dev: umfangreiches Refactoring, Trakt-Integration und Code-Review-Fixes (0.1.69-dev)

Core & Architektur:
- Neues Verzeichnis addon/core/ mit router.py, trakt.py, metadata.py,
  gui.py, playstate.py, plugin_manager.py, updater.py
- Tests-Verzeichnis hinzugefügt (24 Tests, pytest + Coverage)

Trakt-Integration:
- OAuth Device Flow, Scrobbling, Watchlist, History, Calendar
- Upcoming Episodes, Weiterschauen (Continue Watching)
- Watched-Status in Episodenlisten
- _trakt_find_in_plugins() mit 5-Min-Cache

Serienstream-Suche:
- API-Ergebnisse werden immer mit Katalog-Cache ergänzt (serverseitiges 10-Treffer-Limit)
- Katalog-Cache wird beim Addon-Start im Daemon-Thread vorgewärmt
- Notification nach Cache-Load via xbmc.executebuiltin() (thread-sicher)

Bugfixes (Code-Review):
- Race Condition auf _TRAKT_WATCHED_CACHE: _TRAKT_WATCHED_CACHE_LOCK hinzugefügt
- GUI-Dialog aus Daemon-Thread: xbmcgui -> xbmc.executebuiltin()
- ValueError in Trakt-Watchlist-Routen abgesichert
- Token expires_at==0 Check korrigiert
- get_setting_bool() Kontrollfluss in gui.py bereinigt
- topstreamfilm_plugin: try-finally um xbmcvfs.File.close()

Cleanup:
- default.py.bak und refactor_router.py entfernt
- .gitignore: /tests/ Eintrag entfernt
- Type-Hints vereinheitlicht (Dict/List/Tuple -> dict/list/tuple)

addon/plugins/topstreamfilm_plugin.py

@@ -218,8 +218,10 @@ class TopstreamfilmPlugin(BasisPlugin):
                 if directory and not xbmcvfs.exists(directory):
                     xbmcvfs.mkdirs(directory)
                 handle = xbmcvfs.File(path, "w")
-                handle.write(payload)
-                handle.close()
+                try:
+                    handle.write(payload)
+                finally:
+                    handle.close()
             else:
                 with open(path, "w", encoding="utf-8") as handle:
                     handle.write(payload)
@@ -283,8 +285,10 @@ class TopstreamfilmPlugin(BasisPlugin):
                 if directory and not xbmcvfs.exists(directory):
                     xbmcvfs.mkdirs(directory)
                 handle = xbmcvfs.File(path, "w")
-                handle.write(payload)
-                handle.close()
+                try:
+                    handle.write(payload)
+                finally:
+                    handle.close()
             else:
                 with open(path, "w", encoding="utf-8") as handle:
                     handle.write(payload)
@@ -371,9 +375,6 @@ class TopstreamfilmPlugin(BasisPlugin):
             message=message,
         )
 
-    def capabilities(self) -> set[str]:
-        return {"genres", "popular_series"}
-
     def _popular_url(self) -> str:
         return self._absolute_url("/beliebte-filme-online.html")
 
@@ -1162,14 +1163,80 @@ class TopstreamfilmPlugin(BasisPlugin):
         return hosters.get(first_name)
 
     def resolve_stream_link(self, link: str) -> Optional[str]:
+        from plugin_helpers import resolve_via_resolveurl
+        return resolve_via_resolveurl(link, fallback_to_link=True)
+
+    def capabilities(self) -> set[str]:
+        return {"genres", "popular_series", "year_filter", "latest_titles"}
+
+    def years_available(self) -> List[str]:
+        """Liefert verfügbare Erscheinungsjahre (aktuelles Jahr bis 1980)."""
+        import datetime
+        current_year = datetime.date.today().year
+        return [str(y) for y in range(current_year, 1979, -1)]
+
+    def titles_for_year(self, year: str, page: int = 1) -> List[str]:
+        """Liefert Titel für ein bestimmtes Erscheinungsjahr.
+
+        URL-Muster: /xfsearch/{year}/ oder /xfsearch/{year}/page/{n}/
+        """
+        year = (year or "").strip()
+        if not year or not REQUESTS_AVAILABLE or BeautifulSoup is None:
+            return []
+        page = max(1, int(page or 1))
+        base = self._get_base_url()
+        if page == 1:
+            url = f"{base}/xfsearch/{year}/"
+        else:
+            url = f"{base}/xfsearch/{year}/page/{page}/"
         try:
-            from resolveurl_backend import resolve as resolve_with_resolveurl
+            soup = self._get_soup(url)
         except Exception:
-            resolve_with_resolveurl = None
-        if callable(resolve_with_resolveurl):
-            resolved = resolve_with_resolveurl(link)
-            return resolved or link
-        return link
+            return []
+        hits = self._parse_listing_titles(soup)
+        titles: List[str] = []
+        seen: set[str] = set()
+        for hit in hits:
+            if hit.title in seen:
+                continue
+            seen.add(hit.title)
+            self._title_to_url[hit.title] = hit.url
+            self._store_title_meta(hit.title, plot=hit.description, poster=hit.poster)
+            titles.append(hit.title)
+        if titles:
+            self._save_title_url_cache()
+        return titles
+
+    def latest_titles(self, page: int = 1) -> List[str]:
+        """Liefert neu hinzugefügte Filme.
+
+        URL-Muster: /neueste-filme/ oder /neueste-filme/page/{n}/
+        """
+        if not REQUESTS_AVAILABLE or BeautifulSoup is None:
+            return []
+        page = max(1, int(page or 1))
+        base = self._get_base_url()
+        if page == 1:
+            url = f"{base}/neueste-filme/"
+        else:
+            url = f"{base}/neueste-filme/page/{page}/"
+        try:
+            soup = self._get_soup(url)
+        except Exception:
+            return []
+        hits = self._parse_listing_titles(soup)
+        titles: List[str] = []
+        seen: set[str] = set()
+        for hit in hits:
+            if hit.title in seen:
+                continue
+            seen.add(hit.title)
+            self._title_to_url[hit.title] = hit.url
+            self._store_title_meta(hit.title, plot=hit.description, poster=hit.poster)
+            titles.append(hit.title)
+        if titles:
+            self._save_title_url_cache()
+        return titles
 
 
 # Alias für die automatische Plugin-Erkennung.