* refactor(plugins): split PluginStoreManager into mixins
src/plugin_system/store_manager.py (2,977 lines) keeps the class, its
shared state, locks, the uninstall registry, directory lookup and
uninstall; its methods are split by area into:
- store_registry.py (_RegistryMixin): registry, GitHub metadata, search,
manifest validation
- store_install.py (_InstallMixin): install paths and dependencies
- store_update.py (_UpdateMixin): updates, rollback, local git state
Pure move: all 56 members are byte-identical (checked with ast) and the
assembled class has exactly the same attributes as before (checked at
runtime). PluginStoreManager is imported from store_manager.py as before.
Tests that patched shared modules (subprocess, requests, tempfile, shutil)
through store_manager now reach them through the module whose code they
exercise; a source-text contract test reads all store_*.py modules.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: annotate findings the split moved into new store modules
subprocess imports and a list-form git clone (no shell), and the config
template's placeholder token string -- existing code that Codacy reported
as new because it moved. Annotated with the repo's nosec/nosemgrep style.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: annotate the default-branch git clone the split moved
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix: six bugs found testing main on a real Pi (ledpi)
- Stopping the service now runs cleanup. systemd stops ledmatrix.service
with SIGTERM, whose default action ended Python before run()'s finally
block, so the update worker, Vegas and the panel were never torn down.
main() now turns SIGTERM into KeyboardInterrupt, the Ctrl-C path.
- "Now showing" no longer turns into "unknown". display_current_state was
only written on a mode change and the web UI reads it with max_age=120,
so a live game or a single plugin on screen for longer read as unknown.
It is republished every 30 s while unchanged.
- Switching Vegas on in the web UI works when it was off at startup. The
coordinator was only created at startup; the config watcher now flags it
and the render thread creates it.
- configure_web_sudo.sh finds reboot and poweroff in /usr/sbin. Run as the
web user it could not, silently dropped their rules and still said it
granted them, so the web UI's Reboot/Shutdown stopped working.
- check_system_compatibility.sh reports installed packages as installed.
`dpkg -l | grep -q` under pipefail failed when grep exited early.
- A network failure fetching GitHub repo info logs a WARNING, not ERROR.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs(changelog): fixes found testing on a Pi
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test: run the Linux-only script tests correctly
The sbin-lookup test set PATH=/nonexistent and then could not find bash
itself; call it by absolute path. The dpkg-query stub read $4, but the
package name is the third (last) argument. Both now pass on a Pi.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(display): cover the follower, long-render and startup cases
Review follow-ups on the ledpi fixes:
- The pending Vegas start is applied before the sync-follower branch too
(_apply_pending_vegas_init), which skips _is_vegas_mode_active() while a
follower is connected but needs the coordinator for the leader's image.
- _service_pending_changes(), which runs inside Vegas iterations and long
screens, republishes a stale display_current_state as well; the main
loop alone could be away for a 240 s Vegas iteration.
- The SIGTERM handler is installed after DisplayController() is built, so a
stop during parallel plugin loading keeps the default immediate exit.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
