quran-verify: per-ayah SHA-256 checksums to verify Quran text integrity

Maajid

Assalamu alaikum,

I built a small open-source tool: a JSON manifest with a SHA-256 hash for every ayah of the Tanzil Uthmani text (Hafs), plus per-surah hashes and one root hash for the whole Quran.

The point: any app or script can check that its Quran text is intact — after OCR, a font pipeline, or copying between systems — and see exactly which ayah differs if something is wrong.

This is not a new text or a claim about which edition is correct. It's just a verification layer over the existing Tanzil text (based on the Madinah Mushaf). The full provenance chain and normalization rules are documented in the repo:

https://github.com/spqrxi/quranchecksum

Feedback and review are very welcome — especially on the normalization rules and on supporting other riwayat in the future.

علا صالح

Maajid

Assalamu alaikum

Thank you for building and sharing this tool. I really like the idea

One small suggestion: consider adding a separate manifest for the plain (Imla'i) text as well
Many Quran apps rely on the search-friendly text so having checksums for both the Uthmani and Imla'i versions would make this even more useful for developers

علي ملص

Maajid

Assalamu alaikum Maajid.

Great tool and I completely agree with Ola's suggestion. As a next step, it would be highly beneficial to support other Mushafs and Riwayat (like Warsh and Qaloon). This would make things much easier for developers, especially in text processing and audio synchronization projects that require high accuracy in matching texts across different Riwayat, such as the Miqat project I worked on. Great effort!

هادي الأحمد

Alaykum Assalam
interesting idea, I'd like to know more about what edge does this method have over strict text checking (word by word programmatic search and match) of the text
what does the hashing step add in this case? speed of checking? or something else

Maajid

Alaykum Assalam — at Quran scale speed isn't really the point; a word-by-word compare is already fast.
The edge is that you don't need the reference text present to check against it. Instead of shipping a full known-good copy, you ship a tiny manifest of hashes and verify against that. It also gives you a stable, citable fingerprint (one root hash for the whole Quran, plus per-surah and per-verse), and the rollup means a single comparison tells you "did anything change," with the verse hashes pinpointing where. Normalization (NFC + whitespace strip) is specified, so a "match" means the same thing across every implementation.
Caveat: if both sides already hold the identical reference, hashing and direct compare detect the same thing — a match only proves the text is byte-identical to the source, not that it's correct.

Maajid

That's the whole argument in one gesture: touch a single ayah and the corruption climbs the tree until the root itself goes red. Since an app only stores that one root value, it catches any change without holding a full copy of the text — and the red trail is exactly the "drill down" that pinpoints which surah and ayah broke.

Maajid

DEMO https://waqf-m.vercel.app/checksum.html

علا صالح

Maajid

Great work

JazakAllah khair for building this drmo

I really like how the interactive visualization makes the hashing concept easy to understand

Maajid

https://github.com/dfordev1/usxv2 universal language for quran check this also we have integrated it together

ثبت مجتمعنا كتطبيق لتجربة افضل

انضم إلى نشرة إتقان 📩

انضم إلى نشرة إتقان 📩

شكراً لك!