[PDFBOX-2800] PDFTextStripper calculates the character bounding boxes incorrectly

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Won't Fix
• Affects Version/s: 1.8.9
• Component/s: Utilities
• Labels: text_extraction
• Environment: java version "1.6.0_35"; Java(TM) SE Runtime Environment (build 1.6.0_35-b10); Java HotSpot(TM) 64-Bit Server VM (build 20.10-b01, mixed mode)
• Assignee: Unassigned
• Reporter: Evgeny Chesnokov
• Created: 13/May/15 18:32
• Updated: 13/May/15 21:23
• Resolved: 13/May/15 21:23

Description

For a specific file the extracted coordinates provided by a TextPosition stored in a charactersByArticle variable do not match the actual positions of the characters of the content. Some of the rectangles return with zero heights, and others appear shifted on a vertical axis. I am attaching the files illustrating the issue, both the sample file itself and a highlighted bounding rectangles on the 2nd page that mismatch.