Added support for binary extraction of attachments from both MSG and EML file types.

Updated NuGet package dependencies to latest versions.

Document Structured Extraction Enhancements:

• New Table Support: Introduced a dedicated Table element type alongside the existing Paragraph type. This allows for structured tabular data retrieval from DOC and DOCX files.
• Human-Readable Styles: Improved DOC extraction to return actual Paragraph Style Names (e.g., "Heading 1") instead of internal style indexes.
• Page Referencing: Added a PageNumber attribute to all document elements extracted from PDFs, enabling easier navigation and source tracking.
• Enhanced PDF Segmentation: Upgraded the PDF page segmentation algorithm from "Recursive XY Cut" to "Docstrum". This change significantly improves the reliability of element detection, especially in complex layouts with tight margins or overlapping content.