[PDFBOX-2912] PDFTextStripper: positionX < previous positionX case

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Cannot Reproduce
• Affects Version/s: 1.8.10
• Component/s: Text extraction
• Flags: Patch
• Assignee: Unassigned
• Reporter: Billow Gao
• Created: 31/Jul/15 18:02
• Updated: 06/Jan/16 17:40
• Resolved: 06/Jan/16 17:40

Description

It's possible that positionX is smaller than previous character's positionX.
Current PDFTextStripper will not add space and extract them to one word.
Added a patch to compare positionX < previous positionX case which can insert a space.

Disabled by default.

Tested with 10002 PDF files which had the negative positionX issue. The result was improved for all of them.
In my test, I set negativeStartXTolerance to 4