Datasets:

quranlab
/

islamic-corpus-graph

QuranLab is a volunteer effort. Our aim is to present these sources carefully and at high quality, and to help them travel faithfully — in the spirit in which they were written — not to claim them as ours.

This dataset was not gathered from anywhere directly. It is derived from our own corpora — quranlab/quran, quranlab/hadith, quranlab/quran-audio — and everything in it can be traced back to them. Whatever thanks are owed are owed one step further up, so they are repeated here rather than left to a link.

The projects the underlying corpora rest on

Tanzil Project — The canonical Uthmani and Simple-Clean Arabic text, kept verbatim (CC BY 3.0).
fawazahmed0/quran-api — An open aggregation through which most translations were collected.
QuranEnc.com — Official translation and tafsīr editions, republished under their stated terms.
Tafsir Center for Quranic Studies — Al-Sa'dī and al-Muyassar corpora (CC BY 4.0).
Quranic Universal Library (QUL) — Current public API for several classical tafsīr resources.
spa5k/tafsir_api — MIT-packaged snapshots of classical tafsīr texts (the works themselves are public domain).
OpenITI RELEASE — Machine-readable classical Islamicate texts, pinned to an exact release (CC BY-NC-SA 4.0).
Diyanet İşleri Başkanlığı — The official Turkish meal and the Kur'an Yolu commentary.
Quranic Arabic Corpus — Word-by-word morphological annotation (GPL).
fawazahmed0/hadith-api — Ḥadīth collections aligned across languages, with multi-grader gradings.
mhashim6/Open-Hadith-Data — Arabic matn for Musnad Aḥmad and al-Dārimī (ODbL-1.0 / DbCL-1.0).
HadeethEnc.com — A graded, many-language selection of prophetic ḥadīth (IslamHouse / Saudi Ministry of Islamic Affairs).
everyayah.com — Per-ayah recitation references (VerseByVerseQuran).
mp3quran.net — Per-surah recitation references across many readings.
cpfair/quran-align — The open word-level timing that existed before ours, for 11 recordings (CC BY 4.0).
jonatasgrosman/wav2vec2-large-xlsr-53-arabic — The open acoustic model we used to align the recordings that had no timing.

Per-work terms and credit stay with the corpora, where each config carries its own entry.

Consulted, but never quoted

The same care applies here. These helped us decide what a work is and where a passage belongs; none of their text is present.

Quranpedia — It told us where one commentary's passages belong, and confirmed the school of two authors. Not a word of it is in the data.
TDV İslâm Ansiklopedisi — Our first stop when we needed to be sure who an author was and where he stood. Read, never reproduced.
sunnah.com — Each ḥadīth carries a link to its page there, and nothing more — their text and gradings stay theirs.
Scholarly studies and critical editions — Consulted whenever a work's standing needed better evidence than our own reading. Cited in metadata/.

If we got something wrong

If you hold rights to a work that reached this dataset, or we have credited someone poorly, write to us on the Community tab of quranlab/islamic-corpus-graph. A removal request is honoured promptly and without argument.