GitHub - argahsuknesib/scryer-graph: A very simple graph database implementation running on Scryer Prolog.

2 min read Original article ↗

Scryer Graph Database

Scryer Graph is a lightweight RDF graph database implemented in Scryer Prolog, built primarily as a learning project to explore logic programming, reasoning, and database concepts. It offers a simple REPL interface for storing and querying RDF triples with graph contexts, including basic operations like adding data, pattern matching, and persistence.

TL;DR: A hands-on Prolog-based RDF database for adding, querying, and persisting the graph to improve the understanding of Scryer Prolog and logical reasoning.

Features

  • RDF Triple Storage: Store triples in the form (Subject, Predicate, Object, Graph) using Prolog's dynamic predicates.
  • Basic Querying: Support for exact matches and placeholders (use "who" to match any value) in queries.
  • Persistence: Save and load database state to/from files.
  • REPL Interface: Interactive command-line interface for operations.
  • Extensible: Built with Prolog for easy addition of reasoning or advanced queries.

Prerequisites

  • Scryer Prolog installed on your system.
  • Make (for using the provided Makefile).

Installation

  1. Clone the repository:

    git clone <repository-url>
cd scryer-graph

  2. Build the project:

Usage

Running the Database

Start the REPL:

In the REPL, enter commands as Prolog terms:

  • Add a triple: add("Alice", "knows", "Bob", "social").
  • Query triples: query("Alice", "knows", "who", "social"). (finds objects where Alice knows someone in the social graph).
  • Save database: save("backup.txt").
  • Load database: load("backup.txt").
  • Quit: quit.

Testing

Run basic tests:

This executes src/test.pl, which adds sample data and verifies queries.

Cleaning

Remove generated files:

Project Structure

  • src/main.pl: Main REPL and command handling.
  • src/database.pl: Core database operations (add, query, save/load).
  • src/test.pl: Basic test suite.
  • Makefile: Build, run, test, and clean targets.

Examples

  1. Adding and Querying:

    scryer-graph> add("Alice", "knows", "Bob", "social").
Added the RDF Graph.
scryer-graph> query("Alice", "knows", "who", "social").
Results: [["Alice", "knows", "Bob", "social"]]

  2. Saving and Loading:

    scryer-graph> save("mydata.txt").
Database saved to mydata.txt.
scryer-graph> load("mydata.txt").
Database loaded from mydata.txt.

Future Enhancements

  • Reasoning: Add inference rules for deriving new triples (e.g., transitivity).
  • Advanced Query Language: Implement a custom DSL or SPARQL subset for complex queries.
  • Performance: Indexing and optimizations for larger datasets.
  • WASM Compilation: Compile to WebAssembly for browser-based usage.

License

This project is open-source under the MIT License. See LICENSE for details.