A command-line tool for querying classical language lexicons and morphology. External services (Heritage, Diogenes, Whitaker's Words) are not bundled and must already be running.
# Enter the devenv shell (preferred)
devenv shell langnet-cli
# Or run a one-off command with the environment activated
devenv shell just -- cli query lat lupus --output json
# Start/restart the API server (requires external services running)
devenv shell just -- restart-server
# Check backend services
devenv shell just -- cli verify| Language | Lexicon | Morphology | Encoding Support |
|---|---|---|---|
| Latin | Diogenes (Lewis & Short) | Whitaker's Words | UTF-8 |
| Greek | Diogenes (Liddell & Scott) | Diogenes + CLTK | UTF-8, Betacode |
| Sanskrit | CDSL (Monier-Williams/AP90) | Heritage Platform | IAST, Devanagari, SLP1, Velthuis |
Services that must be installed and running locally:
- Sanskrit Heritage Platform (
localhost:48080) – Sanskrit morphology - Diogenes (
localhost:8888) – Greek/Latin lexicons - Whitaker's Words (
~/.local/bin/whitakers-words) – Latin morphology
Manually sourced or downloaded data (plan ahead before running indexers or semantic reduction):
- Perseus canonical corpora (
~/perseus):canonical-greekLitandcanonical-latinLittrees needed for CTS URN indexing and citation lookups. - Classics-Data (PHI CD-ROM) legacy corpus (
~/Classics-Data): optional gap-fill for works missing from Perseus. - Stanza resources (
~/stanza_resources/): downloaded automatically on first Stanza use; allow network or preinstall to avoid runtime stalls. - Gensim embeddings (
~/gensim-data/) used by the semantic reducer for similarity scoring - CLTK models (to
~/cltk_data/) - CDSL data (to
~/cdsl_data/)
After code changes, restart your long-running server process so Python modules reload (just restart-server to pick up changes).
- docs/GETTING_STARTED.md - Installation and first queries
- docs/DEVELOPER.md - Development setup and workflow
- docs/PEDAGOGICAL_PHILOSOPHY.md - Educational approach
- docs/OUTPUT_GUIDE.md - How to read CLI/API JSON (pedagogy-first)
- docs/technical/ - Technical reference docs
- AGENTS.md - Multi-model AI personas and workflows
This project uses multi-model AI-assisted development via OpenRouter. See AGENTS.md for specialized agent usage:
- @architect - System design and planning
- @sleuth - Debugging and root cause analysis
- @coder - Feature implementation and testing
- @artisan - Code optimization and style
- @scribe - Documentation and comments
- @auditor - Code review and security
- External services are required for most functionality; without them
langnet-cli verifyand queries will fail. - Several open issues remain: Diogenes sense extraction and CTS URN enrichment are flaky, Sanskrit canonicalization and DICO dictionary integration are not complete, CDSL outputs often contain SLP1 artifacts, and the universal schema/fuzzy search are still in design (see
docs/plans/for status). - Tests and health checks expect the services above; they have not been run in this workspace during this audit. Use
just testandjust lint-allinsidedevenv shell langnet-cliif your environment has all dependencies, and restart long-lived servers after code changes. - Planning docs live under
docs/plans/; active ideas are tracked indocs/plans/todo/and consolidated roadmap items are indocs/ROADMAP.md.