Building an index of words contained in a PDF

simone.p
Topic Author
Offline
New Member

IP: 95.248.179.53 5 years 4 months ago - 5 years 4 months ago #15349 by simone.p

Building an index of words contained in a PDF was created by simone.p

Is it possible to extract all the text words contained in a PDF?
The purpose is to build an index with the number of occurrences of each word in each document (I have many document to examine), and (if possible) their positions.

Last edit: 5 years 4 months ago by simone.p.

radaee
Offline
Platinum Member

IP: 111.196.244.255 5 years 4 months ago #15350 by radaee

Replied by radaee on topic Building an index of words contained in a PDF

char position(x,y) in page is relate to char index of page.
and char index is in UTF16 encoding.
that mean you may need record page NO and char index to database for each found index.

and make to be sure, char index for Page.ObjsGetString() is never changed(use special version)

simone.p
Topic Author
Offline
New Member

IP: 217.141.78.166 5 years 4 months ago #15351 by simone.p

Replied by simone.p on topic Building an index of words contained in a PDF

Sorry it's not clear to me how to do it, I have the following:

Code:

   com.radaee.pdf.Global.Init(this)
   val m_doc = com.radaee.pdf.Document()
   m_doc.Open("my/file/path.pdf", "")
   for (x in 0..m_doc.GetPageCount()){
     val page = m_doc.GetPage(x)
     page.ObjsStart()
     // what do I do here in order to extract all words in the PDF
   }

radaee
Offline
Platinum Member

IP: 111.196.244.255 5 years 4 months ago #15352 by radaee

Replied by radaee on topic Building an index of words contained in a PDF

to extrat all texts on page, you can:

Code:

page.ObjsStart();
int ccnt = page.ObjsGetCharCount();//char count in UTF16.
String scontent = page.ObjsGetString(0, ccnt);//string object is UTF8 encoding.
//todo: convert java string value to UTF16 texts.

and you shall make your search engine, search texts in UTF16.
once you get your char index in UTF16 of texts, you can locate position of searched key by:

float[] rect = page.ObjsGetCharRect(char_index);

Time to create page: 0.396 seconds