettier to frontend 2023-09-01 01:05:06 -04:00
.gitignore Implement entry and feed pagination 2023-09-02 14:01:18 -04:00
.prettierrc.json Add eslint and prettier to frontend 2023-09-01 01:05:06 -04:00
build.rs Also bundle css with bun so it busts caches 2023-06-29 00:38:32 -04:00
Cargo.lock Add hx-boosted support to Layout 2023-09-26 23:36:31 -04:00
Cargo.toml Add hx-boosted support to Layout 2023-09-26 23:36:31 -04:00
drop_all.sql Add basic user auth 2023-09-25 01:38:18 -04:00
justfile Implement entry and feed pagination 2023-09-02 14:01:18 -04:00
README.md Store sessions in redis 2023-09-26 01:47:50 -04:00
watch.sh Add auto-reload for frontend and backend 2023-06-28 01:23:11 -04:00

crawlnicle

My personalized news and blog aggregator. Taking back control over the algorithm. Pining for the days of Google Reader. An excuse to write more Rust.

Development Instructions

Prerequisites

Install these requirements to get started developing crawlnicle.

  • rust

  • postgres

  • redis

  • sqlx-cli

    • Only postgres needed. Install with:
    cargo install sqlx-cli --no-default-features --features native-tls,postgres
    
  • just

  • bun

  • (optional) cargo-watch for auto-recompiling the server in development

  • (optional) mold for faster builds

First-time setup

  1. Create postgres user and database:

    createuser crawlnicle
    createdb crawlnicle
    sudo -u postgres -i psql
    postgres=# ALTER DATABASE crawlnicle OWNER TO crawlnicle;
    \password crawlnicle
    
    # Or, on Windows in PowerShell:
    
    & 'C:\Program Files\PostgreSQL\13\bin\createuser.exe' -U postgres crawlnicle
    & 'C:\Program Files\PostgreSQL\13\bin\createdb.exe' -U postgres crawlnicle
    & 'C:\Program Files\PostgreSQL\13\bin\psql.exe' -U postgres
    postgres=# ALTER DATABASE crawlnicle OWNER TO crawlnicle;
    \password crawlnicle
    
  2. Save password somewhere safe and then and add a .env file to the project directory with the contents:

    RUST_LOG=crawlnicle=debug,cli=debug,lib=debug,tower_http=debug,sqlx=debug
    HOST=127.0.0.1
    PORT=3000
    DATABASE_URL=postgresql://crawlnicle:<password>@localhost/crawlnicle
    DATABASE_MAX_CONNECTIONS=5
    REDIS_URL=redis://localhost
    TITLE=crawlnicle
    MAX_MEM_LOG_SIZE=1000000
    CONTENT_DIR=./content
    
  3. Run just migrate (or sqlx migrate run) which will run all the database migrations.

Running in Development

Run just watch to build and run the server while watching the source-files for changes and triggering a recompilation when modifications are made.

The server also triggers the browser to reload the page when the server binary is updated and the server is restarted.

It also separately watches the files in frontend/ which will trigger a transpilation with bun and then rebuild the server binary so that it includes the new JS bundle names.

Alternatively, you can just run cargo run after building the frontend JavaScript with just build-dev-frontend.

Building for Production

You can also build the binary in release mode for running in production with the just build command. This will first build the minified frontend JavaScript (just build-frontend) and then build the rust binary with cargo build --release.

Using the CLI

This project also comes with a CLI binary which allows you to manipulate the database directly without needing to go through the REST API server. Run cli --help to see all of the available commands.