4 Commits

Include renames

Author	SHA1	Message	Date
ChuckandClaude Opus 5.5	e6e0a16140	ci: run the web UI DOM test suites; fix two stale suites (#660) ... A new "Web UI JS tests" job installs jsdom, starts the web interface in emulator mode and runs test/js/run_all.js with REQUIRE_DOM=1, which makes a DOM suite that can't run a failure rather than a silent skip. (The unit suites were already covered through pytest.) Two suites failed against main when run for real: - test_tools_sections rendered the Tools partial without LEDEscape, which base.html's app-early.js defines; it now installs it in beforeParse, and supplies two sample Starlark apps (one id with a quote) when the server has none, instead of assuming a device with apps and Pixlet. - test_store_dom assumed the live registry had at most 48 plugins; it now checks pagination whichever side of 48 it is. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>	2026-09-28 15:01:08 -04:00
ChuckandClaude Opus 5.5	f6afbdbb15	fix(web): Cache/Logs error mix-up, store errors, tab fallbacks; remove ~2.5k lines of dead JS (#639) ... * fix(web): keep Cache and Logs helpers out of each other's way Both partials declared top-level showError and escapeHtml. Their scripts run at global scope after every HTMX swap, so whichever tab was opened last owned window.showError, and a Cache failure after visiting Logs rendered into the Logs panel (and the other way round). Each script is now an IIFE; Cache still exports deleteCacheFile for its row buttons. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): make the HTMX-failure fallbacks for tab panels actually run - The "HTMX never loaded" fallback read appElement.__x.$data, which is Alpine 2. The page ships Alpine 3, so the check was always false and the Overview never loaded without HTMX. It now reads Alpine.$data(). - The Overview and WiFi panels used hx-on::htmx:response-error, which htmx expands to "htmx:htmx:response-error", an event that never fires. - loadTabContent sent requests with <body> as the source, so htmx fired its events on <body> and no panel's hx-on handler ran at all. The panel is now the source. htmx also resolves its promise on a 4xx/5xx, and the panel was stamped data-loaded anyway, leaving a skeleton that never retried; it is now stamped only when no responseError fired. loadPluginsDirect, loadOverviewDirect and loadWifiDirect are merged into one window.loadPartialDirect(id, url), which also runs the partial's inline scripts before Alpine sees the markup (as htmx-config.js does on htmx:afterSwap). The ~10 s "htmx never arrived" path in loadTabContent uses it for every tab instead of four hard-coded ones. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): store and registry failures no longer wipe the Plugin Manager showError replaced the whole #plugins-content with an error message, so one failed store search, custom-registry install or saved-repository call took the installed list, the store and every control with it, with no way back short of reloading the tab. Those failures are now error notifications. The full-panel message is kept only for a first load of the installed list that failed (nothing to show yet); a failed refresh of an already-rendered list is a notification too. showSuccess's fallback branch, which wrote the message into innerHTML unescaped, is gone: showNotification always exists. The "Please try refreshing your browser" hint tested for the text "Failed to Fetch", which no browser produces (Chrome says "Failed to fetch", Firefox "NetworkError..."), so it never appeared. It now keys on the failure itself: a TypeError from fetch(), or PluginAPI's NETWORK_ERROR wrapper around one. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): escape plugin action and install output on every path executePluginAction escaped data.message and data.output when an action failed but put data.message straight into innerHTML when it succeeded, and set the OAuth step-2 button's innerHTML from the manifest's step2_button_text. Plugin actions run plugin code, so that is plugin- or server-controlled markup in the page. Both paths now escape, and the button label is set with textContent. The same pattern sat in the install-from-GitHub-URL status lines (plugin_id, the server's message, and error.message, which can echo a repository URL) and the custom-registry load error; those are escaped too. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): file-upload widget owns the image list and schedule editor plugins_manager.js loads after the widget bundle, so its older copies of deleteUploadedFile, updateImageList, hideUploadProgress, formatDate, openImageSchedule, toggleImageScheduleEnabled, updateImageSchedule{Mode, Time,Day} and updateCheckboxGroupData replaced the widget's. They are deleted; the widget files are the only definitions. Before switching over, the two sets were diffed and fixed so nothing regresses: - The old copy labelled the schedule/delete buttons for screen readers and lazy-loaded thumbnails; the widget now does both. - The schedule button did nothing on a card rendered by plugin_config.html whenever the image id is a UUID (every upload): the template turns "-" into "_" in the editor's id, and neither JS copy did. Both now use the template's rule. - The widget's "keep the open editor open" copied the editor's innerHTML into the new list. That dropped its event listeners and showed the old values, so after the first change the editor looked live but