[PDFBOX-894] Can`t open XREF as a stream. Latin Chars unreconized

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.3.1
• Fix Version/s: 1.3.1
• Component/s: Parsing
• Environment: Windows XP
• Assignee: Andreas Lehmkühler
• Reporter: Pedro Rodrigues
• Created: 18/Nov/10 17:27
• Updated: 22/Jul/12 14:11
• Resolved: 22/Jul/12 14:10

Description

Latin characters (e.g. á, é, í, õ, ç) just aren`t recognized in PDF, creating text files with strange characters like Œ, ? and ˆ instead of the right ones.

I tryied to investigate through the PDFBox code what could have caused this, but with my limited knowledge on PDF formating, I couldn`t advance too much.

There are 2 points which I`d like to comment:
1 - The PDF has no XREF table, using Stream XREF(e.g. 3076/Type/XRef/W[1 3 1]>>stream) instead;
2 - While encoding the byte stream (COSStream.getFilteredStream().doEncode()) the Filter encoding is FlateDecode (not sure if it`s a real issue, since in other PDFs I get this same encoding and it works fine).

I`m not sure if there`s a problem with the pdf itself, since I can open it and see the Latin Characters with no trouble at Adobe and Foxit Readers.

Attached is on of the PDFs that I`m talking about. The issue is seen from the 5th page on.