[PDFBOX-449] Decomposed extended Latin Characters not normalized

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.8.0-incubator
• Component/s: Text extraction
• Assignee: Unassigned
• Reporter: Justin LeFebvre
• Created: 01/Apr/09 20:01
• Updated: 04/Jun/13 17:15
• Resolved: 07/Apr/09 16:25

Description

03_2_SSL.pdf file has the unicode U+00a8 character which when extracted does not get placed over the previous character. U+0308 is required to do this. This issue applies to most diacritics found in PDF files since pdf writers use the non-combining forms of the diacritics instead of the combining forms. This issue also applies to two files located in the regression test. These are Garcia2004_thesis.pdf and cweb.pdf.

Issue Links

• is related to: PDFBOX-1622 TextNormalize init not thread-safe, may lead to infinite loop