Translation content is immutable once generated — add Cache-Control: public,
max-age=3600, stale-while-revalidate=86400 to handleTranslationRead so the
browser and any CDN reuse the response without hitting MinIO on repeat views.
Forward the header through the SvelteKit /api/translation/[slug]/[n] proxy
which previously stripped it by constructing a bare Content-Type-only Response.
Three bugs:
1. +page.server.ts returned an unawaited Promise — SvelteKit awaits it on
the server anyway so data.jobs arrived as a plain AIJob[] on the client,
not a Promise. The $effect calling .then() on an array silently failed,
leaving jobs=[] and the table empty. Fixed by awaiting in the load fn.
2. +page.svelte $effect updated to assign data.jobs directly (plain array)
instead of calling .then() on it.
3. handlers_textgen.go: final UpdateAIJob (payload+status write) used
r.Context() which is cancelled when the SSE client disconnects. If the
browser navigated away mid-job, results were silently dropped and the
payload stayed as the initial {pattern} stub with no results array.
Fixed by using context.Background() for the final write, matching the
pattern already used in handlers_image.go.
- Add bookstore.NotificationStore interface and wire it directly to *storage.Store
in Dependencies, bypassing the Asynq wrapper that caused Producer.(*storage.Store)
to fail with a 500 on every notification endpoint when Redis is configured
- Replace s.deps.Producer.(*storage.Store) type assertion in all 5 notification
handlers with s.deps.NotificationStore (nil-safe, always works)
- Fix PocketBase filter single-quote bug in ListNotifications, ClearAllNotifications,
and MarkAllNotificationsRead (PocketBase requires double quotes for string values)
The {#await ... then} + {@const} IIFE pattern for assigning to $state
variables stopped working reliably in Svelte 5.53+. Replaced with a
proper $effect that awaits both streamed promises and assigns to state,
which correctly triggers reactivity.
Also: switch library page selection mode entry from long-press to a
'Select' button in the page header.
Three compounding issues caused the 4+ second load:
1. getAllRatings() ran sequentially after the first Promise.all group,
adding it unnecessarily to the critical path. Now runs in parallel
with listBooks/getVotedSlugs/getSavedSlugs (all 4 concurrent).
2. discovery_votes was fetched twice on every page load — once inside
getBooksForDiscovery (via getVotedSlugs) and again by getVotedBooks.
Fixed by caching getVotedSlugs results with a 30s TTL so the second
call hits cache instead of PocketBase.
3. getVotedSlugs and getSavedSlugs were always uncached, hitting
PocketBase on every navigation. Added short-TTL per-user Valkey
cache entries (voted: 30s, saved: 60s). Cache is invalidated
immediately after each write (upsertDiscoveryVote, clearDiscoveryVotes,
undoDiscoveryVote, saveBook) so stale data is never served.
Removed the custom clear button (shown when query is non-empty) and
suppressed the browser-native webkit search cancel button via CSS.
Only the single Cancel button remains, avoiding the double/triple X
clutter on wider screens.
The SSE (non-async) chapter-names handler streamed results to the client
but never wrote them into the PocketBase job payload — only the initial
{pattern} stub was stored. The Review button then fetched the job and
found no results, showing 'No results found in this job's payload.'
Fix: accumulate allResults across batches (same as the async handler) and
write the full {pattern, slug, results:[...]} payload when marking done.
The wrapper div in +layout.svelte had pointer-events:none which blocked
all taps inside ListeningMode (chapter rows, buttons, scrolling). Removed
the wrapper div and moved the fly transition onto ListeningMode's own root
element so the slide-in animation works without stealing pointer events.
- Add notify_new_chapters_push field to AppUser, PATCH /api/profile, and profile loader
- Fix bell panel to reload notifications on every open (not just once on mount)
- Replace flat in-app + push toggles with structured category table (Category | In-app | Push)
- Add browser push master subscribe/unsubscribe row above the table
- Push column toggle disabled until browser is subscribed; shows — when unsupported/denied
- Update Notifications row hint to summarise active channels (In-app · Push / Off)
- getReadyChaptersForSlug(slug, preferredVoice) in pocketbase.ts: per-slug done jobs map, cached 60s, prefers user's voice
- GET /api/audio/chapters?slug=&voice= endpoint
- Chapter list (/books/[slug]/chapters): amber headphones icon on ready rows, play button for instant listen, banner showing ready count
- Chapter reader: audioReady + availableVoice from server load; 'Audio ready — Listen now' banner shown when audio exists and player is not yet active; sets voice preference before expanding player
- getBooksWithAudioCount() in pocketbase.ts aggregates done audio_jobs, deduplicates by chapter per slug, caches 5 min
- GET /api/audio/books endpoint
- home page: readyToListen shelf with headphones badge, chapter count, Listen button, hideable
- /listen page: full grid with search, sort (most narrated / A-Z / recent), empty state
Float mode was broken because AudioPlayer was unmounted the moment
audioStore.active became true — exactly when the float overlay needs
to render. Fix: keep AudioPlayer mounted in float and minimal modes
regardless of audioStore.active; only standard mode shows the
'Controls below' message.
