How this dataset was built

The method behind all 267 entries is written down here: what the collection includes, how each entry is divided, how sources are handled, modern English and transliteration.

01 · What is included

This release follows all 267 numbered entries across the 132 chapters of Hisn al-Muslim. Stable OpenDua IDs preserve that order while allowing titles, categories and delivery formats to evolve.

02 · How entries are divided

Words to recite are kept separate from instructions, narrations, counts and asides. When a numbered entry contains several recitations, each part carries its own text, count and audio reference.

03 · Sources and gradings

Each entry keeps its source references. Gradings are reported, never issued by OpenDua, and always name the authority. Conflicting reported gradings remain separate rather than being averaged into one verdict.

04 · Translation and transliteration

OpenDua English aims for accuracy over fluency and uses “Allah” throughout. Transliteration is generated afresh from recited Arabic under one documented scheme; instructions and narrations are deliberately not transliterated.

What can be corrected The texts themselves are fixed. What we can get wrong is the edition around them: a transcription, a diacritic, a reference number, a translation choice, a grading attribution or an audio file. Every reported error of that kind is triaged and, if upheld, listed here with the version that fixed it. Nothing is edited silently.
Status Every entry in dataset v0.0.4 names its source and has checked usage rights. OpenDua English and transliteration are versioned editorial layers, and upheld corrections are recorded publicly. Published 2026-09-09
Changelog and corrections Every correction, dated, with the version that fixed it. READ → Credits Sources, reciter, contributors and how to reach us. READ →