Coran Codex

DATA-001

Validated dataset

Quran dataset and counting rules

This record defines the source corpus, analytical views and baseline counting rules used by Coran Codex research.

Results calculated under RAW and BODY are not interchangeable. Every study must state which view it uses.

Source identity

Primary analysis corpus

Source
Tanzil Quran Text
Edition
Simple Clean
Version
1.1
Filename
quran-simple-clean-tanzil-v1.1-official.txt
File size
781,176 bytes
Surahs
114
Verses
6,236
SHA-256
228DF2A717671AEB9D2FF573002BD28D6B3F973F4BC7153554E3A81663D67610

The source file is treated as immutable. Coran Codex does not alter the original Quran text in order to produce numerical results.

Validated baseline

RAW corpus metrics

6,236Verses
78,248Words
332,837Arabic letters
36Distinct letters

A word is defined as one whitespace-separated unit. attached arabic prefixes are not split into separate words.

Analytical views

RAW and BODY

RAW

Source view

Primary

RAW analyzes the verse text exactly as represented in the primary Tanzil Simple Clean corpus.

  • No ad-hoc spelling changes.
  • No silent Unicode folding.
  • No merging of Arabic letter forms.
  • No Basmala removal.
78,248words332,837letters

BODY

Derived analytical view

Derived

BODY preserves all 6,236 verse identifiers but removes a precisely defined opening Basmala prefix where present.

بسم الله الرحمن الرحيم

  • S1:1 remains unchanged as the Basmala verse.
  • S9:1 remains unchanged because it has no opening Basmala prefix.
  • The four-token opening prefix is removed from the first verse of exactly 112 other surahs.
  • No other text transformation is part of the BODY rule.
77,800words330,709letters

Counting policy

Baseline exact-character rules

Words

Words are whitespace-separated units after the selected analytical view has been applied. Attached Arabic prefixes remain part of the same token.

Letters

Letters are counted as non-space Arabic letter characters exactly as represented by the Simple Clean analytical text.

No silent normalization

The baseline policy does not:

  • merge Hamza or Alif variants;
  • convert ة to another character;
  • convert ى to another character;
  • remove or introduce letters;
  • apply semantic word segmentation;
  • change spelling to obtain a desired numerical result.

Any future normalization must receive its own explicit policy name and may not silently replace the baseline exact-character analysis.

Integrity audit

Unicode and identifier audit

Encoding
UTF-8
Malformed verse lines
0
Duplicate identifiers
0
First identifier
1|1
Last identifier
114|6
Non-space characters
332,837
Combining marks
0
Punctuation characters
0
Non-Arabic non-space characters
0

Source attribution

Tanzil attribution and license

Quranic source text:Tanzil Quran Text (Simple Clean, Version 1.1)

Copyright
Copyright © 2007–2026 Tanzil Project
License
Creative Commons Attribution 3.0

Tanzil permits distribution of verbatim copies under the supplied notice, requires clear attribution to Tanzil Project, and states that the source Quran text must not be changed.

Visit Tanzil Project →

BODY is therefore documented by Coran Codex as an analytical view, not as a replacement or edited edition of the Tanzil source corpus.

Reproducibility

Study requirements

Every Coran Codex study must identify at least:

  • DatasetDATA-001
  • ViewRAW or BODY
  • NormalizationAny additional named policy
  • Counting unitThe exact unit being tested
  • TransformationAny transformation applied before analysis

A result obtained under RAW must not be presented as though it had been obtained under BODY, and vice versa.

Record history

Revision history

2026-09-15
DATA-001 baseline established with source checksum, validated RAW/BODY totals and explicit Basmala handling.