feat(stats): reading statistics — session capture wired + /stats page + widget + per-book card

The design parked in the handoff (multi-tier stats off reading_history;
bookorbit-style in data, not looks) is in. The real prerequisite work
was data: reading_history existed but NOTHING wrote to it — the whole
stats feature was sitting on an empty table.

- CAPTURE (no client changes, no repo-boundary crossings): every
  successful progress write through SaveProgress (KOReader + app + web
  + Kobo + offline queue — the one chokepoint) now records one
  reading_history session row: span = the interval since the PREVIOUS
  write instant, pages = positive page delta when both sides know the
  page index, marker rows for first contact. Spans over 2h are kept as
  zero-second markers (overnight suspend→resume, offline flushes must
  not inflate totals) — the honest 'sync cadence, not a stopwatch'
  caveat stays. device_id only when a REAL device row matches (the
  web/app progress path historically stuffs the USER id into DeviceID
  — probe instead of trusting it; FK violation would silently drop
  every capture). Source preserved in device_metadata.

- API: /api/stats/summary?range=7|30|90|year (totals, zero-filled
  daily arrays, streaks, most-active day/weekday, top books, library
  breakdown — one call), /api/stats/this-week (widget: rolling 7d vs
  previous + streak), /api/stats/books/:id (lifetime + recent
  sessions). All bucketed in the user's timezone; finished ≡ any push
  ≥0.995; visible-library lens enforced at source. Legacy
  /api/analytics/* kept untouched. Range windows anchor end-of-day to
  the START of today (blasting the clock time into tomorrow resolved
  'today' a calendar day late — found by the streak test).

- /stats page (sidebar Stats between Progress and Devices, previously
  /analytics was an orphan): range selector, five stat cards, two
  Chart.js v4 charts (daily minutes bar, pages/day line — chunked via
  dynamic import so main.js stays untouched), per-library breakdown,
  most-read list, empty-state explainer. /analytics now 301s to /stats.
  The orphaned analytics.templ + analytics.ts removed (their containers
  and the page JSON never matched; the page was astub).

- Dashboard 'This Week' widget (server-rendered, suppressed on service
  failure) and the book detail 'Your Reading Stats' card (lifetime
  totals of the viewer's own sessions) complete the tiers: per-user
  default, per-library breakdown inline, per-book on the page,
  per-system/admin aggregate parked as before.

Tests: statistics_test.go (capture semantics at the chokepoint,
tz/day buckets, zero-fill, streak math, finished detection, range
validation, auth gate, widget, per-book lifetime). Full suite PASS.

internal/database/queries/queries.sql

@@ -1467,6 +1467,103 @@ SELECT
  ORDER BY read_count DESC
  LIMIT $2;
 