Adds a third 'minimal' style: a compact single-row bar (skip ±,
play/pause, seek, time) with no voice picker or chapter browser.
Voice picker and chapter button are hidden in the idle pill too.
Settings UI updated to show all three options with a live
description of what each style does.
Up/down chevron buttons fixed to the bottom-right of the viewport.
At the top of the chapter the up button becomes a Prev chapter link;
at the bottom the down button becomes an amber Next chapter link.
Hidden in focus mode (uses its own pill). Lifts above the audio
mini-player when it is active.
- Replace bell dropdown with full-screen NotificationsModal (mirrors SearchModal pattern)
- Notifications visible to all logged-in users (not just admin)
- Admin users excluded from new-chapter fan-out (dedup vs Scrape Complete notification)
- Users with notify_new_chapters=false opted out of new-chapter in-app notifications
- Toggle in profile page to enable/disable in-app new-chapter notifications
- PATCH /api/profile endpoint to save notification preferences
- User-facing /notifications page (admin redirects to /admin/notifications)
- Service worker (src/service-worker.ts) handles push events and
notification clicks, navigating to the book page on tap
- Web app manifest (manifest.webmanifest) linked in app.html
- Profile page: push notification toggle (subscribe/unsubscribe)
using the browser Notification + PushManager API with VAPID
- API route POST/DELETE /api/push-subscription proxies to backend
- Go backend: push_subscriptions PocketBase collection storage
methods (SavePushSubscription, DeletePushSubscription,
ListPushSubscriptionsByBook) in storage/store.go
- handlers_push.go: GET vapid-public-key, POST/DELETE subscription
- webpush package: VAPID-signed sends via webpush-go, SendToBook
fans out to all users who have the book in their library
- Runner fires push to subscribers whenever ChaptersScraped > 0
after a successful book scrape
- Config: VAPID_PUBLIC_KEY, VAPID_PRIVATE_KEY, VAPID_SUBJECT env vars
- domain.ScrapeResult gets a Slug field; orchestrator populates it
Demote per-book and per-chapter-list-page Info logs to Debug — these
fire hundreds of times per catalogue run and drown out meaningful signals:
- orchestrator: RunBook starting (per book)
- metadata saved (per book)
- chapter list fetched (per book)
- scraping chapter list page N (per pagination page per book)
The 'book scrape finished' summary log (with scraped/skipped/errors
counters) remains at Info — it is the useful signal per book.
Two bugs fixed in runScrapeTask / runCatalogueTask:
1. FinishScrapeTask was called with the task's own context, which is
already cancelled when the task is stopped. The PATCH to PocketBase
failed silently, leaving all counters at their initial zero values.
Fix: use a fresh context.WithTimeout(Background, 15s) for the write.
2. BooksFound was double-counted: RunBook already sets BooksFound=1 on
success, but the accumulation loop added an extra +1 unconditionally,
reporting 2 books per successful scrape.
Fix: result.BooksFound += bookResult.BooksFound (drop the + 1).
Remove the TTS voice section from the profile page — it fetched
/api/voices on every mount, blocking paint for the full round-trip.
Voice selection lives on the chapter page where voices are already loaded.
Rewrite the server load to run avatar, sessions+stats, and reading history
all concurrently via Promise.allSettled instead of sequentially, cutting
SSR latency by ~2-3x on the profile route.
Add og:title, og:description, og:image (book cover), og:url, og:type,
og:site_name, twitter:card, twitter:image, and rel=canonical to the
book detail and chapter reader pages so link previews in Telegram,
WhatsApp, Twitter/X, Discord etc. show the cover image instead of
the site logo.
Unify the duplicated chapter picker overlays from AudioPlayer and
ListeningMode into a single ChapterPickerOverlay component.
Both callers keep their own onselect handlers; the overlay owns
search state internally and includes safe-area insets + scrollIfActive.
