Skip to content

Instantly share code, notes, and snippets.

Last active January 8, 2024 03:08
Show Gist options
  • Save i000313/6372210 to your computer and use it in GitHub Desktop.
Save i000313/6372210 to your computer and use it in GitHub Desktop.
Java: Class to get PDF annotations using the itextpdf lib. #java, #pdf, #annotations, #pdf-annotations, #itextpdf
package com.itextpdf;
import com.itextpdf.text.Rectangle;
import com.itextpdf.text.pdf.PdfArray;
import com.itextpdf.text.pdf.PdfDictionary;
import com.itextpdf.text.pdf.PdfName;
import com.itextpdf.text.pdf.PdfReader;
import com.itextpdf.text.pdf.PdfString;
import com.itextpdf.text.pdf.parser.FilteredTextRenderListener;
import com.itextpdf.text.pdf.parser.LocationTextExtractionStrategy;
import com.itextpdf.text.pdf.parser.PdfTextExtractor;
import com.itextpdf.text.pdf.parser.RegionTextRenderFilter;
import com.itextpdf.text.pdf.parser.RenderFilter;
import com.itextpdf.text.pdf.parser.TextExtractionStrategy;
* Class for getting PDF annotations using the itextpdf lib (tested whith itextpdf-5.4.3).
* At the moment, this class allows to get highlighted text, underlined text, and
* text in comments.
* <br/>
* Document management — Portable document format — Part 1: PDF 1.7
* @author PSantos
public class PDFAnnotations {
public static void main(String args[]) {
try {
// Reads and parses a PDF document
PdfReader reader = new PdfReader("C:\\Users\\PSantos\\Desktop\\lixo\\exemplo.pdf");
// For each PDF page
for (int i = 1; i <= reader.getNumberOfPages(); i++) {
// Get a page a PDF page
PdfDictionary page = reader.getPageN(i);
// Get all the annotations of page i
PdfArray annotsArray = page.getAsArray(PdfName.ANNOTS);
// If page does not have annotations
if (page.getAsArray(PdfName.ANNOTS) == null) {
// For each annotation
for (int j = 0; j < annotsArray.size(); ++j) {
// For current annotation
PdfDictionary curAnnot = annotsArray.getAsDict(j);
// Check the annotation subtype and print its text if not null
writeAnnotation(curAnnot, reader, i);
} catch (Exception e) {
* Check the annotation subtype and print its text.
* @param annot annotation to write.
* @param reader pdf document containing the annotation.
* @param pageNumber pdf page number containing the annotation.
* @throws IOException
public static void writeAnnotation(PdfDictionary annot, PdfReader reader, int pageNumber) throws IOException {
if(annot == null) {
// System.out.print(annot.get(PdfName.SUBTYPE));
// System.out.print(" -> Rect: " + annot.get(PdfName.RECT));
PdfString text = null;
boolean mayHaveTextAnnotated = false;
// Highlights with comments (balloons) and Highliths
if (PdfName.HIGHLIGHT.equals(annot.get(PdfName.SUBTYPE))) {
// Only Highlights with comments may have text
text = (PdfString) annot.get(PdfName.CONTENTS);
mayHaveTextAnnotated = true;
} else if (PdfName.UNDERLINE.equals(annot.get(PdfName.SUBTYPE))) {
text = annot.getAsString(PdfName.CONTENTS);
mayHaveTextAnnotated = true;
// A comment (a balloon with a comment)
} else if (PdfName.TEXT.equals(annot.get(PdfName.SUBTYPE))) {
text = annot.getAsString(PdfName.CONTENTS);
} else {
text = annot.getAsString(PdfName.CONTENTS);
if(text != null) {
System.out.println(" -> " + text);
if(mayHaveTextAnnotated) {
PdfArray rectangle = (PdfArray) annot.get(PdfName.RECT); // ex: [82.1569, 757.575, 124.395, 769.305]
String textHighlighted = getTextFromRectangle(rectangle, reader, pageNumber);
if(textHighlighted != null) {
System.out.println(" Annotated text -> " + textHighlighted);
* Extracts the text {@code rectangle}, located on page {@code pageNumber}
* of the pdf {@code reader}.
* <p>The text extracted by this method is not perfect. Usually extracts an
* unnecessary (not desirable) extra character. For instance, for an annotations
* like "[This is an annotated] text", the method will extract
* "This is an annotated t". </p> The extra character may appear before or
* after the annotated text.
* @param reader
* @param pageNumber
* @param rectangle ex: [82.1569, 757.575, 124.395, 769.305]
* @return the extracted text or null.
* @throws IOException
public static String getTextFromRectangle(PdfArray rectangle,
PdfReader reader, int pageNumber) throws IOException {
if(rectangle == null) {
return null;
// Get the retangle coodinates
float llx = rectangle.getAsNumber(0).floatValue();
float lly = rectangle.getAsNumber(1).floatValue();
float urx = rectangle.getAsNumber(2).floatValue();
float ury = rectangle.getAsNumber(3).floatValue();
Rectangle rect = new Rectangle(llx, lly, urx, ury);
RenderFilter filter = new RegionTextRenderFilter(rect);
TextExtractionStrategy strategy =
new FilteredTextRenderListener(
new LocationTextExtractionStrategy(), filter);
return PdfTextExtractor.getTextFromPage(reader, pageNumber, strategy);
Copy link

May I ask if is there a way to deal with non-ascii words in FreeText.
I have read several words that typed into pdf, it returns with correct ascii part, but the unicode part is mess.
I decipher with utf-16 and get better result: ~70% words correct, but there are still many words weired.

It seems that pdf store a paragraph of freetext using multiple encoding scheme ?

I call text..getEncoding andget nothing


Copy link

i000313 commented Jan 26, 2021

Hi. Sorry, but I can't help you because I never used this code.

Copy link

rmd13 commented Jan 26, 2021

Hi, I have solved. I mistake signed to unsigned.


Copy link

i000313 commented Feb 23, 2021

Glad to know. Thank you for let me know.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment