Kuki-Chin Bible Corpus & Morphological Analysis

A digital philology infrastructure for Kuki-Chin languages, featuring:

• 20 Bible corpora aligned by verse (422,676 unique wordforms)
• Tedim Chin morphological analyzer with Leipzig-style glossing
• SQLite backend for normalized dictionary/grammar data
• Bootstrap lexicons and interlinear glossing tools

Current publication priority

The current publication phase is focused on the Tedim Chin package: grammar, dictionary, and chrestomathy/reader. Scaling the same workflow to Mizo/lus and the remaining Kuki-Chin languages is a later phase, after the Tedim publication outputs are in publishable form.

Architecture

┌─────────────────────────────────────────────────────────────────┐
│                    Authoring Layer                              │
│  scripts/analyze_morphemes.py → data/ctd_analysis/*.tsv        │
└─────────────────────────────────────────────────────────────────┘
                              ↓
┌─────────────────────────────────────────────────────────────────┐
│                    Backend (SQLite)                             │
│  data/ctd_backend.db ← scripts/backend.py                      │
│  Tables: lemmas, senses, wordforms, examples, morphemes...     │
└─────────────────────────────────────────────────────────────────┘
                              ↓
┌─────────────────────────────────────────────────────────────────┐
│                    Publication Outputs                          │
│  output/dictionary/  output/grammar/  output/metrics/          │
└─────────────────────────────────────────────────────────────────┘

Quick Start

# Build the backend from TSV exports
make backend

# Run the standard test workflow
make test

# Generate backend-driven drafting outputs
make grammar-reports
make dictionary

Standard workflow commands:

• make backend - rebuild the Tedim SQLite backend
• make backend-check - verify backend counts and integrity
• make grammar-reports - generate backend-driven grammar reports, including the integration dashboard
• make dictionary - generate backend-driven dictionary sample outputs
• make test - run the analyzer test runner, rebuild backend, then run the full pytest suite

Tedim Chin: Current Metrics

Metric	Value	Notes
Corpus tokens	831,152	Full Bible text
Lemmas	7,339	All with glosses
Senses	9,962	Including polysemous items
Grammatical morphemes	485	Affixes and clitics
Corpus examples	26,898	Linked to senses/morphemes
Coverage (known POS)	100.0%	Full morphological analysis
Senses with examples	5,812	58% of senses

Regenerate metrics: make metrics → output/metrics/ctd_metrics.json

Test Coverage

Suite	Tests	Purpose
Backend tests	61	Database integrity, API behavior, end-to-end workflow
Export tests	56	TSV export correctness
Orthography tests	71	Normalization and tone restoration
Metrics tests	10	Publication readiness gates
Other analyzer tests	166	POS tagging, morphology, disambiguation
Total	364

Run tests: make test for the standard workflow, make test-backend for the backend-native pytest suite, or python3 -m pytest tests/ for direct pytest.