v6.4.59 release

• Support ignoring leading articles in alphabetical title sorting (Settings → User interface). Existing libraries are automatically upgraded with system-language fallback; use "Recreate sort titles" or "Rescrape all" in Settings for maximum accuracy.
• Fix SMB share access issues (root routing, null pointer crash, Windows 11 signing handling) with smbj
• Improve SMB reliability by reconnecting when a cached session is no longer authenticated
• Refresh metadata correctly when manually scraping media
• Fix Trakt resume position conflict handling
• Store the language of localized TMDb titles
• Clean up orphaned movie collections in the database
• Show progress feedback during metadata scraping, collection scraping, NFO export, and post-import cleanup
• Store language information for localized TMDb titles
• Fix text clipping in Leanback file details row
• Fix orb description overlapping the focused orb in Leanback title view
• Show progress during post-import cleanup and NFO export
• Improve title matching to skip unnecessary scraper lookups on exact matches (results in slightly faster scrape)