ߛߓߊ߬ ߞߊ߲ ߛߓߍ (Cross-Script Bridge)

Real-time translation between N'Ko, Arabic, and Latin scripts.

Overview

The Cross-Script Bridge enables seamless transliteration and translation between three major writing systems:

Script	Direction	Unicode Range	Speakers
N'Ko (ߒߞߏ)	RTL	U+07C0–U+07FF	~10M (Manding)
Arabic (العربية)	RTL	U+0600–U+06FF	~400M+
Latin	LTR	Basic ASCII	Billions

Architecture

┌─────────────────────────────────────────────────────────┐
│                  Cross-Script Bridge                     │
├─────────────────────────────────────────────────────────┤
│  ┌───────────┐    ┌───────────┐    ┌───────────┐       │
│  │   N'Ko    │◄──►│  Bridge   │◄──►│  Arabic   │       │
│  │  ߒߞߏ     │    │   Core    │    │  العربية  │       │
│  └─────┬─────┘    └─────┬─────┘    └─────┬─────┘       │
│        │                │                │              │
│        └────────────────┼────────────────┘              │
│                         │                               │
│                   ┌─────▼─────┐                         │
│                   │   Latin   │                         │
│                   │   (IPA)   │                         │
│                   └───────────┘                         │
├─────────────────────────────────────────────────────────┤
│  Features:                                              │
│  • Phonetic mapping via IPA intermediary                │
│  • Real-time streaming transliteration                  │
│  • Bidirectional conversion                             │
│  • Tone mark preservation (N'Ko)                        │
│  • Diacritic handling (Arabic)                          │
└─────────────────────────────────────────────────────────┘

Installation

cd ~/Desktop/cross-script-bridge

# Create virtual environment
python3 -m venv .venv
source .venv/bin/activate

# Install dependencies
pip install -r requirements.txt
pip install -e .

Web Interface (Gen 6)

Real-time translation in your browser:

source .venv/bin/activate
python web/app.py
# Open http://localhost:5001

Features:

🔄 Real-time WebSocket conversion as you type
🔍 Auto-detect input script
⇄ Swap input/output with one click
📊 Character analysis bar
🌐 RTL/LTR auto-switching

Usage

CLI

# N'Ko to Latin
csb transliterate "ߒߞߏ" --from nko --to latin
# Output: n'ko

# Latin to N'Ko
csb transliterate "salam" --from latin --to nko
# Output: ߛߟߊ߯ߡ

# Arabic to N'Ko
csb transliterate "سلام" --from arabic --to nko
# Output: ߛߟߊ߯ߡ

# Interactive mode
csb interactive --source nko --target latin

Linguistic Intelligence (Gen 6, Instance 18)

Beyond character mapping: semantic awareness for culturally-informed translation.

# Analyze context before translating
python -m linguistic.cli analyze "i ni tile"
# Output: greeting detected, strategy=annotate

# Look up cultural terms
python -m linguistic.cli lookup "ߌ ߣߌ ߕߌ߲"
# Output: Good day (lit: you and the day)

# Parse Manding names
python -m linguistic.cli name "Mamadou Keita"
# Output: Given=mamadou, Jamu=keita, N'Ko: ߡߡߊߘߎ ߞߋߕߊ

# Explore proverbs
python -m linguistic.cli proverbs

# Get translation suggestions
python -m linguistic.cli suggest "ko te ta la" --target nko

Features

Component	Description
ContextEngine	Detects content type (greeting, proverb, name, etc.)
CulturalDatabase	21+ terms: proverbs, greetings, blessings, concepts
NameHandler	30+ jamu, 40+ given names, honorifics with sanaku
SuggestionEngine	Intelligent alternatives when direct transliteration loses meaning

Python API

from linguistic import ContextEngine, CulturalDatabase, NameHandler

# Analyze content before translating
engine = ContextEngine()
analysis = engine.analyze("i ni tile")
print(analysis.content_type)  # greeting
print(analysis.strategy)      # annotate

# Look up cultural context
db = CulturalDatabase()
term = db.lookup("ߌ ߣߌ ߕߌ߲")
print(term.meaning)  # Good day (lit: you and the day)
print(term.usage)    # Standard daytime greeting

# Handle names properly
handler = NameHandler()
nko, confidence = handler.to_nko("Mamadou Diomande")
print(nko)  # ߡߡߊߘߎ ߖߏ߬ߡߊ߲߬ߘߋ

# Get jamu cultural info
info = handler.get_jamu_info("keita")
print(info["sanaku"])  # ['coulibaly'] - joking relationship

Content Types Detected

GREETING - Cultural greetings with proper responses
PROVERB - Wisdom sayings needing semantic translation
NAME - Personal names with jamu (clan) context
BLESSING - Religious/traditional blessings
KINSHIP - Family relationship terms (faden, baden)
CONCEPT - Cultural concepts (sanaku, jamu, faso)

