[ 
https://issues.apache.org/jira/browse/PDFBOX-3749?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15961826#comment-15961826
 ] 

Tilman Hausherr commented on PDFBOX-3749:
-----------------------------------------

I was tired yesterday... today it looks to me as if you made an observation, 
and expect that this observation became a feature, which it isn't. The javadoc 
just mentions getting a string. There is not even a guarantee (unless you use 
the sort option) that the glyphs come in a certain order. Some PDFs with 
columns have their glyphs ordered by line in a column, some by line, so it is 
not even sure that the output is always the best possible.

What you might do, is to create a copy of PDFTextStripper and make writeLine 
public and see if that does what you need.

> void writeString(String text, List<TextPosition> textPositions) is not called 
> per line
> --------------------------------------------------------------------------------------
>
>                 Key: PDFBOX-3749
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-3749
>             Project: PDFBox
>          Issue Type: Bug
>    Affects Versions: 2.0.4
>         Environment: Windows 10 64-bit
>            Reporter: Harun Reşit Zafer
>            Priority: Minor
>              Labels: extraction, style
>         Attachments: contract_00105_SEDAR-marked-1.png, 
> contract_00105_SEDAR.pdf, helloworld-marked-1.png, helloworld.pdf
>
>
> We overwrote the {{void writeString(String text, List<TextPosition> 
> textPositions)}} method of the {{PDFTextStripper}} to extract additional 
> position and style information from the PDFs. We thought this method would be 
> called per line and the elements of the parameter {{List<TextPosition> 
> textPositions}} would be all the letters, including the spaces in a line. 
> This is indeed the case for thousands of the documents. However, one 
> particular document, this is not the case and the {{textPositions}} contains 
> just the letters of a word and {{writeString}} is called per word. 
> I am not sure if this would be counted as a bug because the final extracted 
> text is not affected. 
> The problematic PDF is attached. 



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to