[jira] [Updated] (TIKA-1289) Ligatures convert on text extraction

["Alex Andrushchak (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/TIKA-1289?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Alex Andrushchak updated TIKA-1289:
-----------------------------------

    Attachment: PDF_text that can be copied is over the picture.pdf

In first sentense "Replace this file with..." fi is ligature and it is not extracted as text. I see ? mark instead of fi

> Ligatures convert on text extraction
> ------------------------------------
>
>                 Key: TIKA-1289
>                 URL: https://issues.apache.org/jira/browse/TIKA-1289
>             Project: Tika
>          Issue Type: Bug
>          Components: parser
>    Affects Versions: 1.5
>         Environment: win 8, jre 1.5
>            Reporter: Alex Andrushchak
>         Attachments: PDF_text that can be copied is over the picture.pdf
>
>
> According to tika sources review, it uses pdfbox to parse pdf files. 
> I found that pdfbox itself uses icu4j to handle ligatures.
> Unfortunately, when i added icu4j jar to my classpath nothing changed, ligatures are still not converted. Sample pdf file is attached.



--
This message was sent by Atlassian JIRA
(v6.2#6252)