Translation Strategies

Strategy	When Used
TRANSLITERATE	Plain text - direct character mapping
ANNOTATE	Cultural terms - transliterate + provide meaning
TRANSLATE	Proverbs/idioms - semantic translation preserves meaning
HYBRID	Mixed content - different strategies for different parts

CLI Streaming Edition (Gen 7, Instance 18)

The CLI has been evolved with real-time streaming capabilities:

Watch Mode

Monitor files/directories and auto-translate on changes:

# Watch a directory
csb --watch ~/notes/ -t nko

# Watch with custom output directory
csb --watch ~/docs/ --output-dir ~/docs/translated/ -t arabic

# Custom poll interval
csb --watch . --watch-interval 2.0 -t latin

Clipboard Integration

Real-time clipboard monitoring with auto-copy:

# Monitor clipboard continuously
csb --clipboard -t nko

# Translate clipboard once and copy back
csb --clipboard-once -t arabic

# Monitor without auto-copy
csb --clipboard --no-auto-copy -t latin

HTTP Translation Server

Start a mini server for integration with other tools:

# Start server on default port 8080
csb --serve

# Custom port
csb --serve 3000

API endpoints:

GET  /translate?text=hello&target=nko
POST /translate  {"text": "hello", "target": "nko"}
GET  /detect?text=ߛߟߊߡ

Batch Processing

Translate multiple files at once:

# All markdown files
csb --batch "*.md" -t nko

# Recursive with output directory
csb --batch "**/*.txt" --output-dir out/ -t arabic

# Custom suffix
csb --batch "*.md" --suffix ".nko" -t nko

Streaming & JSON Output

# Stream mode (immediate flush per line)
cat large.txt | csb --stream -t nko

# JSON output for scripting
csb --json "hello world" -t nko

Python API

from cross_script_bridge import Bridge

bridge = Bridge()

# Single conversion
result = bridge.convert("ߒߞߏ", source="nko", target="latin")
print(result)  # n'ko

# Stream mode (real-time)
async for char in bridge.stream("Hello", source="latin", target="nko"):
    print(char, end="", flush=True)

Script Mappings

N'Ko ↔ Latin (IPA-based)

N'Ko	IPA	Latin	Name
ߊ	/a/	a	A
ߋ	/o/	o	O
ߌ	/i/	i	I
ߍ	/e/	e	E
ߎ	/u/	u	U
ߏ	/ɔ/	ɔ/o	Open O
ߐ	/ə/	ə	Schwa
ߓ	/b/	b	Ba
ߘ	/d/	d	Da
ߜ	/gb/	gb	Gba
ߝ	/f/	f	Fa
ߞ	/k/	k	Ka
ߟ	/l/	l	La
ߡ	/m/	m	Ma
ߢ	/ɲ/	ny	Nya
ߣ	/n/	n	Na
ߤ	/h/	h	Ha
ߥ	/w/	w	Wa
ߦ	/j/	y	Ya
ߧ	/ŋ/	ng	Nga
ߨ	/p/	p	Pa
ߩ	/r/	r	Ra
ߪ	/s/	s	Sa

Tone Marks (N'Ko)

Mark	Unicode	Meaning
߬	U+07EC	Falling tone
߭	U+07ED	High tone
߯	U+07EF	Long vowel

HEF Evolution

Task: Cross-Script Bridge Instances:

inst_20260131082128_773 (Gen 9)
Instance 18 (Gen 6 - Telegram Bot)

Generation: 10 (PWA + Offline) / 6 (Instance 18 - Telegram Bot) Techniques Applied:

G02 (Attribute Listing) - Script characteristics
G05 (SCAMPER) - Adapt IPA as intermediary
G08 (Combine) - Merge three script systems
R03 (Simplify) - Minimal phonetic mapping
R08 (Integrate) - Unified API
G11 (Add Senses) - Audio/voice capabilities
R05 (User Testing) - Practice mode with progress tracking
D05 (Digital Deployment) - Browser extension for web-wide translation
D14 (Universal Access) - Translate any text on any webpage
D01 (API Distribution) - Public REST API for third-party integration
D06 (SDK Packaging) - Python & JavaScript SDKs for developers
D10 (Mobile Packaging) - PWA for mobile installation
R12 (Backup Plan) - Offline-first architecture with Service Worker
D15 (Desktop Integration) - Raycast extension for system-wide access
R14 (Accessibility) - One-click translation via launcher
D08 (Channel Distribution) - Telegram Bot for messaging platform integration
D15 (Desktop Integration) - macOS Menu Bar for always-on access
D14 (Universal Access) - Apple Shortcuts & Services for system-wide integration
G11 (Add Senses) - Voice dictation mode with real-time transcription

Voice Dictation Mode (Gen 6, Instance 18)

Speak and get N'Ko output in real-time. Perfect for native Manding speakers who want to write in N'Ko without learning keyboard layouts.