- Replace voice <select> with a two-column card grid grouped by engine
(Kokoro GPU / Pocket TTS CPU / Cloudflare AI); each card has a per-voice
sample play/pause button matching AudioPlayer behaviour
- Add Announce chapter and Audio mode (Stream/Generate) toggles to a
unified Playback row in Preferences; Audio mode toggle disabled for
CF AI voices
- Remove duplicate PUT /api/settings from the profile page; all writes
go directly into audioStore / theme context and the layout's single
debounced effect persists them
- Add Danger Zone section: collapsible, requires typing username to
unlock Delete account button; calls DELETE /api/profile
- Add deleteUserAccount() to pocketbase.ts: purges user_settings,
user_library, progress, comment_votes, book_ratings,
user_subscriptions, notifications, user_sessions then the
app_users record
- Add DELETE /api/profile server route (auth-guarded)
Add a user-selectable playback mode stored in user_settings:
- 'stream' (default): /api/audio-stream starts playing within seconds,
saves to MinIO concurrently — low latency
- 'generate': queue runner task, poll until full audio is ready in
MinIO, then play via presigned URL — legacy behaviour
UI toggles in two places:
- AudioPlayer idle pill: compact '· Stream / · Generate' inline next
to voice name and estimated duration
- ListeningMode controls row: pill alongside Auto, Announce, Sleep;
disabled and grayed out for CF AI voices (batch-only, no streaming)
startPlayback() now branches on audioStore.audioMode for non-CF AI
voices; generate mode uses the same runner task + progress bar flow
as CF AI but without the preview clip.
PocketBase: audio_mode text field added to user_settings on
pb.libnovel.cc (live) and in pb-init-v3.sh (create block +
add_field migration line).
- Add `archived` bool to domain.BookMeta, pbBook, and Meilisearch bookDoc
- ArchiveBook / UnarchiveBook patch the PocketBase record; ListBooks filters
archived=false so hidden books disappear from all public responses
- Meilisearch: add `archived` as a filterable attribute; Search and Catalogue
always prepend `archived = false` to exclude archived books from results
- DeleteBook permanently removes the PocketBase record, all chapters_idx rows,
MinIO chapter objects, cover image, and the Meilisearch document
- New BookAdminStore interface with ArchiveBook, UnarchiveBook, DeleteBook
- Admin HTTP endpoints: PATCH /api/admin/books/{slug}/archive|unarchive,
DELETE /api/admin/books/{slug}
- PocketBase schema: archived field added to live pb.libnovel.cc and to
pb-init-v3.sh (both create block and add_field migration)
Split chapter text into ~1000-char sentence-boundary chunks before sending
to kokoro-fastapi. Each chunk is generated individually and the raw MP3 bytes
are concatenated. This prevents the EOF / timeout failures that occur when
the server receives a very large single request (e.g. a full PDF 'Full Text'
chapter). chunkText() breaks at sentence endings (. ! ? newlines) to preserve
natural speech flow.
- parsePDF: return all text as single 'Full Text' chapter (admin splits manually)
- parseEPUB: fix chapter numbering to use sequential counter not spine index
- Remove dead code: chaptersFromBookmarks, cleanChapterText, extractChaptersFromText, chapterHeadingRE; drop pdfcpu alias and regexp imports
- Backend: POST /api/admin/books/:slug/split-chapters endpoint — splits text on '---' dividers, optional '## Title' headers, writes chapters via WriteChapter
- UI: admin panel now shows for all admin users regardless of source_url; chapter split tool shown when book has single 'Full Text' chapter, pre-fills from MinIO content
Three fixes to PDF chapter extraction quality:
1. Page ordering: parse page number from pdfcpu filename (out_Content_page_N.txt)
instead of using lexicographic sort index — fixes chapters bleeding into each
other (e.g. Prologue text appearing inside Chapter 1).
2. Windows-1252 chars: map bytes 0x91-0x9F to proper Unicode (curly quotes U+2018/
U+2019/U+201C/U+201D, em-dash U+2014, etc.) instead of raw Latin-1 control
bytes that rendered as ◆ in the browser.
3. Chapter header cleanup: skip the first page of each bookmark range (decorative
title art page) and strip any run-on title fragment at the start of the first
body page (e.g. 'for New Journeys!I stood atop...' → 'I stood atop...'). The
remaining sentence truncation is a fundamental limitation of this PDF's
PUA-encoded body font (C2_1/Literata) — those glyphs cannot be decoded without
the publisher's private ToUnicode mapping.
Use PDF outline (bookmarks) for chapter titles and page ranges, then
extract text from per-page content streams via pdfcpu ExtractContent.
This avoids the indefinite hang caused by dslipak/pdf trying to resolve
custom font ToUnicode CMaps on publisher PDFs.
- parsePDF: decrypt → ExtractContent → Bookmarks → chaptersFromBookmarks
- chaptersFromBookmarks: flatten bookmark tree, skip front/back matter
(Cover, Insert, Title Page, Copyright, Appendix), assign page ranges
- extractTextFromContentStream: handle TJ arrays (concat literal strings,
skip hex glyph arrays and kerning numbers) + single Tj strings
- Falls back to paragraph-splitting when no bookmarks present
- Build verified; test PDF produces 9 chapters with proper titles
