Am 20.10.2018 um 20:02 schrieb Ankit Inkollu:
Thanks for the reply Til. Then I need to find a way to group the
TextPosition objects in terms of each word based on the text received from
ExtractText. Is there any other way which helps me in fetching a word as a
TextPosition object?

No, you'd need to write your own logic, just grab the source code of PDFTextStripper. There may have been an answer in stackoverflow some time ago but I can't find it.

An older one is here:
https://stackoverflow.com/questions/13971656/how-to-avoid-pdfbox-appending-separate-words

IMHO it is oversimplified - the problem is that the space width is relative.

You can easily group the words gotten from the text extraction but that one doesn't have the positions.

Tilman


Thanks
Ankit


On Sat, 20 Oct 2018 at 11:08 PM, Tilman Hausherr <[email protected]>
wrote:

You get the space in ExtractText but the spaces are often not in the PDF
itself so they won't be in the TextPosition objects. PDFBox uses
heuristics to insert spaces in the final extracted text, i.e. assume
there is a space due to the distance between glyphs.

Tilman

Am 20.10.2018 um 19:35 schrieb Ankit Inkollu:
*Scenario:*
To get each word details such as 'Text', 'Font', 'Size' etc from a PDF.

*Approach:*
*1. *Get 'charactersByArticle' available in the PDFTextStripper class for
each page in the PDF.
*2. *It returns a list of TextPosition objects which contains each
characters' text, font, font-size etc.

*Query:*
I am able to get the TextPosition object for each character in the PDF
text
but in order to define words I required the default word-separator (" ")
from 'charactersByArticle'. Why doesn't it print the space character or
is
there a flag which I can set in the PDFTextStripper so that it prints the
text along with the space character.

Thanks
Ankit


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




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

Reply via email to