CLI Voice Mode

# Single phrase dictation
csb --voice -t nko

# Continuous dictation (keeps listening)
csb --voice --voice-continuous

# Auto-copy results to clipboard
csb --voice --voice-auto-copy

# With Arabic output
csb --voice -t arabic

Features

Feature	Description
🎤 Real-time Transcription	Speech-to-text via Whisper or system APIs
⚡ Instant Transliteration	Immediate conversion to N'Ko/Arabic
📊 Visual Feedback	Waveform display during recording
📋 Clipboard Integration	Auto-copy results
🔄 Continuous Mode	Keep listening phrase after phrase
🌐 Multi-backend	whisper.cpp, OpenAI API, or macOS native

Web Voice Input

The web interface includes Web Speech API integration:

<script src="voice.js"></script>
<script>
const voiceBtn = new VoiceButton('#voice-container', bridge, {
    targetScript: 'nko',
    onResult: (result) => {
        console.log('N\'Ko:', result.text);
        console.log('Confidence:', result.confidence);
    }
});
</script>

Voice Architecture

┌─────────────────────────────────────────────────┐
│              Voice Pipeline                      │
├─────────────────────────────────────────────────┤
│  🎤 Microphone Input                            │
│       │                                          │
│       ▼                                          │
│  ┌─────────────────┐                            │
│  │ Speech-to-Text  │                            │
│  │ • whisper.cpp   │                            │
│  │ • OpenAI API    │                            │
│  │ • macOS Native  │                            │
│  │ • Web Speech    │                            │
│  └────────┬────────┘                            │
│           │                                      │
│           ▼                                      │
│  ┌─────────────────┐                            │
│  │ Latin Text      │  "salam alaikum"           │
│  └────────┬────────┘                            │
│           │                                      │
│           ▼                                      │
│  ┌─────────────────┐                            │
│  │ Bridge Core     │                            │
│  │ Transliterate   │                            │
│  └────────┬────────┘                            │
│           │                                      │
│           ▼                                      │
│  ┌─────────────────┐                            │
│  │ N'Ko Output     │  ߛߟߊ߯ߡ ߊ߬ߟߊ߯ߌߞߎߡ          │
│  └─────────────────┘                            │
└─────────────────────────────────────────────────┘

Cultural Impact

Voice input addresses a critical accessibility gap:

Oral Tradition: Manding languages have strong oral traditions; voice input feels natural
Script Learning: Speakers can produce N'Ko text before mastering the script
Literacy Bridge: Helps bridge oral fluency to written literacy
Mobile Access: Easier than typing on small screens

ߞߊ߲ ߦߋ߫ ߣߎ߲߫ ߕߍ߫ — The language is between us

Apple Shortcuts & Services (Gen 6, Instance 18)

Access translation from Siri, Shortcuts app, and right-click menus.

Quick Start

# Translate via CLI
echo "hello" | python shortcuts/bridge_shortcuts.py translate
# Output: ߤߋߟߟߏ

# Translate to Arabic
python shortcuts/bridge_shortcuts.py translate --target arabic "peace"

# Detect script
python shortcuts/bridge_shortcuts.py detect "ߒߞߏ"
# {"script": "nko", "confidence": 0.9, "name": "ߒߞߏ (N'Ko)"}

# Install macOS Quick Action
python shortcuts/bridge_shortcuts.py install-service

Siri Integration

After setting up a Shortcut:

"Hey Siri, translate to N'Ko"

Right-Click Translation

Run python shortcuts/bridge_shortcuts.py install-service
Select any text in any app
Right-click → Services → "Translate with Cross-Script Bridge"

Available Commands

Command	Description
`translate`	Translate text (stdin or argument)
`detect`	Identify script of text
`greeting`	Get greeting in script
`scripts`	List available scripts
`install-service`	Install macOS Quick Action

See shortcuts/README.md for full documentation.

Progressive Web App (Gen 10)

Mobile-first, offline-capable translation app.

Start PWA Server

cd pwa
python server.py
# Open http://localhost:5003

Features

📱 Installable - Add to home screen on iOS/Android
⚡ Offline Mode - Full translation works without internet
🔄 Auto-detect - Automatically detects input script
📋 Quick Copy - Tap output to copy
🔗 Share Target - Share text from other apps
🔔 Push Notifications - Learning reminders (optional)
⇄ Swap Scripts - One-tap source/target swap

Installation

Android/Chrome:

Visit http://localhost:5003 in Chrome
Tap "Add to Home Screen" prompt
App icon appears on home screen

iOS/Safari:

Visit http://localhost:5003 in Safari
Tap Share → Add to Home Screen
App appears as native icon

Offline Architecture

┌─────────────────────────────────────────────┐
│              Service Worker                  │
├─────────────────────────────────────────────┤
│  Cache Strategy:                             │
│  • Core Assets: Cache-first                  │
│  • API Calls: Network-first + cache fallback │
│  • Pages: Network-first + offline fallback   │
├─────────────────────────────────────────────┤
│  Cached Resources:                           │
│  • HTML, CSS, JS                             │
│  • Script mapping data (JSON)                │
│  • N'Ko & Arabic fonts                       │
│  • Translation engine (client-side)          │
└─────────────────────────────────────────────┘

