Using PDFbox to determine the coordinates of words

Question

I'm using PDFbox to extract the coordinates of words/strings in a PDF document, and have so far had success determining the position of individual characters. this is the code thus far, from the PDFbox doc:

package printtextlocations;

import java.io.*;
import org.apache.pdfbox.exceptions.InvalidPasswordException;

import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.common.PDStream;
import org.apache.pdfbox.util.PDFTextStripper;
import org.apache.pdfbox.util.TextPosition;

import java.io.IOException;
import java.util.List;

public class PrintTextLocations extends PDFTextStripper {

    public PrintTextLocations() throws IOException {
        super.setSortByPosition(true);
    }

    public static void main(String[] args) throws Exception {

        PDDocument document = null;
        try {
            File input = new File("C:\\path\\to\\PDF.pdf");
            document = PDDocument.load(input);
            if (document.isEncrypted()) {
                try {
                    document.decrypt("");
                } catch (InvalidPasswordException e) {
                    System.err.println("Error: Document is encrypted with a password.");
                    System.exit(1);
                }
            }
            PrintTextLocations printer = new PrintTextLocations();
            List allPages = document.getDocumentCatalog().getAllPages();
            for (int i = 0; i < allPages.size(); i++) {
                PDPage page = (PDPage) allPages.get(i);
                System.out.println("Processing page: " + i);
                PDStream contents = page.getContents();
                if (contents != null) {
                    printer.processStream(page, page.findResources(), page.getContents().getStream());
                }
            }
        } finally {
            if (document != null) {
                document.close();
            }
        }
    }

    /**
     * @param text The text to be processed
     */
    @Override /* this is questionable, not sure if needed... */
    protected void processTextPosition(TextPosition text) {
        System.out.println("String[" + text.getXDirAdj() + ","
                + text.getYDirAdj() + " fs=" + text.getFontSize() + " xscale="
                + text.getXScale() + " height=" + text.getHeightDir() + " space="
                + text.getWidthOfSpace() + " width="
                + text.getWidthDirAdj() + "]" + text.getCharacter());
    }
}

This produces a series of lines containing the position of each character, including spaces, that looks like this:

String[202.5604,41.880127 fs=1.0 xscale=13.98 height=9.68814 space=3.8864403 width=9.324661]P

Where 'P' is the character. I have not been able to find a function in PDFbox to find words, and I am not familiar enough with Java to be able to accurately concatenate these characters back into words to search through even though the spaces are also included. Has anyone else been in a similar situation, and if so how did you approach it? I really only need the coordinate of the first character in the word so that parts simplified, but as to how I'm going to match a string against that kind of output is beyond me.

Answer 1

Based on the original idea here is a version of the text search for PDFBox 2. The code itself is rough, but simple. It should get you started fairly quickly.

import java.io.IOException;
import java.io.Writer;
import java.util.List;
import java.util.Set;
import lu.abac.pdfclient.data.PDFTextLocation;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.text.PDFTextStripper;
import org.apache.pdfbox.text.TextPosition;

public class PrintTextLocator extends PDFTextStripper {

    private final Set<PDFTextLocation> locations;

    public PrintTextLocator(PDDocument document, Set<PDFTextLocation> locations) throws IOException {
        super.setSortByPosition(true);
        this.document = document;
        this.locations = locations;
        this.output = new Writer() {
            @Override
            public void write(char[] cbuf, int off, int len) throws IOException {
            }
            @Override
            public void flush() throws IOException {
            }

            @Override
            public void close() throws IOException {
            }
        };
    }

    public Set<PDFTextLocation> doSearch() throws IOException {

        processPages(document.getDocumentCatalog().getPages());
        return locations;
    }

    @Override
    protected void writeString(String text, List<TextPosition> textPositions) throws IOException {
        super.writeString(text);

        String searchText = text.toLowerCase();
        for (PDFTextLocation textLoc:locations) {
            int start = searchText.indexOf(textLoc.getText().toLowerCase());
            if (start!=-1) {
                // found
                TextPosition pos = textPositions.get(start);
                textLoc.setFound(true);
                textLoc.setPage(getCurrentPageNo());
                textLoc.setX(pos.getXDirAdj());
                textLoc.setY(pos.getYDirAdj());
            }
        }

    }


}

Answer 2

There is no function in PDFBox that allows you to extract words automatically. I'm currently working on extracting data to gather it into blocks and here is my process:

1. I extract all the characters of the document (called glyphs) and store them in a list.

2. I do an analysis of the coordinates of each glyph, looping over the list. If they overlap (if the top of the current glyph is contained between the top and bottom of the preceding/or the bottom of the current glyph is contained between the top and bottom of the preceding one), I add it to the same line.

3. At this point, I have extracted the different lines of the document (be careful, if your document is multi-column, the expression "lines" means all the glyphs that overlap vertically, ie the text of all the columns that have the same vertical coordinates).

