[
https://issues.apache.org/jira/browse/PDFBOX-2272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14628012#comment-14628012
]
Andreas Meier commented on PDFBOX-2272:
---------------------------------------
Sorry that I didn't answer yesterday, I got some technical problems with my
computer.
The only use of handleTextPosition is to perform text extraction for a specific
rotation.
In fact my code does something similar to the sortByPosition flag.
The difference is, that I search for TextPositions with direction other than
0°. If the code finds other directions it handles these cases, by running
handleTextPosition with rotations of 90° and 270° exclusively.
Therefore vertical text starting at the same YPosition inside the PDFDocument
will not be mixed up in a single output line, as it appears to be the case with
sortByPosition. (unfortunately I guess that my approach is even slower than the
sortByPosition flag and too much overhead for now)
I recommend to wait until I can provide a clearly structured rework of
PDFTextStripper.java, but first I need to set up my development environment
properly.
> Can't extract vertical text correctly
> -------------------------------------
>
> Key: PDFBOX-2272
> URL: https://issues.apache.org/jira/browse/PDFBOX-2272
> Project: PDFBox
> Issue Type: Bug
> Components: Text extraction
> Affects Versions: 1.8.6, 2.0.0
> Reporter: Biligsaikhan Batjargal
> Attachments: PDFTextStripper.java, test.pdf, test.txt, vertical.patch
>
>
> - -1.8.6 can't extract the Unicode due to failing to map the UCS2 CMap for
> 90ms-RKSJ-V.-
> - 2.0 extracts the text but can't handle the vertical layout
> Also see the file from PDFBOX-2294 which contains both horizontal and
> vertical text.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