+-- STATISTICS (reading sessions) — time spent, pages, streaks, breakdowns.
+-- Every stats query scopes to the requesting user's VISIBLE libraries
+-- (family lens at the source) and single scan of reading_history.
+
+-- Full session rows for a range (summary math runs in Go: tz buckets,
+-- streaks, finished-book detection). The bucket/range key is the session
+-- END instant (the push time; NULL falls back to created_at) — the same
+-- value the capture stamps every write with.
+-- name: GetUserReadingSessionsRange :many
+SELECT
+    rh.id, rh.media_item_id, rh.device_id, rh.progress_percentage,
+    rh.reading_session_start, rh.reading_session_end,
+    rh.pages_read, rh.time_spent_seconds, rh.device_metadata, rh.created_at,
+    mi.title, mi.author, mi.library_id, mi.cover_image_path
+FROM reading_history rh
+JOIN media_items mi ON mi.id = rh.media_item_id
+WHERE rh.user_id = sqlc.arg('user_id')
+  AND COALESCE(rh.reading_session_end, rh.created_at) >= sqlc.arg('start_time')
+  AND COALESCE(rh.reading_session_end, rh.created_at) <= sqlc.arg('end_time')
+  AND (sqlc.narg('visible_library_ids')::uuid[] IS NULL OR mi.library_id = ANY(sqlc.narg('visible_library_ids')::uuid[]))
+ORDER BY rh.created_at ASC;
+
+-- Per-library aggregation for the same range.
+-- name: GetUserLibraryBreakdown :many
+SELECT
+    l.id::uuid AS library_id,
+    l.name,
+    COUNT(*)::int AS session_count,
+    SUM(COALESCE(rh.time_spent_seconds, 0))::int AS total_seconds,
+    SUM(COALESCE(rh.pages_read, 0))::int AS total_pages,
+    COUNT(DISTINCT rh.media_item_id)::int AS book_count
+FROM reading_history rh
+JOIN media_items mi ON mi.id = rh.media_item_id
+JOIN libraries l ON l.id = mi.library_id
+WHERE rh.user_id = sqlc.arg('user_id')
+  AND COALESCE(rh.reading_session_end, rh.created_at) >= sqlc.arg('start_time')
+  AND COALESCE(rh.reading_session_end, rh.created_at) <= sqlc.arg('end_time')
+  AND (sqlc.narg('visible_library_ids')::uuid[] IS NULL OR mi.library_id = ANY(sqlc.narg('visible_library_ids')::uuid[]))
+GROUP BY l.id, l.name
+ORDER BY total_seconds DESC;
+
+-- Most-read books for a range (minutes), for the top list.
+-- name: GetTopBooksByMinutes :many
+SELECT
+    mi.id::uuid AS media_item_id,
+    mi.title,
+    mi.author,
+    mi.library_id,
+    mi.cover_image_path,
+    SUM(COALESCE(rh.time_spent_seconds, 0))::int AS total_seconds,
+    SUM(COALESCE(rh.pages_read, 0))::int AS total_pages,
+    COUNT(*)::int AS session_count,
+    MAX(rh.created_at) AS last_read
+FROM reading_history rh
+JOIN media_items mi ON mi.id = rh.media_item_id
+WHERE rh.user_id = sqlc.arg('user_id')
+  AND COALESCE(rh.reading_session_end, rh.created_at) >= sqlc.arg('start_time')
+  AND COALESCE(rh.reading_session_end, rh.created_at) <= sqlc.arg('end_time')
+  AND (sqlc.narg('visible_library_ids')::uuid[] IS NULL OR mi.library_id = ANY(sqlc.narg('visible_library_ids')::uuid[]))
+GROUP BY mi.id, mi.title, mi.author, mi.library_id, mi.cover_image_path
+ORDER BY total_seconds DESC
+LIMIT sqlc.arg('limit');
+
+-- Lifetime session rows for ONE book (the book-detail stats card).
+-- name: GetBookReadingSessions :many
+SELECT
+    rh.id, rh.device_id, rh.progress_percentage,
+    rh.reading_session_start, rh.reading_session_end,
+    rh.pages_read, rh.time_spent_seconds, rh.device_metadata, rh.created_at
+FROM reading_history rh
+WHERE rh.user_id = sqlc.arg('user_id')
+  AND rh.media_item_id = sqlc.arg('media_item_id')
+ORDER BY rh.created_at ASC;
+
+-- Distinct ACTIVE days (any captured session) bucketed in the user's
+-- timezone — the streak calendar. Scoped to a long look-back so the
+-- "longest streak" and a current streak longer than the selected range
+-- are both real.
+-- name: GetUserReadingActiveDays :many
+SELECT DISTINCT (COALESCE(rh.reading_session_end, rh.created_at) AT TIME ZONE sqlc.arg('tz')::text)::date::text AS day
+FROM reading_history rh
+JOIN media_items mi ON mi.id = rh.media_item_id
+WHERE rh.user_id = sqlc.arg('user_id')
+  AND COALESCE(rh.reading_session_end, rh.created_at) >= sqlc.arg('start_time')
+  AND (sqlc.narg('visible_library_ids')::uuid[] IS NULL OR mi.library_id = ANY(sqlc.narg('visible_library_ids')::uuid[]))
+ORDER BY day ASC;
+
+-- Device existence probe for capture: the web/app progress paths
+-- historically stuff the user id into DeviceID (it only fed a broadcast
+-- label before stats); a capture row must only reference a REAL device.
+-- name: DeviceExistsForUser :one
+SELECT EXISTS(
+    SELECT 1 FROM devices
+    WHERE id = sqlc.arg('device_id') AND user_id = sqlc.arg('user_id')
+) AS has_device;
+
+
  -- ============================================
 -- DEVICE MANAGEMENT & AUTH
 -- ============================================