ignored input. A schedule edit now saves to the hidden input and updates the card's summary in place without re-rendering the list; a list re-render (upload, delete) rebuilds an open editor from the data. Editor controls are routed by one delegated change listener, so there are no per-element listeners to lose. - The old deleteUploadedFile had a JSON branch that removed a #file_<id> element and skipped the re-render. No template or script renders such an element, and JSON uploads are listed through updateImageList like images, so re-rendering (the widget's behaviour) is the consistent one; the branch was not carried over. - The template always renders the summary line (".image-schedule-summary", "Always shown" when unscheduled) so an edit has a line to update. The inline-handler test evaluated plugins_manager.js's updateImageList; test_file_upload_widget.js now covers the widget's list and editor. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): delete the unused handleCredentialsUpload Its last caller went when plugin_config.html switched credential uploads to the file-upload widget's handleSingleFileSelect. Nothing in the web UI, the tests or the plugin monorepo references it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): delete dead and shadowed front-end code Nothing calls any of these (checked across web_interface/, test/ and the ledmatrix-plugins monorepo, including hx-*/x-*/onclick attributes): - app-shell.js: the Alpine methods refreshPlugins (it called a nonexistent this.searchPluginStore), loadPluginConfig, savePluginConfig, getSchemaPropertyType, escapeCssSelector, formatCommitInfo and formatDateInfo, and the top-level copies of savePluginConfig, getSchemaPropertyType, escapeCssSelector, formatCommitInfo, formatDateInfo and togglePluginFromTab. Plugin config forms save through hx-post in plugin_config.html. - window.reconnectSSE (app-shell.js); window.updateArrayTableAddButtonState (array-table.js). - toggleNestedSection, defined twice (app-shell.js and plugins_manager.js) and called from nowhere. - plugins_manager.js: the window.initializePlugins wrapper around an IIFE-local origInit that was always undefined, and __pluginDomReady, which was written but never read. - display.html's fixInvalidNumberInputs fallback: app-shell.js defines it before any partial loads. - base.html's window.loadCodeMirror and the two CodeMirror stylesheet preloads, and the .CodeMirror rules in plugins.html. The raw JSON editor is a plain textarea. Also deleted: app-shell.js definitions that a later script always replaced, so they never ran: executePluginAction (plugins_manager.js assigns its own), uninstallPlugin and its pollUninstallOperation (plugins_manager.js), and updateAllPlugins (install_manager.js). vendor/codemirror stays: test/test_web_smoke.py still requests codemirror.min.js as a sample static asset. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): call showNotification without checking it exists app-shell.js defines window.showNotification (a stand-in that queues until the notification widget loads) and base.html runs it, deferred, before every other script that notifies: app.js, the utilities, the widget bundle, plugins_manager.js, and all partials, which HTMX loads after the page. The 81 `typeof showNotification === 'function'` / `!== 'undefined'` checks, the `window.showNotification || console.log` and `|| alert` fallbacks, and their else branches (alert(), console output, and schedule.html's own hand-built toast) could never take the fallback path. They are removed, as is fonts.html's second copy of the queueing stand-in. The stand-in in app-shell.js keeps its guard (it must not replace the widget's implementation if load order ever changes), and BaseWidget's public notify()/getNotificationFunction() keep their shape for widgets that plugins ship. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): one HTML escaper, window.LEDEscape About 30 files each carried their own escapeHtml / escapeAttr / escHtml / _esc / escapeJs. They disagreed: several (notification.js, display.html's escapeHtml, operation_history.html, the app() stub) did not escape quotes, google-calendar-picker.js and tools.html's escHtml left ' alone, and some turned 0 into ''. Most were fine only because the quote-safe widget copies were preferred at runtime. window.LEDEscape now lives at the top of app-early.js, a blocking script in <head>, so it exists before any other script runs: html(v) & < > " ' as entities, null/undefined as '' attr(v) the same, for call sites that want to say "attribute" jsStringAttr(v) a JS string literal safe inside an inline handler Every former copy is now a one-line name for it (kept so call sites do not change), widgets included, with no fallback. plugins_manager.js loses its four escapeJs wrappers (callers use jsStringAttr), the duplicate escapeAttr and escapeHtml inside renderInstalledCards and renderCustomRegistryPlugins, and the window.escapeHtml / window.escapeAttribute exports, which nothing read. addArrayObjectItem's fallback markup (with a sixth hand-written escape chain) is gone too: window.renderArrayObjectItem is defined earlier in the same file, so the fallback could not run. The unused escapeHtml methods on the Alpine app (app-early.js stub and app-shell.js) are deleted. test_html_escaping.js now runs LEDEscape and every remaining name for it, and fails if a hand-rolled escaper reappears anywhere in web_interface/. Suites that evaluate slices of plugins_manager.js or widget files load LEDEscape from app-early.js through test/js/led_escape.js. