• JDK 25.0.5.0.3 with the fix for CVE-2026-114820
• JavaFX 25.0.5
• Sejda 6.0.0.M8
• SAMBox 4.0.0.M9
• Updated dependencies
• Closed #663: lists of PDF files are now read using the system charset before falling back to ISO-8859-1, so file names with non-Latin characters load correctly
• Fixed split outputs that could lose changes, like unused fonts removed by compression, when pages share objects in the same object stream
• Fixed document metadata being lost when loading encrypted PDFs with compressed metadata
• Fixed corrupted content when decoding ASCII85 and ASCIIHex encoded streams
• Fixed content that was ignored after stray ), ] or > characters in a page content stream
• A CropBox with zero or negative width or height now falls back to the MediaBox, as other PDF readers do
• Removed the spurious Missing 'endobj' token and Xref stream ended prematurely warnings on valid documents