Split EPG IPC orchestration, worker lifecycle, and query logic into focused services. Harden clear-worker lifecycle after review with timeout/exit handling and regression coverage.
Split the Electron external-player IPC monolith into focused launch-context, playback-request, runtime, MPV session, and VLC session modules. Includes Greptile follow-up fixes for Homebrew Cask VLC path resolution and VLC spawn-error promise handling, with regression coverage.
- merge origin/master into PR #964 and keep embedded MPV test on the isolated playback sub-entrypoint
- centralize EPG capability through DataService.supportsEpg and update PWA web-e2e expectations
- split BrowserAccessError copy between Electron and PWA diagnostics
Two Electron E2E regressions from this branch's redesigns:
1. Hiding the playlist switcher on /settings (a playlist-scoped control)
also hid the global "Add playlist" button, which lives in the same
header-actions block. A user configuring Settings before importing
their first source had no way to add a playlist — and the settings /
playlist-switcher E2E suites, which add a portal right after saving a
setting, timed out waiting for the button. The Add Playlist button is
a global action, so it now stays visible on Settings; only the
per-playlist context shortcut and bulk actions are hidden there.
2. dashboard-activation.e2e.ts asserted on `dashboard-global-favorites-
rail`, the mixed Favorites rail removed earlier in this branch. The
test now targets the v0.22 rails: live favorites resolve through the
favorites-first `dashboard-live-recent-rail`, and the played movie /
series resolve through `dashboard-continue-watching-rail` (renamed
from `dashboard-recently-watched-rail`). The movie-from-favorites
assertion was dropped — that dashboard surface no longer exists; the
favorites collection still has its own page-level coverage.
Verified locally: dashboard-activation, settings, and playlist-switcher
E2E specs all pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The redesign rollup in this branch replaced the Add Playlist dialog's
2-level type × subtype tabs with a single flat 5-card radiogroup, and
replaced the Settings theme list with a compact segmented control whose
options are just "Light"/"Dark"/"System" (no "theme" suffix). E2E tests
that pinned to the old roles/labels failed in CI:
- basic / xtream / stalker / self-hosted: getByRole('tab', ...) on
"Add via file upload" / "Xtream" / "Stalker" — there are no tabs in
the new dialog. Switched to getByRole('radio', { name: /M3U file/i })
and friends, matched against the new radio labels.
- settings: getByRole('radio', { name: 'System theme' }) — the new
picker uses just "System". Scoped to the [data-test-id="select-theme"]
radiogroup so it doesn't collide with the identically-labelled
cover-size options below.
The Electron `clickDialogSegmentedOption` helper grew a radio-role lookup
as its primary path and keeps the old tab/button/legacy-selector
fallbacks so a future redesign won't break every fixture again. The
"M3U" parent category becomes a no-op (the new picker has no standalone
M3U tile — callers always specialise to M3U URL/file/text immediately
after, which is what we want anyway).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
## Summary
- Normalize Xtream recently-added timestamps across UI, import, and dashboard query paths.
- Filter future/invalid provider dates before ranking rails and migrate legacy millisecond cache rows.
- Add regression coverage for future timestamps, series date priority, and DB migration behavior.
## Validation
- GitHub checks passed, including Unit Tests and Typechecks, Web E2E, Electron E2E on macOS/Ubuntu/Windows, CodeQL, builds, and Greptile Review.
For series in the Continue Watching rail, surface which episode the user
left off on as a small "S2·E5" chip next to the card subtitle. Sources
the season/episode straight off the matched PlaybackPositionData so the
information is always in sync with the resume target.
Fixes a latent bug in the same patch: getPlaybackPositionForItem could
not resolve series whose recent_items row carried the series id (the
landing-page path), because playback_positions are keyed by the episode
id. The lookup now matches both shapes (contentXtreamId === xtreamId OR
seriesXtreamId === xtreamId) and prefers the most recently updated
episode, which also restores the resume progress bar that had been
silently dropping out for series.
Localised in all 17 non-English locales — most use language-specific
short forms (St·F in German, T·E in Spanish/Portuguese, С·Э in
Russian/Belarusian, 시즌N N화 in Korean) rather than the English S·E.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Zoom/Filter/Search Programs buttons in the Multi-EPG header
hard-coded English matTooltip strings, so they never localized — even
when the rest of the toolbar (Close, Previous/Next day) used keys. Adds
six new EPG.* keys (ZOOM_IN, ZOOM_OUT, FILTER_CHANNELS, CLOSE_FILTER,
SEARCH_PROGRAMS, CLOSE_PROGRAM_SEARCH), wires both [matTooltip] and
[attr.aria-label] through `| translate`, and fans the strings out to
all 17 non-English locales via the i18n-fill pipeline.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Radio audio-player buttons (skip-prev, play/pause FAB, skip-next, mute,
volume slider) had zero matTooltip or aria-label. The UX audit flagged
this: "Skip-prev / Skip-next — what do they skip? In radio, is the
next 'track' the next station? Next in favourites? Random? Without a
tooltip the buttons trade silently." Same critique applied to mute
and play/pause for screen-reader users.
Add hover tooltips and screen-reader labels on all 5 controls. Skip
prev/next say "Previous station" / "Next station" so the behaviour is
explicit (they walk the active filtered list). Play/Pause flips its
label with state. Mute/unmute flips when isMuted() || volume === 0.
Volume slider gets its own label so the slider thumb is announced
correctly.
i18n: 7 new AUDIO_PLAYER.* keys translated across all 17 non-English
locales by per-locale agents; all 18 locales pass coverage and
placeholder integrity checks.
52/52 ui-playback tests still pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
`--settings-action-bar-space: 80px` added padding-bottom to the form
*after* its last child (the action bar), giving 80px of scrollable
empty space below the bar — visible when scrolling to the end as a
big dark band the page kept scrolling past the last setting.
The reservation was originally tuned for the OLD floating-chip
layout, where the bar hovered 24px above the bottom edge and content
needed to be pushed above it. In the new sticky-footer model the bar
IS the bottom edge, so no padding is needed below it.
Set to 0px. CDP verifies form.bottom === bar.bottom after the change
— the scroll now ends cleanly at the action bar.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The settings-section-scroll directive ran a 260ms Element.animate() on
whichever section became "active" as the user scrolled — three
keyframes painting an `inset 0 0 0 1px var(--settings-group-active-ring)`
ring plus a soft glow on the section block. With the new flat layout
(no card chrome, no static active ring), that pulse drew a brief 1px
border around each block as you scrolled past — the user-reported
"short border kind of highlight around blocks" effect.
The rail's left active state already announces which section you're on,
so the inline pulse is redundant. Removed the entire active-section
animation effect (the pendingScrollTarget scroll-into-view effect is
untouched — that's the click-the-nav-and-jump behaviour).
CDP verifies: all .settings-group elements report `box-shadow: none`
after scrolling. 50/50 settings tests still pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
`--settings-action-bar-offset: 24px` was tuned for the previous
floating-chip layout (border-radius:18px, margin-left:auto) so the bar
visually hovered above the bottom edge like a card. After converting
it to a full-width sticky footer the 24px gap just left a strip of
empty content peeking out below the bar — the user-reported "not
perfectly aligned to the footer" gap.
Set the offset to 0px so the bar sits flush against the bottom of the
scroll container. CDP verifies: barBottom === viewport height after
the change.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
4gray