PWA Files

pwa/
├── manifest.json       # PWA manifest
├── service-worker.js   # Offline caching
├── index.html          # Main app
├── offline.html        # Offline fallback
├── server.py           # Development server
├── js/
│   ├── bridge-core.js  # Translation engine
│   └── app.js          # App logic
├── css/
│   └── style.css       # Styles
├── fonts/              # Cached fonts
└── icons/              # App icons

Cultural Impact

This PWA is designed for West African users where:

Internet connectivity can be intermittent
Mobile-first is essential
N'Ko script preservation is culturally important
Offline capability enables learning anywhere

iOS Keyboard Extension (Gen 7, Instance 18)

Native iOS custom keyboard for system-wide N'Ko typing.

Location: ios-keyboard/

Features

Feature	Description
🔄 Real-time Transliteration	Type in Latin, see N'Ko instantly
⌨️ Dual Input Modes	Latin-to-N'Ko or direct N'Ko character input
💡 Smart Suggestions	Common word predictions with meanings
📱 System-wide	Works in any app - Messages, Notes, Safari
🔢 N'Ko Numerals	Full digit support (߀-߉)
⚡ Offline	No network required

Installation (Development)

cd ios-keyboard
open .  # Create Xcode project from files

Create iOS App project in Xcode
Add Keyboard Extension target
Copy KeyboardViewController.swift to extension
Build and run on device
Enable in Settings → General → Keyboard → Add New Keyboard

Usage

Latin Mode (Default):

Type salam → ߛߟߊ߯ߡ
Type nko → ߒߞߏ
Digraphs: ny → ߢ, gb → ߜ, ng → ߣ

N'Ko Direct Mode:

Tap ⇧ to switch to direct N'Ko input
Double-tap ⇧ for caps lock

Keyboard Layout

Latin Mode:          N'Ko Mode (⇧):
┌──────────────┐     ┌──────────────┐
│q w e r t y u│     │ߊ ߓ ߔ ߕ ߖ ߗ ߘ│
│ a s d f g h │     │ ߜ ߝ ߞ ߟ ߠ ߡ│
│⇧ z x c v b ⌫│     │⇧ ߥ ߦ ߧ ߨ ߋ ⌫│
│123 🌐 space ⏎│     │ABC 🌐 ߛߓߊ ⏎│
└──────────────┘     └──────────────┘

Cultural Impact

Enables N'Ko typing anywhere on iOS — Messages, Notes, Safari, social media. Brings script accessibility to the mobile-first generation.

Public API (Gen 9)

Start the API server:

source .venv/bin/activate
pip install flask-socketio flask-cors
python api/server.py
# Running on http://localhost:5002

Endpoints

Method	Endpoint	Description
`POST`	`/v1/translate`	Translate between scripts
`POST`	`/v1/detect`	Detect script type
`POST`	`/v1/analyze`	Phonetic analysis
`POST`	`/v1/batch`	Batch translation
`GET`	`/v1/scripts`	List supported scripts
`WS`	`/v1/stream`	Real-time streaming

Example Request

curl -X POST http://localhost:5002/v1/translate \
  -H "Content-Type: application/json" \
  -H "X-API-Key: demo-free-key" \
  -d '{"text": "ߒߞߏ", "target": "latin"}'

Response:

{
  "success": true,
  "input": "ߒߞߏ",
  "output": "n'ko",
  "source": "nko",
  "target": "latin",
  "detected": true
}

Rate Limits

Tier	Requests/min	Batch Size
Free	100	10
Pro	1000	100
Unlimited	∞	1000

SDKs (Gen 9)

Python SDK

cd sdk/python
pip install -e .

from cross_script_bridge import CrossScriptBridge

csb = CrossScriptBridge(api_key='demo-free-key')

# Translate
result = csb.translate("ߒߞߏ", target='latin')
print(result.output)  # n'ko

# Quick methods
nko = csb.to_nko("Hello")        # ߤߋߟߏ
latin = csb.to_latin("ߒߞߏ")     # n'ko

# Batch
results = csb.batch([
    {'text': 'ߒߞߏ', 'target': 'latin'},
    {'text': 'Hello', 'target': 'nko'}
])

JavaScript SDK

cd sdk/javascript
npm install

import { CrossScriptBridge } from 'cross-script-bridge';

const csb = new CrossScriptBridge({ apiKey: 'demo-free-key' });

// Translate
const result = await csb.translate('ߒߞߏ', { target: 'latin' });
console.log(result.output); // n'ko

// Quick methods
const nko = await csb.toNko('Hello');   // ߤߋߟߏ
const latin = await csb.toLatin('ߒߞߏ'); // n'ko

Browser Extension (Gen 8)

Translate any text on any webpage with the Cross-Script Bridge extension.

