Region-Based Extraction using Java

Extract text from a rectangular page region

Use TextSearchOptions with a Rectangle to restrict extraction to a defined area on a page.

1. Open the source PDF in a Document instance.
2. Create a TextAbsorber to collect text from the selected page area.
3. Create TextSearchOptions for the target Rectangle and enable setLimitToPageBounds(true) so extraction stays inside the visible page box.
4. Apply the configured search options to the absorber and visit the target Page.
5. Write the extracted text buffer to the output file.

public static void extractTextFromRegion(Path inputFile, Path outputFile, int pageNumber, Rectangle rectangle)
        throws Exception {
    try (Document document = new Document(inputFile.toString())) {
        TextAbsorber absorber = new TextAbsorber();
        TextSearchOptions options = new TextSearchOptions(rectangle);
        options.setLimitToPageBounds(true);
        absorber.setTextSearchOptions(options);
        document.getPages().get_Item(pageNumber).accept(absorber);
        Files.writeString(outputFile, absorber.getText());
    }
}

Extract paragraphs with geometry information

Use ParagraphAbsorber to inspect section rectangles and paragraph polygons together with the extracted text.

1. Open the source PDF in a Document instance.
2. Create a ParagraphAbsorber and visit the target Page to build page markup information.
3. Read the first page-markup result and iterate through its sections and paragraphs.
4. Collect each section rectangle, paragraph polygon, and the paragraph text reconstructed from its TextFragment lines.
5. Build the output report with geometry and extracted text details.
6. Write the extracted details to the output file.

public static void extractParagraphsWithGeometry(Path inputFile, Path outputFile) throws Exception {
    try (Document document = new Document(inputFile.toString())) {
        ParagraphAbsorber absorber = new ParagraphAbsorber();
        absorber.visit(document.getPages().get_Item(1));

        PageMarkup pageMarkup = absorber.getPageMarkups().get(0);
        StringBuilder text = new StringBuilder();
        int sectionIndex = 1;
        for (MarkupSection section : pageMarkup.getSections()) {
            text.append("Section ").append(sectionIndex)
                    .append(": rectangle = ").append(section.getRectangle()).append("\n");
            int paragraphIndex = 1;
            for (MarkupParagraph paragraph : section.getParagraphs()) {
                text.append("  Paragraph ").append(paragraphIndex)
                        .append(": polygon = ").append(Arrays.toString(paragraph.getPoints())).append("\n");
                StringBuilder paragraphText = new StringBuilder();
                for (List<TextFragment> line : paragraph.getLines()) {
                    for (TextFragment fragment : line) {
                        paragraphText.append(fragment.getText());
                    }
                    paragraphText.append("\r\n");
                }
                text.append("    Text: ").append(paragraphText).append("\n\n");
                paragraphIndex++;
            }
            sectionIndex++;
        }

        Files.writeString(outputFile, text.toString());
    }
}