Epic 1a — Script / Unicode / font research

[Hetchy]

Parent: #27 · Plan``

Goal: Choose the Warsh script and prove it is machine-tractable and display-compatible. Engineering-flavored; runs parallel to 1b.

Tasks

• Survey Warsh scripts/sources (Quran.com / King Fahd, QUL, Tanzil, others). Per source: encoding, Unicode codepoints, diacritic conventions, verse/word segmentation.
• Assess orthography information content: does the script encode phoneme hints (preferred) or hide them so they must be inferred? Score candidates.
• Build a Unicode-audit validator (shippable tool): For a given script file: every codepoint is explainable (in an allowlist with a meaning), no stray/random characters, report unknown codepoints with locations.
• Digital font compatibility: Survey Warsh fonts compatible with the script. Asses if Digital Khatt is compatible with Warsh, or which fonts support Warsh. Assess how well each script matches the chosen display font. Document the mapping required to go from script -> font agreement.
• Comparison matrix + recommendation feeding the Epic 0 script/font decisions.

Acceptance

• dev/ validator script that fails on unexplained codepoints.
• Comparison matrix committed under quranic_phonemizer/resources/warsh/research/script-survey.md.
• A recommended (script source, font) pair with the required mapping documented.

Key refs

dev/README.md, dev/build_quran_db.py, quranic_phonemizer/resources/surah_info.json, dev/unicode_occurences/.