---

---

📜 Lontar Balinese Dictionary (LBD)

The historical dictionary of the Balinese language — Balinese · Indonesian · English

Named after the palm leaf manuscripts of Bali — the primary medium through which Balinese knowledge, literature, and language have been preserved for over a millennium. LBD is the first attempt at an OED-grade historical dictionary of the Balinese language.

---

What is LBD?

LBD is an open, corpus-backed, historically grounded dictionary of the Balinese language — trilingual across Balinese, Indonesian, and English. Each entry traces a word from its earliest attested form in lontar manuscripts and copper plate inscriptions (prasasti) through its modern usage, with full register (unda-usuk) stratification and etymological depth.

The OED is the reference model: not a translation dictionary, not a learner's dictionary, but a scholarly record of what words mean, what they have meant, and where they come from — grounded in primary evidence.

lema: padem
aksara: ᬧᬤᭂᬫ᭄
register: alus singgih
pos: verba
etymology: Kawi paḍem < Sanskrit padma (lotus) → euphemistic register shift
senses:
  1. [ritual/formal] meninggal dunia (of royalty, high caste) — to die (honorific)
attestations:
  - Prasasti Sukawana A1 (~882 CE): "...sang hyang paḍem..."
  - Lontar Usada Rare (undated, Gedong Kirtya LT 1247): "yan padem sapisan..."
register_equivalents:
  andap: mati
  alus_sor: seda
  alus_singgih: padem

---

Relationship to the Lontar Ecosystem

LBD is built on two existing projects:

Project	Role in LBD
lontar-ocr	Corpus ingestion — converts manuscript images to Unicode text, feeds the attestation layer
lontar	Rendering and export — Aksara Bali text shaping (lontar-aksara), scholarly output formats (DOCX, PDF, LaTeX)

LBD is not a standalone project. It is the dictionary layer that makes the outputs of lontar-ocr and lontar meaningful as preservation infrastructure.

---

What Makes LBD Different

Feature	Modern Kamus (Balai Bahasa)	BASAbali Wiki	LBD
Trilingual (Bali–ID–EN)	Bali–ID only	Bali–ID–EN	✅
Etymology	❌	❌	✅ Full chain
Earliest attestation	❌	❌	✅ Corpus-backed
Unda-usuk register cross-links	Partial	Partial	✅ Structured
Aksara Bali per entry	❌	Partial	✅
IPA transcription	❌	❌	✅
Dialect tagging	❌	❌	✅
Open structured data	❌	Partial	✅ CC BY-SA
Historical depth	❌	❌	✅ OED-model

---

Project Status

Phase	Status	Goal
0	🔴 Design	Schema, data model, editorial governance
1	🔴 Design	Seed database (5,000–10,000 entries), REST + GraphQL API
1.5	🔴 Design	Public web UI — entry pages, search, browse, corpus viewer
2	🔴 Design	Aksara Bali rendering, lontar-aksara integration
3	🔴 Design	Corpus attestation pipeline via lontar-ocr
4	🔴 Design	Editorial interface — review queue, diff view, dispute workflow
5	🔴 Design	Institutional partnerships, Van der Tuuk extraction

Status legend: 🔴 Design · 🟡 In Progress · 🟢 Usable · ✅ Stable

---

Primary Sources

Manuscript Collections

• Gedong Kirtya, Singaraja — largest lontar collection in the world (~3,000+ texts)
• Leiden University Library — Van der Tuuk bequest, partially digitized
• Udayana University — active transliteration program
• Museum Bali, Denpasar

Foundational Lexicographic Works

• Van der Tuuk, Kawi-Balineesch-Nederlandsch Woordenboek (1897–1912) — 4 vols, ~3,600 pp — the historical backbone, public domain
• Zoetmulder, Old Javanese-English Dictionary (1982) — Kawi layer reference
• Balai Bahasa Provinsi Bali, Kamus Bali–Indonesia (3rd ed.) — modern lemma base
• Sutjaja, Tuttle Concise Balinese Dictionary — ~18,000 entries, register-aware

---

Architecture

See ARCHITECTURE.md for the full technical design.

Roadmap

See TODO.md for the detailed task breakdown.

---

Getting Started (Scaffold)

Prerequisites

• Rust toolchain (1.76+ recommended)
• PostgreSQL 17
• Meilisearch running locally (default http://localhost:7700)

Setup

cp .env.example .env
# edit DATABASE_URL, MEILISEARCH_URL, MEILISEARCH_KEY

# Run migrations (requires database exists)
sqlx migrate run

# Seed sample data
psql -d lbd -f scripts/seed_data.sql

# Run server
cargo run

API Smoke Test

curl http://127.0.0.1:3000/health
curl http://127.0.0.1:3000/entries/padem

Current Core Dependencies

• axum 0.7.5
• tokio 1.37.0
• sqlx 0.7.4
• meilisearch-sdk 0.27.0
• serde/serde_json 1.0.x

Runtime Check (latest deps)

• cargo run boots successfully (health endpoint OK)
• cargo check clean on updated dependencies

License

Dictionary data: Creative Commons Attribution-ShareAlike 4.0 (CC BY-SA 4.0)

Software: Licensed under Apache-2.0.

---

ᬮᭀᬦ᭄ᬢᬭ᭄ ᬩᬲ ᬩᬮᬶ — Lontar Basa Bali