Installation

Open Chrome/Edge and go to chrome://extensions/
Enable "Developer mode" (top right)
Click "Load unpacked"
Select the extension/ folder

Features

🔍 Auto-detect - Automatically identifies N'Ko, Arabic, or Latin text
🖱️ Right-click menu - Select text → Right-click → "Translate Script"
💬 Instant tooltip - Hover translation appears on selection
⌨️ Keyboard shortcut - Alt+Shift+T translates selected text
📋 One-click copy - Copy translation to clipboard
⚙️ Customizable - Set default target script, toggle tooltip

Popup Translator

Click the extension icon for a full translator interface:

Enter text to translate
Swap source/target with one click
Quick phrase buttons for common greetings
Settings panel for preferences

Files

extension/
├── manifest.json          # Extension manifest (v3)
├── background/
│   └── service-worker.js  # Context menus & shortcuts
├── content/
│   ├── bridge-core.js     # Translation engine (JS port)
│   ├── content-script.js  # Selection handler
│   └── tooltip.css        # Tooltip styles
├── popup/
│   ├── popup.html         # Popup interface
│   ├── popup.css          # Popup styles
│   └── popup.js           # Popup logic
└── icons/
    └── icon*.png          # Extension icons

CLI Tool (Gen 6 - Instance 18)

Command-line interface for translation, scripting, and automation.

Usage

# Direct translation (auto-detect source)
csb "ߛߊ߬ߟߊ߯ߡߎ߲"                    # → salaamun

# Specify target script
csb -t nko "salam"                     # → ߪߑߟߑߡ
csb -t arabic "peace"                  # → Arabic output

# Pipe support for scripting
echo "ߒ'ߞߏ ߞߊ߲" | csb                  # → n'ko kan
cat document.txt | csb -t nko > translated.txt

# File translation
csb -f input.txt -o output.txt -t nko

# Verbose mode (show source detection)
csb -v "ߛߊ߬ߟߊ߯ߡߎ߲"                   # → [ߒ'ߞߏ → Latin] salaamun

# Interactive REPL
csb -i

Interactive REPL

╭─────────────────────────────────────────╮
│ ߞ Cross-Script Bridge ع A              │
│   Real-time script transliteration     │
╰─────────────────────────────────────────╯

Target script: Latin
Type text to translate. Commands: /nko /arabic /latin /quit

csb → Latin: ߒߞߏ
  [ߒ'ߞߏ → Latin] n'ko

csb → Latin: /nko
Target set to ߒ'ߞߏ

csb → ߒ'ߞߏ: hello world
  [Latin → ߒ'ߞߏ] ߤߋߟߏ ߥߏߙߟߘ

Options

Flag	Long	Description
`-t`	`--target`	Target script: `nko`, `arabic`, `latin`
`-s`	`--source`	Source script (auto-detect if omitted)
`-i`	`--interactive`	Interactive REPL mode
`-f`	`--file`	Input file to translate
`-o`	`--output`	Output file (stdout if omitted)
`-v`	`--verbose`	Show source/target info
	`--no-color`	Disable colored output

Integration Examples

# Translate clipboard content (macOS)
pbpaste | csb -t nko | pbcopy

# Watch file for changes
fswatch -o input.txt | xargs -n1 -I{} csb -f input.txt -o output.txt

# Git hook: translate commit message
csb -t nko "$(git log -1 --format=%s)" >> nko_changelog.txt

Raycast Extension (Gen 7, Instance 18)

System-wide instant translation from your macOS launcher.

Installation

cd ~/Desktop/cross-script-bridge/raycast
npm install
npm run dev  # Development mode

Commands

Command	Description
Translate Script	Full translation interface with real-time conversion
Translate Clipboard	One-shot clipboard translation (⌘C to copy result)
Quick N'Ko Input	Fast Latin → N'Ko with phrase suggestions

Features

⚡ Real-time translation as you type
🔍 Auto-detect source script
📋 Clipboard integration with auto-copy
⇄ Swap source/target with ⌘S
📖 Quick phrases library (10 common greetings)
🕐 Recent translations history
🖥️ Works offline — standalone TypeScript engine

Keyboard Shortcuts

Shortcut	Action
⌘C	Copy translation
⌘⇧V	Paste translation
⌘S	Swap scripts
⌘P	Quick phrases
⌘K	Clear

Configuration

Set preferences in Raycast → Extensions → Cross-Script Bridge:

Default source/target scripts
Auto-copy behavior
Bridge server URL (for advanced translation)

macOS Menu Bar App (Gen 6.1, Instance 18)

Always-on translation from your menu bar. No browser needed.

Installation

source .venv/bin/activate
pip install rumps
./menubar/launch.sh

 ┌─────────────────────────────────────────┐
 │  ߛ⟷  📋 Translate Clipboard     ⌘T    │
 │       ✏️ Type to Translate...   ⌘N    │
 │       ─────────────────────────       │
 │       Source Script        ▸          │
 │       Target Script        ▸          │
 │       ⇄ Swap Source/Target  ⌘S       │
 │       ─────────────────────────       │
 │       📜 History           ▸          │
 │       🗣️ Quick Phrases     ▸          │
 │       ─────────────────────────       │
 │       🌐 Open Web Interface ⌘W       │
 │       ⚙️ Settings...                  │
 └─────────────────────────────────────────┘

Features

Feature	Description
📋 Clipboard Translation	⌘T translates clipboard and copies result back
✏️ Type to Translate	Quick input window for instant conversion
🔍 Auto-detect	Automatically identifies N'Ko/Arabic/Latin
⇄ Swap Scripts	One-click source/target swap
📜 History	Last 10 translations saved
🗣️ Quick Phrases	Common N'Ko greetings ready to copy
🌐 Web Link	Jump to full web interface

Keyboard Shortcuts

Shortcut	Action
⌘T	Translate clipboard
⌘N	Type to translate
⌘S	Swap source/target scripts
⌘W	Open web interface
⌘Q	Quit

Menu Bar Icon

The icon shows your current translation direction:

ߛ⟷A — N'Ko to Latin
?⟷ߛ — Auto-detect to N'Ko
ع⟷ߛ — Arabic to N'Ko

Launch at Login

System Settings → General → Login Items
Click + under "Open at Login"
Add ~/Desktop/cross-script-bridge/menubar/launch.sh

Audio & Pronunciation (Gen 7)

Pronunciation Guide Web UI

source .venv/bin/activate
python web/app.py
# Open http://localhost:5001/pronunciation

Features:

🔊 Text-to-speech for any script
📊 IPA phonetic transcription
🎯 Character-by-character breakdown
🏆 Spaced repetition practice mode
📈 Progress tracking with mastery levels

Python API

from audio.pronunciation import PronunciationEngine
from audio.phoneme import PhonemeMapper
from audio.tts import ScriptTTS

# Get pronunciation breakdown
engine = PronunciationEngine()
breakdown = engine.breakdown("ߒߞߏ")
print(breakdown['ipa'])  # nko
print(breakdown['difficulty'])  # beginner

# IPA transcription
mapper = PhonemeMapper()
ipa = mapper.to_ipa_string("سلام")  # sla:m

# Text-to-speech (async)
import asyncio
tts = ScriptTTS()
audio = asyncio.run(tts.speak("ߒߞߏ"))

Practice Mode

# Get characters to practice
practice_set = engine.get_practice_set('nko', count=5)

# Record practice results
result = engine.practice_character('ߓ', was_correct=True)
print(result['mastery_level'])  # 0-5

# Check learning stats
stats = engine.get_stats()

Telegram Bot (Gen 6, Instance 18)

Real-time translation in Telegram chats — highly impactful for West African communities.

Setup

cd telegram-bot
pip install -r requirements.txt

# Get token from @BotFather on Telegram
export TELEGRAM_BOT_TOKEN='your_token'
python bot.py

Features

Feature	Description
📝 Direct Translation	Send any text, get instant translation
⚡ Inline Mode	`@botname text` works in any chat
💬 Group Support	`/translate` command for groups
🎯 Quick Commands	`/nko`, `/arabic`, `/latin` shortcuts
⚙️ User Preferences	Set default target script with `/target`
📖 Quick Phrases	Common N'Ko greetings via `/phrases`
📊 Statistics	Track usage with `/stats`

Usage

Direct message:

You: ߛߟߊ߯ߡ
Bot: 🇬🇳 N'Ko → 🔤 Latin: salam

Commands:

/nko hello        → ߤߋߟߏ
/arabic peace     → سلام
/latin ߒߞߏ       → n'ko

Inline mode (any chat):

@your_bot_username ߌ ߣߌ ߞߍ

Cultural Impact

Telegram is widely used in West Africa. This bot:

Enables N'Ko in everyday messaging
Supports language communities in group chats
Makes the script accessible without installing keyboards
Works offline with standalone translation engine

Cultural Context

This bridge serves multiple purposes:

N'Ko Preservation - Making the script accessible to learners
Cross-cultural Communication - Manding ↔ Arabic ↔ Western
AI Training Data - Generating parallel corpora
Digital Griot Tradition - Oral history in written form

