Opening the Library with a large grouped library was slow because each folder's 2x2 cover mosaic was recomposed from scratch on every paint (up to 4 MuPDF cover decodes plus scales per cell). On a 685-book library this dominated the synchronous open path (254ms of 300ms) and ran again on the post-sync refresh. Navigation felt fast only because drilling into a group shows single covers, not mosaics. Mirror cloud_covers' async pattern in group_covers: - Cache the composed master bb per group, keyed by a signature that flips when the child set or any child's cover availability changes; serve cheap copies on a hit. Cache a nil result too, so a group whose covers are not ready yet keeps its placeholder without recomposing on every refresh; a later cover download flips the signature and recomposes once. - Compose off the first-paint path: a miss enqueues a background job (one mosaic per UI tick) and returns nil so the cell paints its FakeCover placeholder immediately; finished mosaics coalesce into one refresh. - Free cached masters when the Library closes. Measured synchronous open path drops from 300ms to 151ms on a 685-book library; mosaic compositing moves off the blocking paint and fills in progressively. Adds open-path timing logs to librarywidget and localscanner for on-device diagnosis of future large-library reports. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
readest.koplugin tests
Unit tests for apps/readest.koplugin/library/ modules. Runs under LuaJIT
2.1 (the runtime KOReader uses) via busted.
Toolchain
One-time per machine:
# macOS
brew install luajit luarocks
# Linux (Debian/Ubuntu)
sudo apt-get install luajit luarocks
# Then, regardless of OS:
luarocks --lua-version=5.1 install busted
luarocks --lua-version=5.1 install lsqlite3complete
The --lua-version=5.1 flag is required: LuaJIT identifies itself as Lua 5.1, and we install rocks against that runtime so production code (which targets LuaJIT) and test code share a Lua interpreter.
Running
From the repo root:
pnpm test:lua
Or from this directory:
eval "$(luarocks --lua-version=5.1 path)"
busted --lua=$(which luajit)
Layout
spec/
├── spec_helper.lua # KOReader stubs + lua-ljsqlite3 shim (loaded once)
├── library/
│ ├── smoke_spec.lua # Sanity check that the harness boots
│ └── *_spec.lua # One per module under library/
└── README.md # This file
What spec_helper provides
require("lua-ljsqlite3/init")→ returns a SQLite shim wrappinglsqlite3complete. Exposes the subset of the lua-ljsqlite3 API our library modules use (open,exec,prepare,bind1,step,reset,clearbind,close, etc).require("logger")→ no-op logger (warn/info/dbg/errcallable).require("datastorage")→ fakeDataStorage:getSettingsDir()returning a per-testmktemp -dpath.require("device")→ stubDevice.canUseWAL() == true,Device.screenwithgetWidth/getHeight.G_reader_settings(global) → in-memoryreadSetting/saveSetting/flush.
Each spec calls require("spec_helper").reset() in before_each to wipe state.
Adding a new module
- Write production code at
apps/readest.koplugin/library/foo.lua. - Write
apps/readest.koplugin/spec/library/foo_spec.lua. - Run
pnpm test:luafrom the repo root. - Run
pnpm lint:luato syntax-check (LuaJIT bytecode compile).
Why LuaJIT and not stock Lua?
KOReader runs LuaJIT exclusively. LuaJIT extends Lua 5.1 with FFI and a few syntax tweaks; stock Lua 5.4 has features (integer division //, bit operators ~, <const> annotations) that LuaJIT rejects. Running tests under LuaJIT catches these incompatibilities at test time instead of when KOReader fails to load the plugin.