How to export paragraphs in one line using Apache Tika

Asked Jun 10 '20 at 20:15

Active Jun 11 '20 at 18:31

Viewed 239 times

I am passing a PDF document to the Apache Tika software, which is in this format:

PDF document with paragraphs like these:

(iii) 50% of Text Text Text Text Text Text Text Text Text 
Text Text Text Text. 

Text Text Text Text Text Text Text 1 Text Text Text Text Text 
Text Text.

I am getting the text in the same format as the input text provided in the PDF file.

But the output expected is :

(iii) 50% of Text Text Text Text Text Text Text Text Text Text Text Text Text. 

Text Text Text Text Text Text Text 1 Text Text Text Text Text Text Text.

I want to export the paragraphs in one line rather than in the same format as provided in the input file.

I am making the call to Tika in this way:

private Tika tika = new Tika();
String content = tika.parseToString(file);

I receive the content of the file in the content variable.

Is there any configuration through which I can make this happen?

edited Jun 11 '20 at 18:31

asked Jun 10 '20 at 20:15

mshikher

How are you calling Apache Tika? – Gagravarr Jun 11 '20 at 06:18
updated the question – mshikher Jun 11 '20 at 18:31
Ask Apache Tika for the XHTML version instead of plain text, then split yourself by `
` tags?
– Gagravarr Jun 12 '20 at 04:37
@Gagravarr, is there a way that I can extract all the text present inside the
tags one by one?
– mshikher Jun 16 '20 at 07:14
A quick-and-dirty solution would probably be just to split on a regexp based on the tags... "proper" solution is to write a custom ContentHandler, pass that to Tika, and have that route the text into chunks by paragraph tag – Gagravarr Jun 16 '20 at 07:36

How to export paragraphs in one line using Apache Tika

0 Answers0