Files

cross-script-bridge/
├── README.md               # This file
├── setup.py                # Package setup
├── requirements.txt        # Dependencies
├── core/
│   ├── __init__.py
│   ├── bridge.py           # Main Bridge class
│   ├── nko.py              # N'Ko handler
│   ├── arabic.py           # Arabic handler
│   └── latin.py            # Latin/IPA handler
├── audio/                  # Gen 7
│   ├── __init__.py
│   ├── phoneme.py          # PhonemeMapper - IPA transcription
│   ├── pronunciation.py    # PronunciationEngine - Learning mode
│   └── tts.py              # ScriptTTS - Text-to-speech
├── api/                    # Gen 9 - Public REST API
│   └── server.py           # Flask + SocketIO API server
├── sdk/                    # Gen 9 - Developer SDKs
│   ├── python/             # Python SDK
│   │   ├── cross_script_bridge/
│   │   │   ├── __init__.py
│   │   │   └── client.py
│   │   ├── setup.py
│   │   └── README.md
│   └── javascript/         # JavaScript SDK
│       ├── index.js
│       ├── package.json
│       └── README.md
├── pwa/                    # Gen 10 - Progressive Web App
│   ├── manifest.json       # PWA manifest
│   ├── service-worker.js   # Offline caching
│   ├── index.html          # Main app UI
│   ├── offline.html        # Offline fallback
│   ├── server.py           # Development server
│   └── js/                 # Client-side bridge
├── raycast/                # Gen 7 (Instance 18) - Raycast Extension
│   ├── package.json        # Extension manifest
│   ├── tsconfig.json       # TypeScript config
│   └── src/
│       ├── lib/bridge.ts   # Standalone translation engine
│       ├── translate.tsx   # Full translation UI
│       ├── clipboard-translate.tsx  # One-shot clipboard
│       └── quick-nko.tsx   # Fast Latin → N'Ko
├── telegram-bot/           # Gen 6 (Instance 18) - Telegram Bot
│   ├── bot.py              # Main Telegram bot
│   ├── bridge_core.py      # Standalone translation fallback
│   ├── requirements.txt    # Bot dependencies
│   └── README.md           # Bot documentation
├── extension/              # Gen 8 - Browser Extension
│   ├── manifest.json       # Chrome/Edge manifest v3
│   ├── background/         # Service worker
│   ├── content/            # Content scripts & styles
│   ├── popup/              # Extension popup UI
│   └── icons/              # Extension icons
├── web/
│   ├── app.py              # Flask + SocketIO server
│   └── templates/
│       ├── index.html      # Real-time translator
│       └── pronunciation.html  # Pronunciation learning UI
├── data/
│   ├── nko_map.json        # N'Ko ↔ IPA mappings
│   ├── arabic_map.json     # Arabic ↔ IPA mappings
│   └── cognates.json       # Cross-language cognates
├── cli/                    # Gen 7 (Instance 18) - Streaming CLI
│   ├── __init__.py
│   └── csb.py              # CLI with watch/clipboard/server modes
├── cache/                  # Audio cache (generated)
├── scripts/                # Utility scripts
└── tests/
    └── test_bridge.py      # Unit tests

License

MIT - For the preservation of N'Ko and cross-cultural understanding.

ߊ߬ ߣߌ߫ ߛߓߍ ߕߐ߭ ߓߍ߯ ߦߋ߫ - "The word and all writing are one"

Collaborative Translation (Gen 6, Instance 18)

Real-time multi-user translation sessions for language learning and cultural preservation.

Quick Start

source .venv/bin/activate
python collab/server.py
# Open http://localhost:5002

Features

Feature	Description
🔴 Live Sessions	Multiple users contribute text simultaneously
⚡ Real-time Sync	See translations as others type
🔄 Script Auto-detect	Input in any script, get all translations
✅ Verification	Moderators can verify contributions
📚 Phrase Libraries	Build shared vocabulary in Rooms
🔗 Invite Links	Share session/room access
💾 Corpus Export	Generate parallel corpora for AI training

Architecture

┌─────────────────────────────────────────────────────────┐
│               Collaborative Translation                  │
├─────────────────────────────────────────────────────────┤
│  ┌─────────────────────────────────────────────────┐   │
│  │              CollaborativeSession                │   │
│  │  • Multiple participants                         │   │
│  │  • Real-time contributions                       │   │
│  │  • Auto-translate to all scripts                 │   │
│  └─────────────────────────────────────────────────┘   │
│                         │                               │
│  ┌───────────┐   ┌─────▼─────┐   ┌───────────────┐    │
│  │BridgeRoom │   │ScriptSync │   │ConflictResolve│    │
│  │(persistent)│   │(real-time)│   │(merge logic) │    │
│  └───────────┘   └───────────┘   └───────────────┘    │
│                         │                               │
│  ┌─────────────────────▼───────────────────────────┐   │
│  │              WebSocket Server                    │   │
│  │  • Socket.IO real-time events                    │   │
│  │  • HTTP REST API                                 │   │
│  │  • Web UI at localhost:5002                      │   │
│  └─────────────────────────────────────────────────┘   │
└─────────────────────────────────────────────────────────┘

Python API

from collab import CollaborativeSession, BridgeRoom, MemberRole

# Create a translation session
session = CollaborativeSession("N'Ko Learning Group")

# Add participants
host = session.add_participant("Moussa", preferred_script="nko", is_moderator=True)
learner = session.add_participant("David", preferred_script="latin")

