The Problem

If your notes contain vocalized Hebrew like בְּרֵאשִׁית, searching for בראשית (without vowels) won't find it. The same happens with café vs cafe, Arabic بِسْمِ vs بسم, and any other diacritical marks.

Diacritics-Free Search solves this by making all diacritics transparent during search — type without marks and find text that has them.

Search query:    בראשית        →  Finds:  בְּרֵאשִׁית ✓
Search query:    cafe           →  Finds:  café ✓
Search query:    بسم            →  Finds:  بِسْمِ ✓

---

Features

🔍 Local Search — Find & Replace in Current Note

Shortcut	Option + F
Behavior	Opens a search bar at the top of the active note (Chrome-style)

Capabilities:

• Real-time highlighting — All matches highlighted in yellow; current match in orange
• Match navigation — Use ▲▼ buttons or Enter / Shift+Enter to jump between matches
• Scrollbar markers — Yellow ticks on the scrollbar show where matches are (clickable!)
• Replace one / Replace all — Replace the current match or all matches at once
• Case sensitivity toggle — Click "Aa" to switch between case-sensitive and insensitive
• Remembers last search — Reopen the bar and your previous query is still there
• Escape to close — Press Esc from anywhere (even while editing) to dismiss

---

🌐 Global Search — Search Entire Vault

Shortcut	Shift + Option + F
Behavior	Opens a modal to search across all markdown files in your vault

Capabilities:

• Vault-wide search — Scans every .md file in your vault
• Highlighted context — Shows the matching line with the found text highlighted
• Click to navigate — Click any result to open the file and jump to the exact match
• Match count per file — See how many matches each file contains
• Global Replace All — Replace a term across all files in one click
• Debounced input — Waits 300ms after you stop typing to avoid lag on large vaults

---

🎯 How Matching Works

The plugin uses Unicode NFD normalization to decompose characters, then removes all combining marks (Unicode category Mn):

Original	Normalized	Script
בְּרֵאשִׁית	בראשית	Hebrew (nikud + cantillation)
café	cafe	Latin (acute accent)
بِسْمِ ٱللَّهِ	بسم الله	Arabic (tashkil)
naïve	naive	Latin (diaeresis)
Ἀθῆναι	Αθηναι	Greek (polytonic)
résumé	resume	Latin (multiple accents)
über	uber	Latin (umlaut)

💡 Both sides are normalized — if you search with diacritics, it still matches text without them (and vice versa).

---

⌨️ Keyboard Shortcuts

Action	Shortcut
Open local search	Option + F
Open global search	Shift + Option + F
Next match	Enter or ▼ button
Previous match	Shift + Enter or ▲ button
Close search	Esc (works from anywhere)
Replace current	Enter in replace field

Both commands are also available via the Command Palette (Cmd + P).

---

🗂️ Supported Scripts

Script	Example	What's stripped
🇮🇱 Hebrew	בְּרֵאשִׁית → בראשית	Nikud, cantillation marks
🇸🇦 Arabic	بِسْمِ → بسم	Tashkil / harakat
🇫🇷 Latin	café → cafe	Accents, umlauts, cedillas, tildes
🇬🇷 Greek	Ἀθῆναι → Αθηναι	Polytonic accents, breathings
🇷🇺 Cyrillic	й → и	Combining breve
🇮🇳 Devanagari	—	Anusvara, virama, etc.
🌍 Any	—	All Unicode Mn (Mark, Nonspacing)

---

Installation

Manual Installation

1. Download main.js and manifest.json from the latest release
2. Create folder: <your-vault>/.obsidian/plugins/diacritics-free-search/
3. Place both files inside
4. Restart Obsidian
5. Go to Settings → Community Plugins → Enable "Diacritics-Free Search"

Building from Source

git clone https://github.com/spenhos/obsidian-diacritics-free-search.git
cd obsidian-diacritics-free-search
npm install
npm run build

Copy main.js + manifest.json to your vault's plugin folder.

---

Settings

Setting	Description	Default
Case sensitive by default	Start searches with case sensitivity enabled	Off

---

Contributing

Issues and PRs are welcome! If you find a script or language where diacritics aren't being properly stripped, please open an issue with an example.

---