4. Then, you can compare the left coordinate of the current glyph to the right coordinate of the preceding one to determine if they belong to the same word or not (the PDFTextStripper class provides a getSpacingTolerance() method that gives you, based on trials and errors, the value of a "normal" space. If the difference between the right and the left coordinates is lower than this value, both glyphs belong to the same word.

I applied this method to my work and it works good.

Answer 3

take a look on this, I think it's what you need.

https://jackson-brain.com/using-pdfbox-to-locate-text-coordinates-within-a-pdf-in-java/

Here is the code:

import java.io.File;
import java.io.IOException;
import java.text.DecimalFormat;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;

import org.apache.pdfbox.exceptions.InvalidPasswordException;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.common.PDStream;
import org.apache.pdfbox.util.PDFTextStripper;
import org.apache.pdfbox.util.TextPosition;

public class PrintTextLocations extends PDFTextStripper {

public static StringBuilder tWord = new StringBuilder();
public static String seek;
public static String[] seekA;
public static List wordList = new ArrayList();
public static boolean is1stChar = true;
public static boolean lineMatch;
public static int pageNo = 1;
public static double lastYVal;

public PrintTextLocations()
        throws IOException {
    super.setSortByPosition(true);
}

public static void main(String[] args)
        throws Exception {
    PDDocument document = null;
    seekA = args[1].split(",");
    seek = args[1];
    try {
        File input = new File(args[0]);
        document = PDDocument.load(input);
        if (document.isEncrypted()) {
            try {
                document.decrypt("");
            } catch (InvalidPasswordException e) {
                System.err.println("Error: Document is encrypted with a password.");
                System.exit(1);
            }
        }
        PrintTextLocations printer = new PrintTextLocations();
        List allPages = document.getDocumentCatalog().getAllPages();

        for (int i = 0; i < allPages.size(); i++) {
            PDPage page = (PDPage) allPages.get(i);
            PDStream contents = page.getContents();

            if (contents != null) {
                printer.processStream(page, page.findResources(), page.getContents().getStream());
            }
            pageNo += 1;
        }
    } finally {
        if (document != null) {
            System.out.println(wordList);
            document.close();
        }
    }
}

@Override
protected void processTextPosition(TextPosition text) {
    String tChar = text.getCharacter();
    System.out.println("String[" + text.getXDirAdj() + ","
            + text.getYDirAdj() + " fs=" + text.getFontSize() + " xscale="
            + text.getXScale() + " height=" + text.getHeightDir() + " space="
            + text.getWidthOfSpace() + " width="
            + text.getWidthDirAdj() + "]" + text.getCharacter());
    String REGEX = "[,.\\[\\](:;!?)/]";
    char c = tChar.charAt(0);
    lineMatch = matchCharLine(text);
    if ((!tChar.matches(REGEX)) && (!Character.isWhitespace(c))) {
        if ((!is1stChar) && (lineMatch == true)) {
            appendChar(tChar);
        } else if (is1stChar == true) {
            setWordCoord(text, tChar);
        }
    } else {
        endWord();
    }
}

protected void appendChar(String tChar) {
    tWord.append(tChar);
    is1stChar = false;
}

protected void setWordCoord(TextPosition text, String tChar) {
    tWord.append("(").append(pageNo).append(")[").append(roundVal(Float.valueOf(text.getXDirAdj()))).append(" : ").append(roundVal(Float.valueOf(text.getYDirAdj()))).append("] ").append(tChar);
    is1stChar = false;
}

protected void endWord() {
    String newWord = tWord.toString().replaceAll("[^\\x00-\\x7F]", "");
    String sWord = newWord.substring(newWord.lastIndexOf(' ') + 1);
    if (!"".equals(sWord)) {
        if (Arrays.asList(seekA).contains(sWord)) {
            wordList.add(newWord);
        } else if ("SHOWMETHEMONEY".equals(seek)) {
            wordList.add(newWord);
        }
    }
    tWord.delete(0, tWord.length());
    is1stChar = true;
}

protected boolean matchCharLine(TextPosition text) {
    Double yVal = roundVal(Float.valueOf(text.getYDirAdj()));
    if (yVal.doubleValue() == lastYVal) {
        return true;
    }
    lastYVal = yVal.doubleValue();
    endWord();
    return false;
}

protected Double roundVal(Float yVal) {
    DecimalFormat rounded = new DecimalFormat("0.0'0'");
    Double yValDub = new Double(rounded.format(yVal));
    return yValDub;
}
}

Dependencies:

PDFBox, FontBox, Apache Common Logging Interface.

You can run it by typing on command line:

javac PrintTextLocations.java 
sudo java PrintTextLocations file.pdf WORD1,WORD2,....

the output is similar to:

[(1)[190.3 : 286.8] WORD1, (1)[283.3 : 286.8] WORD2, ...]