Livtet Core

The shared Rust workspace for the Livtet personal book-collection manager. This is the engine: the database layer, business logic, cover subsystem, search index, plugin runtime, sync engine, and CLI used by the desktop and mobile apps.

Everything in this repo is published under the Mozilla Public License 2.0.

The rest of Livtet (the desktop, Android and iOS apps) lives in sibling repositories and links against the crates built here.

What It Is

Every Livtet client embeds this workspace, directly or indirectly:

• The Tauri desktop app links the individual crates it needs.
• The Android and iOS apps reach the same surface through livtet-ffi, a uniffi-proc-macro crate exposing the library over the language boundary (Kotlin / Swift).

livtet-core is the unison crate. It composes livtet-data, livtet-covers, and livtet-search into the public surface every client depends on, and re-exports the curated names so consumers can import a single crate.

Layout

livtet-data          DB owner (sea-orm, sqlx, migrator, seed, error)
livtet-types         Leaf type definitions (DbId, URN, identifiers, format metadata)
livtet-search        Search index (Tantivy)
livtet-core          Unison crate (data + covers + search + types)
livtet-sync          Sync engine (server + client + change log)
livtet-cli           CLI binary
livtet-test-utils    Shared test helpers

The dependency graph is acyclic. livtet-data is the only crate that declares a direct sea-orm or sqlx dependency; everyone else obtains them through livtet_data::orm and livtet_data::sql. livtet-types is the one sanctioned exception — it sits below livtet-data and uses sea-orm derive paths.

livtet-types    (leaf; direct sea-orm dep)
   ^
   |
livtet-data     (DB owner)
   ^
   |
livtet-search   --> livtet-data, livtet-types
livtet-core     --> livtet-data, livtet-search, livtet-types

livtet-cli      --> livtet-core, livtet-data

livtet-test-utils (test helpers)

Crate map

Each crate has its own README.md describing its surface and responsibilities:

• livtet-data — DB owner; sea-orm + sqlx re-exports, migration runner, seed helpers.
• livtet-types — leaf; strong-typed DbId, URN, identifiers, format metadata, device types.
• livtet-search — Tantivy index, label resolver, sea-orm resource lookup glue.
• livtet-core — unison; works, editions, identifiers, plugins, sync, OPDS, backup, and the user-agent / cover fetcher used by the desktop and mobile apps.
• livtet-cli — developer CLI. Seed demo data and print canonical app-data directories.
• livtet-test-utils — shared test helpers used by integration tests across the workspace.

Conventions

• sea-orm / sqlx ownership. livtet-data is the only crate that declares a direct sea-orm or sqlx dependency. Other crates obtain them through livtet_data::orm / livtet_data::sql. livtet-types is the sanctioned exception (it sits below livtet-data and uses sea-orm derive paths).
• Public API via livtet-core. Consumers SHOULD depend on livtet-core, not on the individual sub-crates. livtet-core re-exports the curated surface (pub use livtet_data as data;, pub use livtet_covers as covers;, pub use livtet_search as search;).

Quick start

See GETTING_STARTED.md for the full setup. The short version:

mise install
cargo check --workspace --all-targets
cargo test  --workspace

Build commands

cargo check --workspace --all-targets   # verify every crate compiles
cargo test  --workspace                  # run every crate's tests
cargo clippy --workspace --all-targets -- -D warnings
cargo fmt --all                         # apply workspace formatting

License

Mozilla Public License 2.0. Copyright (c) 2026 Jacky Alcine [email protected].

See CONTRIBUTING.md for contribution guidelines (DCO sign-off, AI-assisted-by attribution, scope-prefixed commit messages).