SOLID FRAMEWORK 10.0.17650

Solid Framework SDK has been updated.

Feature Update: 

  • Enabled language detection for Arabic language documents. 

Bugfixes: 

  • [docx] Improved detection of bold font style on a specific document. 
  • [docx] Resolved a page count issue on macOS. 
  • [docx] Improved detection of small caps font style effect. 

Office Fidelity: 

  • [office] Improved whitespace detection between Arabic words. 
  • [docx] Improved header and footer detection targeting one-page documents with a horizontal line below body content. 
  • [docx] Improved detection of tables that overlap with body content. 
  • [docx] Improved detection of word order with right-to-left aligned text. 
  • [docx] Improved detection of Arabic characters when the text layer does not match the character glyphs. 

Security 

  • A limited number of third-party libraries have been updated to include the latest security fixes.   

SOLID FRAMEWORK 10.0.17490

Solid Framework SDK has been updated.

Improvements: 

  • Improvements to pdf rendering clipping algorithms: 
    • Added method for detecting a polygon that had degenerated into a polyline. 
    • Added methods for detecting self-intersections of polygon contours. 
    • Added method for automatic error detection in the clipping algorithm. 
    • Added method for changing the direction of a polygon. 
    • Support for winding/alternate rules has been added prior to polygon clipping. 
    • Rewrote the method for finding polygon intersections. 
    • Rewrote the method for adding the found intersections of polygons to the polygon structures. 
    • All clipping algorithm methods have been updated to operate with the same tolerance. 
    • Improved the accuracy of determining the type of vertices found near a polygon. 
    • Improved processing of polygon edges located very close to each other. 
  • Header and footer improvements specifically targeting one-page documents: 
    • Improved the exclusion of graphic lines, images and labels. 
    • Improved the exclusion of large tables, images and footnotes. 
    • Improved the exclusion of headings and titles. 
    • Improved the exclusion of images or text located close to other page content. 

Bugfixes: 

  • [pdf] Fixed an issue preventing conversion with PDF/A-1a and A-2b standards due to a specific page structure. 
  • [docx] Resolved an issue with CAD source content where vertical text around architectural details is displaced. 
  • [docx] Fixed an issue resulting in the top of characters in one line of text to be clipped. 
  • [docx] Resolved an issue causing five rows of a table to be incorrectly merged. 
  • [docx] Fixed an issue causing two columns of a table to be merged into one. 
  • [docx] Fixed an issue in the clipping engine preventing successful rendering of a specific pdf. 
  • [docx] Fixed an issue causing the last line of a right-to-left direction paragraph to have a hanging indent. 
  • [docx] Resolved an issue where right-to-left text was incorrectly left aligned. 
  • [docx] Fixed an issue preventing the rendering of leader (tabbing) characters in table of contents containing right-to-left text. 
  • [docx] Fixed incorrectly wrapped right-to-left text causing a page overflow issue.  
  • [docx] Resolved an indentation and alignment issue at list items in a right-to-left document. .  

Office Fidelity: 

  • [docx] Improved table of contents detection by optimizing sections across pages. 
  • [office] Improved GNSE detection to independently recognize glyphs and unicodes in separate stages. 
  • [office] Improved support for Arabic diacritical marks using analysis of scale and character spacing. 
  • [docx] Improved border line termination in specific table cases 
  • Improved the left margin alignment of a document. 
  • [docx] Resolved an issue causing text misplacement when viewed on Office 2016 only. 
  • [docx] Fixed a hybrid table detection issue resulting in two additional columns. 
  • [docx] Fixed an issue causing line shapes to be rendered as underlines. 
  • [pptx] Fixed an issue resulting in a block of text in a table to be incorrectly divided into six rows. 
  • [docx] Resolved an issue that caused one table to be incorrectly split into two tables. 
  • [docx] Resolved an issue causing a textbox to be divided in two parts. 
  • [docx] Improved Arabic language character unicode detection 
  • [docx] Improved alignment and indentation of content with right-to-left text direction.

SOLID FRAMEWORK 10.0.17360

Solid Framework SDK has been updated.

Feature Updates: 

  • Added the ability to set the text recovery language to any language when the corresponding Tesseract traineddata file is available. 

Bugfixes: 

  • [docx] Fixed an issue that caused an extra paragraph to be inserted after a specific graphic group. 
  • [docx] Resolved an issue preventing the detection of a first page header. 
  • [docx] Fixed an issue that caused an extra space to be inserted incorrectly affecting content layout. 
  • [docx] Resolved an issue preventing the detection of a small portion of red background color on a graphic. 
  • [docx] Fixed an issue converting invisible text as visible when OCR detection results in very few words. 
  • [docx] Fixed an issue preventing successful conversion of a file on Linux arm64 operating systems only. 
  • [docx] Fixed an issue preventing successful GNSE detection of text when Roboto font is used. 
  • [docx] Resolved an issue preventing the successful conversion of a file containing detailed images. 
  • [docx] Fixed an issue that caused header content to be shifted down one line on certain pages of a document. 
  • [docx] Resolved an issue that caused a document to be incorrectly clipped diagonally causing content loss.  
  • [docx] Fixed an issue preventing text from being detected on certain pages of a specific scanned document. 
  • [docx] Resolved an issue that caused a font style change in the empty space following an underlined descending letter. 

Office Fidelity: 

  • [docx] Improved detection of page X of Y page number format. 
  • [docx] Improved consistency of detection of multi-line headers. 
  • [docx] Improved detection of repeated table headers as body content instead of as header content. 
  • [docx] Improved the recovery of Korean text when GNSE is enabled. 
  • [docx] Improved borderless table detection. 
  • [docx] Improved detection of alternating headers supporting odd and even pages. 
  • [docx] Improved detection of alternating footers on odd and even pages. 
  • [docx] Improved table detection. 
  • [docx] Improved detection of merged cells. 
  • [docx] Improved detection of self intersecting glyph outlines when GNSE is enabled. 
  • [docx] Improved recognition of multi column layouts. 
  • [docx] Extend our character detection for non standard encoding to use tesseract OCR when required. 

Security: 

  • Security scanning of our codebase is automated as part of our compilation process.