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): stop htmx re-running partial scripts after every tab load htmx-config.js runs each swapped-in <script> itself on htmx:afterSwap and meant to turn htmx's own script handling off with htmx.config.allowScriptTags = false. It did that once, while setting up, but base.html loads htmx with a dynamic <script>, so htmx was not defined yet and the setting never applied. On every tab load htmx then tried to run each script again in its settle phase, found it already replaced (no parent node) and threw "Cannot read properties of null (reading 'insertBefore')" into the console, which also skipped the rest of that swap's settle tasks. The setting is now applied in the afterSwap handler, which always runs after htmx exists and before htmx settles the same swap. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): show "--" for a system stat the server could not read The stats stream and /system/status now send null for a metric they cannot read (cpu_temp off a Pi, for one) instead of 0. updateSystemStats built the header and Overview text as value + unit, so a null showed as "null°C". CPU, memory and temperature, in the header and on the Overview, now render "--" plus the unit for null or a missing field -- the same placeholder the page starts with, and what tools.html already shows. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): one Alpine accessor and one plugin-list signal window.getApp() (app-early.js) returns the root <body x-data="app()"> component through Alpine's public Alpine.$data, or null before Alpine has initialised it. It replaces the private el._x_dataStack[0] reads in app.js, app-early.js, app-shell.js, settings-search.js, overview.html and plugins_manager.js, the three local getAppComponent/appData/getAppData copies, and the Alpine 2 el.__x.$data fallbacks, which Alpine 3 never provides. Publishing the installed-plugin list: one load set window.installedPlugins and dispatched pluginsUpdated twice (loadInstalledPlugins, then renderInstalledPlugins), then wrote into the Alpine component through _x_dataStack[0] and called its updatePluginTabs() directly, and app-early.js's global listener set window.installedPlugins a third time and called updatePluginTabs() again. Now renderInstalledPlugins is the one publisher: it sets window.installedPlugins and dispatches pluginsUpdated once, and the full app()'s listener (app-shell.js) is the receiver. The app-early.js listener only builds the tab row while the app is not the full implementation yet. The "grid not loaded yet" case is a normal state (Plugin Manager tab not opened), so it logs through pluginLog instead of console.warn. updatePluginTabs had a "Debounce" comment and clearTimeout over a timer nothing ever set, and two identical branches; it now just calls _doUpdatePluginTabs (app-early.js detects the full implementation by that name in its source, which the new comment says). app()'s unused baseComponent lookup is removed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): reload the plugin list after installs and failed toggles Several callers refreshed the installed list with if (typeof loadInstalledPlugins === 'function') loadInstalledPlugins(); else if (typeof window.loadInstalledPlugins === 'function') ... but loadInstalledPlugins is local to the plugin-manager IIFE and window.loadInstalledPlugins is never defined, so from outside that IIFE both tests were false and nothing reloaded: - A failed plugin toggle left the switch drawn in the new state while the data said the old one. It now re-renders from the reverted data. The optimistic in-place edit also has to forget the grid's last-rendered markup, or setGridHtmlIfChanged sees identical HTML and skips the revert. A successful toggle still keeps the switch (and focus) as drawn. - Installing from a GitHub URL (the early handleGitHubPluginInstall), installing or uploading a Starlark app, and toggling a Starlark app on its config tab never refreshed the list, so the new app had no tab or Installed badge until the page was reloaded. They now force a reload through window.pluginManager.loadInstalledPlugins(true), and the Starlark grid redraws when that finishes instead of after a fixed 500 ms. - The Starlark uninstall inside the IIFE reloaded from the 3 s cache, which could still hold the app; it now forces a reload. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): route debug output through debugLog base.html defines window.debugLog, gated on localStorage.pluginDebug. plugins_manager.js read the same key twice more into its own flags (_PLUGIN_DEBUG_EARLY, and PLUGIN_DEBUG behind a pluginLog() wrapper), and api_client.js's RequestThrottler had a separate `debug` property with a setDebug() that nothing called. All of it now goes through debugLog. The "functions defined" dumps with their ✓ lines, and two per-plugin "enabled=" loops that ran on every render, are dropped; "[PLUGINS STUB]" labels on code that has not been a stub for a long time read "[PLUGINS]". Ungated console.log calls that announced normal events on every page load or action (settings search and tooltips registering, every toast repeated to the console, the schedule pickers initialising, widget registry unregister/clear) go through debugLog too. What remains on console.log is the widget registry's on-demand LEDMatrixWidgets.debug() dump and BaseWidget.notify's no-notifier fallback. