[jira] [Updated] (PDFBOX-2023) Text extraction gets zero font height for type3 fonts

["John Hewson (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/PDFBOX-2023?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

John Hewson updated PDFBOX-2023:
--------------------------------
    Fix Version/s: 2.0.0

> Text extraction gets zero font height for type3 fonts
> -----------------------------------------------------
>
>                 Key: PDFBOX-2023
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-2023
>             Project: PDFBox
>          Issue Type: Bug
>          Components: Text extraction
>    Affects Versions: 2.0.0
>            Reporter: Tilman Hausherr
>              Labels: type3
>             Fix For: 2.0.0
>
>         Attachments: PDFBOX-2023.pdf, zero_height.pdf
>
>
> Fred Andrews posted this to the user list and I can confirm that text extraction gets nothing:
> I am using PDFTextStripper on some PDF statements from Bank of America, and everything is coming through as zero height. I traced it down to getFontHeight in org.apache.pdfbox.pdmodel.font.PDSimpleFont, which is indeed getting zero.  The font is a type 3 font and I'm not sure how it should work, but getFontHeight is calling getAFM() and that is returning a null because its not a type 1 font.  Then in the next section in getFontHeight there are no font descriptors, and the zero just flows through all the way through getFontHeight. 
> I searched for anything I could key on to calculate the font height but couldn't find it.  The font size is claimed to be 20 by getFontSize(), although it appears to be more like 8. I did trace to where it got a font size command of twenty, but somehow I'm assuming that would need to be scaled, and I can't see where that might come from.
> The font width on the other hand looks accurate, and I would think something similar to that would be needed, but would really appreciate some guidance on how it should work.  If I have clue on how it should work I can see what I can do to implement it.
> This file displays fine in Acrobat and edits fine in Nitro, so it can't be that invalid.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)