[jira] Updated: (PDFBOX-269) ExtractText errors

["Justin LeFebvre (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/PDFBOX-269?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Justin LeFebvre updated PDFBOX-269:
-----------------------------------

    Attachment: endstream_missing_fix.diff

The reason for this issue is that some of the stream objects in the pdf file are missing endstream tags. Because of this, pdfbox was skipping right over objects until it found the next endstream tag in the file. The objects did, however have endobj tags so our fix basically ignores the fact that they don't have endstream tags and terminates parsing those objects at their endobj tags. 

> ExtractText errors
> ------------------
>
>                 Key: PDFBOX-269
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-269
>             Project: PDFBox
>          Issue Type: Bug
>          Components: Text extraction
>            Priority: Minor
>         Attachments: endstream_missing_fix.diff
>
>
> [imported from SourceForge]
> http://sourceforge.net/tracker/index.php?group_id=78314&atid=552832&aid=1706491
> Originally submitted by wrwessel on 2007-04-24 04:31.
> Wrote a batch file to convert over 500 powerpoint files I had to pdf (using DocumentConverter.py and OpenOffice) then the batch file uses ExtractText.exe to extract the text.  Most of these files converted fine but I had 4 files where ExtractText could not get any text and threw various error messages.  I have attached one of these as a sample.  Using version 0.7.4 from 19/5/07 and same problem with 0.7.3 release.  It is easy enough for me to convert the last 4 by hand, but might be a bug you can fix.
> Many thanks for the ExtractText program, saved a long time converting these by hand.
> [attachment on SourceForge]
> http://sourceforge.net/tracker/download.php?group_id=78314&atid=552832&aid=1706491&file_id=226382
> Sample.zip (application/x-zip-compressed), 216343 bytes
> [comment on SourceForge]
> Originally sent by benlitchfield.
> Logged In: YES 
> user_id=601708
> Originator: NO
> I've looked at the attached PDF, technically I believe the root issue is that OpenOffice is not correctly writing the PDF.  I have submitted the issue with those guys and can be monitored by going to http://www.openoffice.org/issues/show_bug.cgi?id=76879
> The issue is that the PDF is sometimes missing the 'endstream' tag; which PDFBox looks for to tell it that the stream is done.
> My rule of thumb is that if Acrobat can open it, then so should PDFBox, so this is still a 'bug' with PDFBox.  Fixing this is possible but is not straightforward, so it may be a little bit before it is complete.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.