[ 
https://issues.apache.org/jira/browse/PDFBOX-62?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13990540#comment-13990540
 ] 

Juraj Lonc commented on PDFBOX-62:
----------------------------------

I made 2 quick fixes for this (actually it is for PDFBOX-2059). Both are for 
PDTrueTypeFont. I don't know whether it is useful or you already have some 
other plans for this bug.

1. added getFontWidth() where I calculate widths from TTF font. Right now you 
are relying only on "widths" defined within PDF
2. modified getExternalFontFile2() so I am looking for system fonts too. Right 
now you are using only fonts defined in PDFBox_External_Fonts.properties

It works, but take it just for inspiration. It should be moved do PDFont (I 
guess) and make it more robust.

> Incorrect (zero) character widths returned in some docs
> -------------------------------------------------------
>
>                 Key: PDFBOX-62
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-62
>             Project: PDFBox
>          Issue Type: Bug
>          Components: Rendering, Text extraction
>            Assignee: Andreas Lehmkühler
>         Attachments: 5542.pdf, PDFBOX-2059_PDTrueTypeFont.diff, 
> PDTrueTypeFont.diff, pdfbox-2006-zerowidth.pdf-1.png, 
> pdfbox-62-zerowidth.pdf-1.png
>
>
> [imported from SourceForge]
> http://sourceforge.net/tracker/index.php?group_id=78314&atid=552832&aid=1216674
> Originally submitted by tamirhassan on 2005-06-07 13:42.
> For certain PDF documents (such as the one attached) 
> the character/string widths (as obtained e.g. by the 
> PDFont.getStringWidth method) are not returned 
> correctly, i.e. they appear to be correct for punctuation 
> characters but are zero for alphanumeric characters.  
> It seems as if these alphanumeric characters are NOT 
> within PDFont.firstChar and PDFont.lastChar in the 
> Type 1 font.  The method therefore attempts to obtain 
> the font widths from the AFM (font metric) file, but fails 
> (silently) with a 'resource is null' logline message.
> (Note that this problem doesn't seem to occur with Type 
> 1 fonts in other documents.)
> A more detailed discussion regarding this issue can be 
> found in this link:
> http://sourceforge.net/forum/forum.php?
> thread_id=1260349&forum_id=267205
> Thanks in advance for any help that can be obtained,
> Tam



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

Reply via email to