# Contribute text (auto-translated to all scripts)
contribution = session.contribute(host.id, "ߒ ߕߐ߯ߞߏ ߦߋ߫ ߡߎߛߊ")
print(contribution.translations["latin"])  # "n tooko ye musa"

# Create persistent room for ongoing collaboration
room = BridgeRoom("West African Languages Circle")
admin = room.add_member("Aissata", role=MemberRole.ADMIN)

# Build shared phrase library
room.add_phrase(
    nko="ߌ ߣߌ ߕߌ߲",
    arabic="إي ني تيل",
    latin="i ni tile",
    meaning="Good day",
    category="greeting",
    contributor_id=admin.id,
    cultural_note="Daytime greeting in Manding languages"
)

# Export corpus for AI training
corpus = session.export_corpus("training_data.json")

Real-time Sync

from collab import ScriptSyncEngine, ConflictResolver

# Create sync engine for a session
sync = ScriptSyncEngine()

# Register participants
sync.register("user_1", "Moussa")
sync.register("user_2", "Fatou")

# Operations are synchronized with vector clocks
op1 = sync.insert("user_1", "doc_1", 0, "Hello")
op2 = sync.insert("user_2", "doc_1", 0, "Hi")  # Potential conflict!

# Auto-resolve conflicts
resolver = ConflictResolver()
for conflict in sync.get_pending_conflicts():
    suggestion = resolver.suggest_resolution(
        conflict.operation_a.content,
        conflict.operation_b.content,
        script="latin"
    )
    print(f"Suggested: {suggestion['strategy']} @ {suggestion['confidence']*100}%")

Use Cases

Language Learning Groups - Students practice N'Ko together with instant feedback
Digital Preservation - Communities collaboratively document oral traditions
AI Training Data - Generate verified parallel corpora for translation models
Cross-cultural Communication - Bridge Manding ↔ Arabic ↔ Western languages

Cultural Impact

This collaborative system enables:

Distributed N'Ko teaching - Teachers can work with students remotely
Community phrase building - Preserve regional vocabulary and proverbs
Verified translations - Moderator approval ensures quality
Living corpora - Continuously growing datasets for AI

ߒߞߏ ߓߍ߯ ߦߋ߫ ߓߏ߲߬ ߞߋߟߋ߲ ߘߐ߫ — "All N'Ko speakers are one family"

Name		Name	Last commit message	Last commit date
.github/workflows		.github/workflows
api		api
audio		audio
cli		cli
collab		collab
core		core
data		data
dialect		dialect
extension		extension
ios-keyboard		ios-keyboard
linguistic		linguistic
memory		memory
menubar		menubar
morphology		morphology
pipeline		pipeline
pwa		pwa
quality		quality
raycast		raycast
scripts		scripts
sdk		sdk
shortcuts		shortcuts
telegram-bot		telegram-bot
tests		tests
voice		voice
web		web
DEP_REPORT.md		DEP_REPORT.md
README.md		README.md
requirements.txt		requirements.txt
setup.py		setup.py

Latest commit

History

Folders and files

Repository files navigation

ߛߓߊ߬ ߞߊ߲ ߛߓߍ (Cross-Script Bridge)

Overview

Architecture

Installation

Web Interface (Gen 6)

Usage

CLI

Linguistic Intelligence (Gen 6, Instance 18)

Features

Python API

Content Types Detected

Translation Strategies

CLI Streaming Edition (Gen 7, Instance 18)

Watch Mode

Clipboard Integration

HTTP Translation Server

Batch Processing

Streaming & JSON Output

Python API

Script Mappings

N'Ko ↔ Latin (IPA-based)

Tone Marks (N'Ko)

HEF Evolution

Voice Dictation Mode (Gen 6, Instance 18)

CLI Voice Mode

Features

Web Voice Input

Voice Architecture

Cultural Impact

Apple Shortcuts & Services (Gen 6, Instance 18)

Quick Start

Siri Integration

Right-Click Translation

Available Commands

Progressive Web App (Gen 10)

Start PWA Server

Features

Installation

Offline Architecture

PWA Files

Cultural Impact

iOS Keyboard Extension (Gen 7, Instance 18)

Features

Installation (Development)

Usage

Keyboard Layout

Cultural Impact

Public API (Gen 9)

Endpoints

Example Request

Rate Limits

SDKs (Gen 9)

Python SDK

JavaScript SDK

Browser Extension (Gen 8)

Installation

Features

Popup Translator

Files

CLI Tool (Gen 6 - Instance 18)

Usage

Interactive REPL

Options

Integration Examples

Raycast Extension (Gen 7, Instance 18)

Installation

Commands

Features

Keyboard Shortcuts

Configuration

macOS Menu Bar App (Gen 6.1, Instance 18)

Installation

Menu

Features

Keyboard Shortcuts

Menu Bar Icon