Hyland.DocumentFilters.Linux-arm64 25.4.0

Learn more about Target Frameworks and .NET Standard.

• • Hyland.DocumentFilters (>= 25.4.0)

# Enhancements

-     - Margin values can be specified in pixels, points, or as percentages of the page dimensions. (DF-2621)
-     - This ensures cleaner extracted text for AI workflows, data curation, and search indexing. (DF-2621)
-     - Use a combined shorthand option (`PDF_EXCLUSION_MARGINS={value}`) or individual options (`PDF_EXCLUSION_MARGIN_TOP`, `PDF_EXCLUSION_MARGIN_BOTTOM`, `PDF_EXCLUSION_MARGIN_LEFT`, `PDF_EXCLUSION_MARGIN_RIGHT`). (DF-2621)
- - Introduced support for specifying top, bottom, left, and right exclusion margins on each page of PDF input, to omit headers, footers, or side notes from extracted text and Markdown output — without affecting HiDef rendering. (DF-2621)
- Introduced option `PDF_RENDER_MODE` to specify whether PDF output should be rendered for on-screen display or printing. This affects certain rendering behaviors, including how annotations are drawn. (DF-2622)
- Introduced support for OCR of embedded images in Excel (XLS, XLSX, and CELL), PDF, and XFA PDF input files using the `OCR_INLINE_IMAGES` option. Extracted text appears as searchable, accurately positioned content in HiDef output and as inline content in Markdown output, providing enhanced searchability and text extraction without impacting the visual fidelity of documents. (DF-2560)

# Updates

- HD Mode: Resolved a condition for DOC and DOCX files in HD mode where vanished paragraphs could cause incorrect list numbering during PDF conversion. This ensures paragraph numbering in the output matches the original Word document, improving accuracy for legal and structured documents. (DF-2589)
- HD Mode: Resolved a condition for DOC and DOCX files where floating tables with pinned header rows could repeat previous row when splitting large rows across pages. (DF-2293)
- HD Mode: Resolved a condition where DOCX files containing bidirectional text could be missing text content. (DF-2347)
- HD Mode: Resolved a condition where charts with logarithmic scale axes were rendered incorrectly. (DF-2520)
- HD Mode: Resolved a condition where conversion of a non-PDF input file into PDF output fails when the `PDF_PRESERVE_ORIGINAL` option is enabled. (DF-2399)
- Resolved a condition for DOC files where only the first bullet in each list was retained when converting to Markdown. This ensures all list items are correctly interpreted, improving accuracy and completeness in Markdown exports. (DF-2609)
- Resolved a condition for DOCX files where a segmentation fault could occur during conversion to PDF when processing certain table content. This stability improvement prevents crashes during automated document transformations, ensuring more reliable handling of complex DOCX structures. (DF-2667)
- Resolved a condition for EML files where calendar decline messages produced incorrect output, displaying VCAL metadata instead of the expected footer text. This ensures that converted EML files now retain footer content consistent with how Outlook displays these messages, improving fidelity for end users. (DF-2634)
- Resolved a condition for MS Word 6 files where files created on MacOS may extract or render extended ASCII characters incorrectly. (DF-2532)
- Resolved a condition for PDF files where some tables with full borders may not be extracted as tables. (DF-2620)
- Resolved a condition for PPTX files where a corrupted file could cause a memory leak. (DF-2668)
- Resolved a condition for the Java API where Popup annotations could have infinitely nested child Popup annotations. This prevents excessive memory usage and ensures annotation structures are correctly terminated, improving stability and reliability for downstream processing. (DF-2583)
- Resolved a condition for the Java API where calling getDefaultAppearance on an annotation could cause a SIGSEGV. (DF-2580)
- Resolved a condition for the Java API where setting and then retrieving an action on an annotation object would not return the saved action. (DF-2582)
- Security: 7-zip: CVE-2025-53816: not-affected: The vulnerable code is part of the full 7-Zip application, not the LZMA SDK, and is not compiled into Document Filters. (DF-2572)
- Security: 7-zip: CVE-2025-53817: not-affected: The vulnerable code is not compiled into Document Filters. (DF-2571)
- Security: Protobuf: CVE-2025-10220: CVE affects .NET NuGet packages (Google.Protobuf, DynamicData, etc.) used in AxxonSoft Axxon One VMS. Document Filters uses C++ protobuf library v3.0.0 from source, not .NET NuGet packages. None of the vulnerable .NET packages are present in Document Filters codebase. (DF-2607)
- Security: libtiff: CVE-2025-9165: not-affected: The vulnerable code is in tiff CLI tool tiffcmp, and is not compiled into Document Filters. (DF-2573)
- Security: libtiff: CVE-2025-9900: patched: Applied security patch to address this issue. (DF-2638)
- Security: xpdf: CVE-2025-11896: patched: Applied security patch to address this issue. (DF-2653)