Leed

Leed is a lightweight, self‑hosted RSS/ATOM aggregator written in PHP that prioritizes speed and minimalism. It decouples feed fetching from the user experience by running a cron‑driven background task (action.php) that pulls, parses, and stores feed entries in a MySQL database. When a user logs in through the web interface, Leed simply reads from this pre‑populated cache, ensuring instant page loads even on low‑powered hardware.

Key Features

• Cron‑based ingestion: All network I/O and XML parsing occur in a scheduled PHP script, keeping the web server stateless during user sessions.
• OPML import/export: Enables migration from other aggregators and bulk subscription management via a standard XML format.
• Responsive UI: Built with vanilla PHP/HTML/CSS; no heavy JavaScript frameworks, making it trivial to host on shared servers or containers.
• Synchronization token: When invoking the ingestion script via HTTP (e.g., wget), a secret code protects against unauthorized triggers.

Technical Stack

Leed’s codebase follows a classic Model–View–Controller pattern with procedural PHP. The models/ directory contains data access objects that wrap PDO calls, while views/ holds minimal PHP templates. There is no ORM; direct SQL queries are used for performance and simplicity.

Core Capabilities & APIs

• REST‑like endpoints: action.php?action=… exposes a few read/write operations (synchronize, delete entry) that can be consumed by external scripts or custom front‑ends.
• Webhooks: By configuring a cron job to hit an external URL after each sync, developers can integrate Leed with notification services (Slack, Discord).
• Plugin hooks: The hooks/ directory allows third‑party PHP files to be loaded during feed parsing, enabling custom filtering or enrichment (e.g., Markdown rendering of descriptions).

Deployment & Infrastructure

Leed is designed for self‑hosting on LAMP stacks. Its minimal dependencies mean it can run in a small Docker container:

docker run -p 8080:80 ghcr.io/leedrss/leed:latest

For scalability, the cron job can be distributed across multiple nodes by pointing each to a shared MySQL instance and using a unique sync code. Horizontal scaling is limited only by the database’s read/write throughput, as each node processes all feeds independently.

Integration & Extensibility

• OPML: Import/export facilitates integration with other aggregators or data pipelines.
• Custom feed processors: Developers can drop PHP scripts into hooks/ to transform RSS items (e.g., strip HTML, add tags).
• API consumption: The lightweight REST endpoints can be wrapped by a GraphQL layer or consumed directly from a mobile app.

Developer Experience

• Configuration: A single config.php holds database credentials, sync code, and cron schedule. No complex YAML or JSON files.
• Documentation: The README is comprehensive, with clear sections for prerequisites, installation, and cron setup. Inline comments in the code provide additional context.
• Community: The project is hosted on GitHub with an active issue tracker. Contributors include multiple maintainers, and the AGPL‑3.0 license encourages open collaboration.

Use Cases

Advantages Over Alternatives

• Performance: By separating ingestion from rendering, page load times stay sub‑200 ms even with dozens of feeds.
• Simplicity: No JavaScript frameworks or heavy dependencies; a single PHP binary and MySQL suffice.
• Open‑source freedom: AGPL licensing ensures that any derivative must remain open, fostering community contributions.
• Ease of deployment: Works out‑of‑the‑box on shared hosting with Apache; Docker images reduce friction for cloud deployments.

Leed offers developers a lean, well‑documented foundation to build or integrate RSS aggregation into larger systems without the overhead of full‑featured readers.