SPARQL-Anything/sparql.anything

SPARQL Anything is a system for Semantic Web re-engineering that allows users to ... query anything with SPARQL.

What it solves

Most data lives in messy, heterogeneous formats like CSV, JSON, XML, spreadsheets, HTML, Markdown, and even archives. Normally you'd need to convert each format into a common structure before you can query it. SPARQL is a powerful query language for graph data, but it historically requires you to build an RDF graph first. SPARQL Anything removes that step: you can run SPARQL queries directly on files in their original format, without any preprocessing or conversion.

How it works

SPARQL Anything uses a model called Façade-X, a minimal meta-model that treats any data source as if it were already RDF. When you write a SPARQL query, you refer to a data source using a special service URL (like x-sparql-anything:https://example.com/data.json) inside the query. Behind the scenes, SPARQL Anything reads the source, translates it on the fly into RDF triples, and returns results that match your query. It's bundled as a single Java jar file, so you can run it from the command line. It also works as a server, a Java library, a Python library, or a Docker container, making it flexible for different workflows.

Who it's for

This tool is for developers, data engineers, and semantic web researchers who work with mixed data sources and want to integrate them into RDF-based systems without building custom parsers for each format. It's especially useful for knowledge graph construction and data integration projects where the data comes from many different files and endpoints. The project is also being standardised within the W3C Data Façades Community Group, so it appeals to those who care about interoperable, community-driven approaches.

Highlights

  • Supports a wide range of formats: CSV, JSON, XML, spreadsheets, HTML, Markdown, archives, and more.
  • No upfront conversion – query files directly with SPARQL via a simple service syntax.
  • Runs anywhere: command line, server, Java library, Python library, or Docker.
  • Open source under Apache 2.0 license, with active community discussions and issue tracking.
  • Backed by published research and part of W3C standardisation efforts.

Related

  • Project
    dorylab/dory

    Dory is an AI-native SQL client that provides a persistent, editable workspace for both humans and AI agents to query, visualize, and manage database operations via the Model Context Protocol.

  • Project
    typesafe-ai/typesafe-sdk-python

    The TypeSafe AI Python SDK provides a client interface to interact with the TypeSafe AI platform, enabling developers to perform structured data classification and processing tasks.

  • Project
    Mengqi97/chinese-medical-dataset

    A curated collection of open-source Chinese medical datasets for tasks like medical QA, entity recognition, and knowledge graph construction to support the development of medical AI.

  • Project
    knowsys/nemo

    Nemo is a Rust‑implemented, in‑memory Datalog rule engine for fast, declarative reasoning over RDF‑compatible data. It offers a CLI (`nmo`), language bindings, and a browser UI, supporting many data formats (CSV, Turtle, SPARQL, etc.) and advanced Datalog features like stratified negation and existential rules. The project is open‑source (Apache‑2.0/MIT) and geared toward knowledge‑graph and analytic workloads.