Extract pdf vector objects

Question

Is there any way to extract the coordinates and dimensions of vector objects with a specific color with C#? Like a "dieline" or a "cut line", for example? I tried with PDFSharp library, but it doesn't seem to have such function. I can extract the swatches but not the coordinates or dimensions of the objects using that respective color.

Answer 1

You can list the vector graphics from a PDF page with PDF4NET as follows (assuming your dieline and cut line are separation colors):

PDFFixedDocument document = new PDFFixedDocument(pdfFile);

PDFContentExtractor ce = new PDFContentExtractor(document.Pages[pageNumber]);
PDFVisualObjectCollection pageVisualObjects = ce.ExtractVisualObjects(false);

for (int i = 0; i < pageVisualObjects.Count; i++)
{
    switch (pageVisualObjects[i].Type)
    {
        case PDFVisualObjectType.Path:
            PDFPathVisualObject pathVisualObject = pageVisualObjects[i] as PDFPathVisualObject;
            if ((pathVisualObject.Pen != null) && 
                (pathVisualObject.Pen.Color.ColorSpace.Type == PDFColorSpaceType.Separation))
            {
                PDFSeparationColorSpace cs = 
                    pathVisualObject.Pen.Color.ColorSpace as PDFSeparationColorSpace;
                if (cs.Colorant == "dieline")
                {
                    Console.WriteLine("dieline stroke");
                    for (int j = 0; j < pathVisualObject.PathItems.Count; j++)
                    {
                        Console.Write("{0}: ", pathVisualObject.PathItems[j].Type);
                        if (pathVisualObject.PathItems[j].Points != null)
                        {
                            for (int k = 0; k < pathVisualObject.PathItems[j].Points.Length; k++)
                            {
                                Console.Write("[{0:0.######}, {1:0.######}] ", 
                                    pathVisualObject.PathItems[j].Points[k].X, 
                                    pathVisualObject.PathItems[j].Points[k].Y);
                            }
                        }
                        Console.WriteLine("");
                    }
                }
            }
            if ((pathVisualObject.Brush != null) &&
                (pathVisualObject.Brush.Color.ColorSpace.Type == PDFColorSpaceType.Separation))
            {
                PDFSeparationColorSpace cs =
                    pathVisualObject.Brush.Color.ColorSpace as PDFSeparationColorSpace;
                if (cs.Colorant == "dieline")
                {
                    Console.WriteLine("dieline fill");
                    for (int j = 0; j < pathVisualObject.PathItems.Count; j++)
                    {
                        Console.Write("{0}: ", pathVisualObject.PathItems[j].Type);
                        if (pathVisualObject.PathItems[j].Points != null)
                        {
                            for (int k = 0; k < pathVisualObject.PathItems[j].Points.Length; k++)
                            {
                                Console.Write("[{0:0.######}, {1:0.######}] ",
                                    pathVisualObject.PathItems[j].Points[k].X,
                                    pathVisualObject.PathItems[j].Points[k].Y);
                            }
                        }
                        Console.WriteLine("");
                    }
                }
            }
            break;
    }
}

This is the output from a sample PDF file that contains a filled and stroked rectangle:

dieline stroke
MoveTo: [20, 605]
LineTo: [270, 605]
LineTo: [270, 705]
LineTo: [20, 705]
CloseSubpath:
dieline fill
MoveTo: [20, 605]
LineTo: [270, 605]
LineTo: [270, 705]
LineTo: [20, 705]
CloseSubpath:

Disclaimer: I work for the company that develops PDF4NET library.