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): drop waits and guards that could never fire - handlePluginAction polled up to 10 x 50 ms for window.togglePlugin, configurePlugin, updatePlugin and uninstallPlugin before calling them. All four are defined when the scripts load, before any card can be clicked, so the poll always succeeded at once; it now calls them. The long thinking-aloud comment over the toggle state is replaced by two lines on why the stored state, not the checkbox, decides. - initializePlugins checked typeof on setupGitHubInstallHandlers and applyStoreFiltersAndSort, function declarations in the same IIFE, and wrapped window.checkGitHubAuthStatus(), which returns a promise with its own .catch, in try/catch. - searchPluginStore wrapped each "#store-count" update (a getElementById and an innerHTML assignment) in try/catch four times; one setStoreCount() helper does it. The store's post-render re-attach of the GitHub token handler dropped its try/catch and existence checks for the same reason. - The load-time fallback outside the IIFE tested typeof initializePluginPageWhenReady, which is IIFE-local and so always undefined there; it calls window.initPluginsPage directly. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): delete two unused plugin-manager helpers stopOnDemand (IIFE-local; the page's stop button calls window.stopOnDemand from app-shell.js) and debounce had no callers. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): document the plugin-config handlers templates call validatePluginConfigForm, handleConfigSave, handleToggleResponse, handlePluginUpdate and refreshPluginConfig each get a JSDoc naming the attribute in partials/plugin_config.html that calls it and what the return value means (only validatePluginConfigForm's matters: false cancels the submit). - The `if (!window.__pluginConfigHandlersInitialized)` wrapper is gone: app-shell.js runs once per page, so it was never false. The block is dedented one level; `git diff -w` shows the real change. - The three handlers read xhr.responseJSON first. XMLHttpRequest has no such property (it is jQuery's), so that branch never ran; one xhrJson(xhr) helper parses responseText for all of them, with the same fallbacks as before. - runPluginOnDemand and stopOnDemand checked that plugins_manager.js's openOnDemandModal/requestOnDemandStop exist; plugins_manager.js is on every page, so they call them. - fixInvalidNumberInputs had a stray "Notification helper function" comment on top of its own; a leftover "section toggle ... duplicate definition removed" note is gone. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): one toast per save, and a failed durations save says so app.js's global htmx:afterRequest listener showed the server's message for every htmx request, and every form and button that posts through htmx (plugin config save/toggle/update, Display, Durations, General, Schedule, Dim schedule, the Overview actions) also reports its own result from hx-on after-request. Each save showed two toasts. The global listener now stays quiet for a request whose element, or its form, has its own after-request handler. That exposed the Rotation & Durations form's handler, which read xhr.responseJSON: XMLHttpRequest has no such property, so it always said "Durations saved" in green, even when the save failed (the global toast had been the only place the error showed). display.html already had a correct version (2xx only counts as saved; the server's message wins; its status may refine success but never overturn failure). That is now window.showSaveResult(xhr, savedText, failedText) in app.js, used by the Display, Durations and General forms; General's inline copy of the same logic is gone. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * docs(web): file headers and comments that say what the code does now - plugins_manager.js, app-shell.js, app.js and app-early.js open with a header: what the file owns, how base.html loads it and in what order relative to the others, and the globals it defines. app-early.js's app() stub also says why it exists and that, with app-shell.js now loaded before Alpine, it does not run in practice. - base.html's note on plugins_manager.js said it must load last to win over same-named functions in app.js/app-shell.js; there are none left, so it now gives the real reason (it uses everything loaded before it). - Change-narration and "already defined at the top, no need to redefine" notes are gone or rewritten as present-tense reasons; comments that were wrong are fixed ("Toggle password visibility" over the function that opens the token panel, "Insert before the closing </nav>" over an appendChild, "(from v2)", the export note that still listed escapeHtml). About forty comments that restated the line below them are removed, and a second window.currentPluginConfig = null outside the IIFE is dropped (the IIFE sets it). - The file-upload, checkbox-group and custom-feeds widgets' render() stubs say plainly that the widget is rendered server-side, instead of "for now" / "placeholder for future client-side rendering". test_plugin_action_delegation.js sliced the source up to one of the removed notes; it now ends the slice at the next section header. