Language detection, page orientation detection and OCR character recognition have been improved for non Latin languages. To include these improvements please update required files using traineddata.zip from Solid Framework downloads.

Next release expected on 27 May 26

Improvements: 

• [docx] Improved detection of vector text. 
• [docx] Improved paragraph detection. 
• [docx] Improved soft hyphen detection. 
• [docx] Improved detection of certain non-standard encoded characters. 
• [docx] Implemented detection of additional list items in Chinese, Japanese and Korean language documents. 
• [office] Implemented the inclusion of producer information in conversion output. 

Bugfixes: 

• [docx] Fixed a bug preventing the detection of text in a specific document. 
• [docx] Fixed a bug where the height of an empty textbox prevented the document from being opened in Microsoft Word. 
• [docx] Fixed a bug causing Type 3 text in a font bounding box to be clipped. 
• [docx] Fixed a bug causing the substitution of a symbol for a non-standard encoded character. 
• [docx] Fixed a bug causing performance issues in a document. 
• [docx] Fixed a bug causing underline style to be detected as a line shape. 
• [xlsx] Fixed a bug resulting in rows becoming merged on a document. 
• [pptx] Fixed a bug interfering with image transparency detection.