This page brings together the lightweight general data files that can be used to inspect the structure of the atlas or build reading tools on top of it. The data here are derived from the published pages and semantic search layers, but they do not publish the full text of RAG passages, and they are not an alternative source to Muhammad Shahrur’s books or to an automatic interpretation of the Qur’an.
Index
- Manifest: file index, record counts, and usage limits.
- Atlas pages: pages, links, editorial types, and content length.
- Atlas RAG index: titles, links, and passage types with a short excerpt only, without the full text.
- Qur’anic RAG index: indexed verses and Qur’anic passages with a short excerpt only, without the full text.
- Shahrur-Qur’an bridge: preliminary relations between Shahrur’s concepts and verse locations.
- Atlas graph: nodes and links for pages and conceptual relations.
- Graph readings: centers of gravity, bridges between sections, and weakly connected pages.
- README: brief description of the files.
How is the data read?
The best starting point is the atlas-manifest.json file, because it lists the available files and the number of records in each file. jsonl files are read as one independent JSON record per line, which makes them suitable for programmatic processing or for ingestion into research databases.
In the public files, RAG records use fields such as textPreview, textLength, and textSha256 instead of the full text field. The full version remains internal and is built locally outside the publication folder.
Limits of use
These data help build reading and inspection tools, but they do not determine the validity of Shahrur’s reading, do not interpret the verses, and do not replace returning to the original texts. Any final research quotation must refer back to the original source or the published page that indicates its location.