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): keep the escapeHtml/escapeAttribute globals for plugin pages 6da77363 removed window.escapeHtml and window.escapeAttribute because nothing in core or the plugin monorepo read them. Plugin web UIs served through serve_plugin_web_ui and third-party plugin pages may still call them, so they come back as aliases of window.LEDEscape.html and .attr, defined in app-early.js before any other script runs. test_html_escaping.js checks the aliases exist. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * docs(changelog): web-frontend Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): encode the image thumbnail path; match script tags case-insensitively CodeQL flagged the upload widget building an <img> src from a stored path, and the escaper test extracting inline scripts with a case-sensitive regex. Each path segment is now URL-encoded (still a same-origin path, and correct for names with spaces or * fix(web): clear Codacy findings in the escaper, app shell and upload widget - LEDEscape looks entities up in a Map instead of indexing an object. - showNotification is declared as a global for app-shell.js. - openImageSchedule checks the index is a non-negative integer and reads the image with Array.prototype.at. - The schedule editor calls escapeHtml directly and documents why its innerHTML template is safe: every value is escaped or constrained. The remaining rule hits are suppressed on that line with the reason. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(web): build the image schedule editor with DOM calls Codacy does not honour inline suppressions, and the editor's innerHTML template kept tripping its XSS rules even though every value was escaped. The editor is now built with a small element helper (createElement and setAttribute), so no value is ever parsed as HTML, and the file's own escapeHtml goes away. Also for Codacy: - LEDEscape.attr is its own function rather than a second name for html. - The tab loader records a failed load on the panel (data-load-failed) from a named handler, instead of a closure over a local flag. The fake DOM in test_file_upload_widget.js gains append/replaceChildren, its hostile-id check now asserts the id arrives as attribute data with no innerHTML anywhere in the editor, and test_html_escaping.js drops the file-upload.js escaper it no longer has. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * refactor(web): schedule editor helpers as plain functions Codacy's lint flags arrow functions held in local constants and a forEach callback that returns a value. The editor's pieces are now named function declarations (displayStyle, scheduleModeOption, scheduleRangeTime, scheduleDayTime, scheduleDayRow) taking what they need as arguments, and the element helper loops with for...of. htmx is declared as a global in app-shell.js. Output is unchanged; test_file_upload_widget.js passes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>	2026-09-24 17:36:53 -04:00
ChuckandClaude Opus 5	5137e86d16	feat(tools): MQTT bridge and Pixlet editor, ported onto the api_v3 split (#554) ... * feat(tools): manage the MQTT bridge and Pixlet editor from the Tools tab PR #544's change, ported onto the api_v3 package split (#553). Identical behaviour; only the placement of the new code differs. The original added 508 lines to web_interface/blueprints/api_v3.py, which #553 deletes, so every hunk of it would conflict irreconcilably. Ported by AST: 26 new top-level items sorted to where the split puts each kind -- __init__.py 2 imports, 11 constants, 7 helpers starlark.py 4 routes (/starlark/editor/{apps,status,start,stop}) misc.py 2 routes (/integrations/mqtt-bridge{,/config}) Everything outside api_v3.py -- the Tools partial, the installer scripts, the JS tests -- applied unchanged. Routes: 111 from the split plus these 6 = 117, and the url-map snapshot is regenerated to match, which is exactly what test_api_v3_url_map.py is designed to make you do when routes are added. Full Python suite: 4,278 passed, 68 skipped, 0 failed. The JS tests this PR ships could not be run here -- node is not installed on this machine -- so test/js/dom/test_tools_sections.js is unverified. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9 * fix(starlark): don't crash the pixlet editor's start/stop routes, and honor an operator-set PIXLET_EDITOR_HOST The AST-based port of #544 onto the api_v3 package split dropped `time` from starlark.py's import list. start_pixlet_editor() and stop_pixlet_editor() both call time.time()/time.sleep() directly, so every start (NameError building `state['started_at']`) and every stop that has to wait out the EXIT trap crashed with a 500. No test caught it because the route's own tests mock subprocess.Popen but never actually invoked it before now. Also carries over #544's later fix that this port branched before: env['PIXLET_EDITOR_HOST'] = '0.0.0.0' unconditionally overrode an operator who had already pinned PIXLET_EDITOR_HOST to loopback, forcing the unauthenticated `pixlet serve` process onto the LAN regardless (CodeQL CWE-1188). Switched to env.setdefault(...), same as api_v3.starlark.py's siblings already do for _pkg-owned names. Both fixes route the shared _pkg.time reference the rest of the package's route modules already use for anything a test might need to patch, rather than a bare `import time` local to this file. Ported the existing regression test from #544 (TestPixletEditorHostDefaultsButDoesNotOverride) onto this branch's module layout (web_interface.blueprints.api_v3.starlark instead of the old monolithic api_v3 module), which is what caught the NameError. Full suite: 4330 passed, 62 skipped, 2 failed -- identical on this branch and on origin/main (missing tzdata package breaks two timezone-alias tests in test_onboarding_checklist.py, unrelated to this change). Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(api-v3): clear the six lint errors this rebase introduced All six were introduced by rebasing this branch onto the merged blueprint split, not by the split itself. Confirmed by diffing pyflakes output against main with line numbers normalised -- everything else it reports is present on main too and is the package's deliberate re-export pattern. starlark.py used _STARLARK_APPS_DIR three times without importing it (F821). The rebase resolved an import-list conflict as a union of both sides, and that symbol was on neither side of the conflict hunk, so it was silently lost. It is defined in __init__.py and is now imported like its neighbours. This was the only one of the six that would fail at runtime rather than merely lint. __init__.py imported contextlib twice (F811): the cherry-pick added one next to the existing import. Removed the duplicate; the original at line 19 is used. __init__.py imported signal purely to re-export it to starlark.py, so pyflakes saw it as unused (F401). signal is stdlib and does not need routing through the blueprint package, so starlark.py imports it directly and __init__.py no longer does. contextlib stays re-exported because this module genuinely uses it. _read_mqtt_bridge_config()'s local `config` shadowed the `config` submodule this module imports at the bottom for its route side effects (F811). Renamed to `settings`, with a comment saying why, since the name is otherwise the obvious one to reach for. Verified: pyflakes now reports nothing on this branch that main does not, the package imports, all nine route modules load, and 117 routes register, matching the pinned URL-map snapshot. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(api-v3): reject MQTT bridge bodies the endpoint cannot apply Two CodeRabbit findings on the bridge settings endpoint, both of which returned 200 while doing something other than what the caller asked. `request.get_json(silent=True) or {}` turned a missing or unparseable body -- and the JSON literals null, [] and false -- into an empty dict, which then satisfied the isinstance(data, dict) guard on the very next line. The guard was there to reject exactly those bodies. Dropping the `or {}` lets None fail it. The same `or {}` on /errors/clear is left alone: its docstring documents the body as optional, so an absent body legitimately means "use the defaults". The difference is that saving settings has nothing sensible to do with no body. `if data.get('clear_password'):` accepted any truthy value, and the string "false" is truthy in Python -- so a client echoing the field back as a string wiped a password it meant to keep. Now coerced through the package's existing _coerce_to_bool, which already maps 'true'/'on'/'1'/'yes' and nothing else. test_mqtt_bridge_config_endpoint.py covers both: five unusable body shapes plus a missing body, and clear_password across truthy and falsy spellings. Verified against the unfixed code -- reverting the body guard fails 5, reverting the coercion fails 3. Not changed here: CodeRabbit also asks this endpoint to reject MQTT credentials when TLS is off (CWE-319). That is a policy decision about the feature rather than a defect -- unencrypted MQTT on a trusted LAN is common and often deliberate -- so it is raised on the PR for a maintainer call instead. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: work through the remaining review findings on the editor and bridge allow_insecure_mqtt (CWE-319, requested): a password with TLS disabled crosses the network in cleartext. Refused now rather than merely warned about -- but refused, not forbidden, because unencrypted MQTT on a trusted LAN is a normal deliberate setup. allow_insecure_mqtt is the explicit acknowledgement, defaults false, and is coerced like the other booleans so the string "false" cannot switch the guard off. starlark.py:796 -- the supported service runs Flask threaded, so two start requests could each see running=False, each launch an editor, and the second state write replace the first PID, orphaning a process that holds the display down with nothing recording it. The check-launch-write sequence now takes a module-level lock. starlark.py:848 -- if the state write failed the route returned success with an editor running and no PID recorded: status and stop both reported no session while the display stayed down until the timeout expired. It now terminates the process group and returns an error. starlark.py:890 -- SIGKILL gives the script's EXIT trap no chance to run, so nothing hands the display back, yet the response said "the display is restarting". After an escalation the display is now restarted explicitly, and a failure to do so returns an error naming the manual step instead of a success. pixlet_config_editor.sh:184 -- find_pixlet supports Darwin but macOS ships no timeout(1); GNU coreutils installs it as gtimeout. Resolved up front so the failure lands before the display is stopped rather than after. pixlet_config_editor.sh:154 -- wildcard, loopback and an explicit interface address are three cases, not two. Collapsing the last two printed a URL saying "localhost" whenever PIXLET_EDITOR_HOST named a LAN address. tools.html:1254 -- escHtml does not encode single quotes, and the app id was interpolated into an inline onclick="startPixletEditor('...')", so a directory containing an apostrophe could break out of the JS string and run script. The handler binds with addEventListener and reads the id from dataset, where it is only ever parsed as an HTML attribute. Tests: test_mqtt_bridge_config_endpoint.py grows to 23 cases covering the opt-in in both directions. The tools DOM suite gains three guards asserting the edit buttons carry no inline onclick and pass the id via dataset -- those need jsdom and did not run here, so CI verifies them. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(api-v3): log the traceback on the editor state-write failure The 848 fix answers 500 when the session state cannot be written, and logged that at error level -- but without exc_info, so the traceback never reached the log. test_web_error_detail.py guards exactly this: a handler returning 5xx must write an error-level record *with* the traceback and return the sanitized detail, because checking that merely something was logged is too weak. Caught by Core unit tests on the previous commit, not locally: the guard parses every module under web_interface/blueprints/api_v3 as one source, so it only fires once the whole package is read together. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-11 11:32:12 -04:00
ChuckandClaude Opus 5	4423ec33d5	feat(plugins): search, filter and sort for Installed Plugins, on a shared ListFilter helper (#540) ... * feat(plugins): add search, filter and sort to Installed Plugins, on a shared helper The Installed Plugins grid had no way to narrow it down: no search, no way to see only what's enabled, disabled, or out of date. On a rig with a couple dozen plugins that means scrolling the whole grid to find one. The two sections below it already solved this, twice, independently — the Plugin Store and Starlark Apps carried a copy-paste fork of the same ~600 lines (filter state, apply-filters-and-sort, page renderer, pagination strip, active-filter badge, listener wiring). Rather than add a third copy, this extracts the shared machinery and builds the new toolbar on it. New: web_interface/static/v3/js/plugins/list_filter.js — ListFilter.create() owns debounced search, filter axes, sort, the active-filter count, Clear, and optional pagination/persistence. Callers keep their own card markup via a `render` callback. Three control types cover every axis the page uses: pills (new), select (store category, starlark author) and cycle (the tri-state All -> Installed -> Not Installed button). Installed Plugins gets a compact toolbar: search box, one-click All / Enabled / Disabled / Updates pills, and a sort dropdown (A-Z, Z-A, updates first, recently updated, category). Filters reset on load, so you never come back to a mysteriously short list. No new CSS — this is the first consumer of the .filter-pill rules already sitting unused in app.css. renderInstalledPlugins() is split so it still publishes canonical state while renderInstalledCards() draws only the visible subset; the filtered list is never assigned to window.installedPlugins, which the toggle handler, isStorePluginInstalled(), runUpdateAllPlugins() and the Alpine config tabs all read as their source of truth. Toggling a plugin while filtered pins its card so it doesn't vanish from under the cursor. The Store and Starlark migrations are behaviour-preserving: same element ids, same localStorage keys (storeSort/storePerPage, starlarkSort/starlarkPerPage), same tri-state button markup, same pagination. Verified by differential tests that run the old and new implementations side by side against identical fixtures and compare every observable after each interaction. The only visible change is the pagination attribute (data-store-page/data-starlark-page -> data-list-page), which nothing outside its own click handler referenced. Net -156 lines in plugins_manager.js while adding a feature. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9 * fix(plugins): keep raw search text, and stop the store search refetching Two review findings from CodeRabbit on #540. Do not write the trimmed search value back into the input. setSearch() trimmed before storing, and syncControls() then copied that trimmed value back over what the user had typed. Pausing longer than the debounce after typing a space deleted the space (and reset the caret), making multi-word terms effectively untypable. The raw text is now kept alongside the trimmed one: filtering and activeCount() still use the trimmed value, while the input keeps exactly what was typed. Remove the legacy #plugin-search / #plugin-category listeners in initializePlugins(). They bound searchPluginStore as the event handler, so the DOM event arrived as its `fetchCommitInfo` argument — always truthy, which skipped the cached-filter fast path and refetched /api/v3/plugins/store/list with commit info on every keystroke burst and category change. The store's ListFilter controller already filters the cached list, which is what those two controls should do. This double-binding predates this PR (the old code guarded with _listenerSetup and _storeFilterInit, two different flags, so both sets stayed live); it is fixed here because the refactor owns that wiring now. Both fixes are covered by tests that fail without them: the trailing-space regressions in the installed-plugins DOM suite, and a new whole-file jsdom test that counts fetches while typing (1 request at init, 0 thereafter; previously 1 -> 2 -> 3 -> 5). Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9 * fix(plugins): build pagination via DOM APIs, drop computed member access Addresses the five Codacy security findings, all in list_filter.js. Pagination no longer assembles an HTML string (3 findings: 2 critical + 1 high, "unsafe assignment to innerHTML"). The interpolated values were only page integers and local class constants, so there was no injection path, but concatenating markup into innerHTML is the pattern the scanners flag and createElement is no less clear. Each button now also owns its click listener directly instead of the container being re-queried afterwards, and the strip is cleared with textContent = '' rather than by assigning empty markup. No innerHTML assignment remains in the file. haystack() now walks Object.entries(item) and keeps the configured fields, instead of reading item[field] per field ("generic object injection sink"). Field order no longer drives the haystack order, which is irrelevant to the substring test. matches() iterates controls with for...of instead of an index ("variable assigned to object injection sink"). The rendered pagination is unchanged: same buttons, labels, page numbers, disabled states and classes. The old-vs-new differential tests now compare pagination structurally (tag, text, page, disabled, sorted class list) rather than as an HTML string, since building nodes legitimately serialises differently — «/» as characters rather than &laquo;/&raquo;, disabled="" rather than a bare attribute. That comparison is stronger than the string one it replaces, and the real-DOM suite still drives the actual page buttons. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9 * fix(plugins): keep configured field order when building the search haystack The previous commit swapped item[field] for Object.entries(item) to clear a static-analysis object-injection warning, and in doing so changed the order of the haystack: entries follow the object's own key insertion order, not the configured `fields` order. Since the values are concatenated, that order decides which values end up adjacent, so a multi-word query spanning a field boundary matched differently. For store fields [name, description, author, id, ...] and API objects keyed {id, name, description, author, ...}, "bob plugin-01" matched before and stopped matching after. That contradicted the behaviour-preservation claim for the store and starlark migrations, and the differential tests missed it because every fixture query was a single word. Values now come out of a Map built from Object.entries, iterated in `fields` order: the original haystack is restored, and there is still no computed member access for the analyser to flag. Regression coverage for the ordering itself, at both levels: - unit: phrases spanning name->id and category->tags, plus the reverse (object-key) order asserted NOT to match - differential: the same class of query compared old-vs-new, with a guard that the phrase actually matches something so a mutual zero-result cannot pass vacuously Verified both fail without this fix (3 unit, 2 differential) and pass with it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9 * test(web): add JS suites for ListFilter and the plugin-manager grids No JS toolchain exists in this repo, so these are plain node scripts with no framework: each prints ok/FAIL lines and exits non-zero. `node test/js/run_all.js` runs everything, skipping the DOM suites (rather than failing) when jsdom is absent or nothing is listening, so it stays useful in a bare checkout. unit/test_list_filter.js ListFilter search/filter/sort/count/sticky, driven through the installed-plugins config eval'd verbatim out of plugins_manager.js so the test cannot drift from the real configuration unit/test_render_cards.js renderInstalledCards markup, both empty states, and escaping of hostile plugin metadata dom/test_installed_dom.js the toolbar in a real DOM, including the HTMX partial re-swap and a getComputedStyle check that .filter-pill[data-active] matches what we emit dom/test_store_dom.js store pagination, per-page, category, tri-state Installed button, persistence across a re-boot dom/test_no_double_fetch.js loads the whole plugins_manager.js and counts requests, so a keystroke cannot refetch the store The DOM suites deliberately fetch the partial and the plugin data from a running web interface instead of using fixtures, so a renamed element id or a changed payload shape fails them loudly. Point them at a rig with a full plugin set when it matters (BASE=http://host:5000); a dev box with two plugins installed passes while exercising very little. Several assertions exist to stop specific bugs recurring: trailing spaces surviving the search debounce, a query spanning two adjacent search fields (haystack field order is load-bearing), and window.installedPlugins staying at full length while the grid is filtered. Others guard against passing vacuously — counting only non-skeleton cards, and checking a search phrase matches something before comparing two result sets. The old-vs-new differential suites that verified the store and starlark migrations are not included: they compared against the pre-refactor code, which now exists only in git history. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014RRtqXDCnvnY6EQwhT5CV9 --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-08 20:04:47 -04:00