Java Requires Chilkat v11.7.0+

Extract a PDF Table with Merged Cells

See more PDF Text Extraction Examples

Demonstrates how merged cells are reported by the "tables" operation: a header cell that spans several columns has colSpan greater than 1, and a row label that spans several rows has rowSpan greater than 1. Also shows the minRows and minColumns options. Chilkat2 Java

import com.chilkatsoft.chilkat2.ChilkatException;
import com.chilkatsoft.chilkat2.JsonObject;
import com.chilkatsoft.chilkat2.Pdf;

public class ChilkatExample {

    public static void main(String[] args) {
        // This example requires the Chilkat API to have been previously unlocked.
        // See Global Unlock Sample for sample code.

        // The PDF used in this example is available at spans_table.pdf

        Pdf pdf = new Pdf();
        try {
            pdf.loadFile("qa_data/pdf/spans_table.pdf");
        } catch (ChilkatException e) {
            System.out.println(e.getLastErrorText());
            return;
        }

        JsonObject json = new JsonObject();
        JsonObject jsonOut = new JsonObject();

        // Grids smaller than minRows x minColumns are not reported as tables.  The defaults are 2 and 2.
        // Setting them higher skips small boxes that are not really tables.
        json.updateInt("minRows", 3);
        json.updateInt("minColumns", 3);

        try {
            pdf.extractText("tables", json, jsonOut);
        } catch (ChilkatException e) {
            System.out.println(e.getLastErrorText());
            return;
        }

        System.out.println("Tables found: " + jsonOut.intOf("count"));
        System.out.println(jsonOut.intOf("tables[0].rows") + " rows x " + jsonOut.intOf("tables[0].columns") + " columns");

        // A merged cell is reported once, at its top-left position, with its spans.
        // The positions it covers have no cell of their own.
        int rowSpan = 0;
        int colSpan = 0;
        int j = 0;
        int count_j = jsonOut.sizeOfArray("tables[0].cells");
        while (j < count_j) {
            jsonOut.setJ(j);
            rowSpan = jsonOut.intOf("tables[0].cells[j].rowSpan");
            colSpan = jsonOut.intOf("tables[0].cells[j].colSpan");
            if ((rowSpan > 1) || (colSpan > 1)) {
                System.out.println("row " + jsonOut.intOf("tables[0].cells[j].row") + ", col " + jsonOut.intOf("tables[0].cells[j].col") + ": \"" + jsonOut.stringOf("tables[0].cells[j].text") + "\" spans " + rowSpan + " row(s) and " + colSpan + " column(s)");
            }

            j = j + 1;
        }

        // Sample output:

        // Tables found: 1
        // 6 rows x 6 columns
        // row 0, col 0: "Area" spans 2 row(s) and 1 column(s)
        // row 0, col 1: "City" spans 2 row(s) and 1 column(s)
        // row 0, col 2: "First Half" spans 1 row(s) and 2 column(s)
        // row 0, col 4: "Second Half" spans 1 row(s) and 2 column(s)
        // row 2, col 0: "East" spans 2 row(s) and 1 column(s)
        // row 4, col 0: "West" spans 2 row(s) and 1 column(s)
    }
}