[TIKA-613] PDF parser is changing letters positions

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.9
• Component/s: parser
• Labels: parser pdf

Environment

running Tika inside VB.NET 2010 with IKVM

• Assignee: Unassigned
• Reporter: Alex
• Created: 09/Mar/11 18:17
• Updated: 01/Mar/15 19:06
• Resolved: 01/Mar/15 19:06

Description

The pdf parser is changing the position of some letters and adding spaces inside the text.

For example:
Parsed text
"O fluox de caixa e os ganhos econmô icos referentes à estocagem dos RSD no aterro sanitário"

Original
"O fluxo de caixa e os ganhos econômicos referentes à estocagem dos RSD no aterro sanitário"

I`ve parsed the same text with iTextsharp and the result was Ok.

The original pdf file is here:
http://www.teses.usp.br/teses/disponiveis/8/8135/tde-04072008-113118/publico/DISSERTACAO_JOSE_EDUARDO_ABBAS.pdf

[UPDATE]

It looks like the "changing positions" is solved in the new version (1.0), but there is some "spaces" between text:

Parsed
"Os processos econômicos e polít icos causadores da atual forma de geração dos RSD"

Original
"Os processos econômicos e políticos causadores da atual forma de geração dos RSD"