Swift 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. Swift

func chilkatTest() {
    var success: Bool = false

    //  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

    let pdf = CkoPdf()!
    success = pdf.loadFile(filePath: "qa_data/pdf/spans_table.pdf")
    if success == false {
        print("\(pdf.lastErrorText!)")
        return
    }

    let json = CkoJsonObject()!
    let jsonOut = CkoJsonObject()!

    //  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(jsonPath: "minRows", value: 3)
    json.updateInt(jsonPath: "minColumns", value: 3)

    success = pdf.extractText(op: "tables", jsonIn: json, jsonOut: jsonOut)
    if success == false {
        print("\(pdf.lastErrorText!)")
        return
    }

    print("Tables found: \(jsonOut.int(of: "count").intValue)")
    print("\(jsonOut.int(of: "tables[0].rows").intValue) rows x \(jsonOut.int(of: "tables[0].columns").intValue) 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.
    var rowSpan: Int
    var colSpan: Int
    var j: Int = 0
    var count_j: Int = jsonOut.size(ofArray: "tables[0].cells").intValue
    while j < count_j {
        jsonOut.j = j
        rowSpan = jsonOut.int(of: "tables[0].cells[j].rowSpan").intValue
        colSpan = jsonOut.int(of: "tables[0].cells[j].colSpan").intValue
        if (rowSpan > 1) || (colSpan > 1) {
            print("row \(jsonOut.int(of: "tables[0].cells[j].row").intValue), col \(jsonOut.int(of: "tables[0].cells[j].col").intValue): \"\(jsonOut.string(of: "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)

}