A PDF document consists of some primitive and high-level PDF objects. Generally, a PDF document contains nine primitive types of objects and can be interpreted as a graph of linked primitive PDF objects, where an object is one of the following types defined in the PDF specification:

• PDF array
• PDF bool
• PDF dictionary
• PDF name
• PDF null
• PDF number
• PDF reference
• PDF stream
• PDF string

All high-level PDF objects in object model (such as Page, AnnotationBase, Action, etc.) are implemented as wrappers around primitive PDF objects. A wrapper contains a reference to the underlying primitive PDF type (PdfDict, PdfArray, PdfDictObject, etc.) and provides methods and properties for accessing and manipulating the underlying object. The root class for all high-level objects is PdfWrapperBase; it contains a reference to the underlying PDF primitive object defined by IPdfObject.

DsPdf allows you to work directly with the primitive objects used to build all the high-level entities in a PDF document, such as DocumentInfo, a PDF dictionary, using the following listed interfaces and classes, and their methods and properties in GrapeCity.Documents.Pdf.Spec namespace:

The PDF specification defines the properties that can be present in this dictionary (Creator, Author, etc.), but PDF producers can add arbitrary custom properties, such as the SourceModified property, which is often found in various real-world PDF files. Types from GrapeCity.Documents.Pdf.Spec namespace allow you to access (read, write, or edit) such custom elements.

Since most high-level objects in a PDF file are PDF dictionaries, in the DsPdf API, the corresponding objects are derived from the PdfDictWrapper class, which in turn is derived from PdfWrapperBase and uses IPdfDict as the underlying object. The GetPdfStream, GetPdfStreamInfo, and GetPdfStreamData methods of PdfWrapperBase can retrieve data from the PDF stream associated with the PDF dictionary.
Each high-level PDF object (depending on its type) implements one of the primitive interfaces so that the user can use the extension methods of GrapeCity.Documents.Pdf.Spec namespace with these high-level objects.

Refer to the following example code to get image properties from a PDF document:

// Initialize GcPdfDocument.
GcPdfDocument doc = new GcPdfDocument();
                
// Load PDF document.
doc.Load(fs);

// Get image from the PDF document.
var imgs = doc.GetImages();
var pi = imgs[0].Image;

// Write image ID.
Console.WriteLine($"PdfImage object ID: {pi.ObjID}");

/* The PdfImage is a descendant of PdfDictWrapper object and has a lot of methods
   that allow you to get properties and data from the underlying PDF stream object. */
using (PdfStreamInfo psi = pi.GetPdfStreamInfo())
{
    // Get image information such as length filter name, filter decode parameters, etc.
    Console.WriteLine($"    Image stream length: {psi.Stream.Length}");
    Console.WriteLine($"        ImageFilterName: {psi.ImageFilterName}");
    Console.WriteLine($"ImageFilterDecodeParams: {psi.ImageFilterDecodeParams}");
    
    // Dump content of ImageFilterDecodeParams.
    foreach (var kvp in psi.ImageFilterDecodeParams.Dict)
    {
        Console.WriteLine($"{kvp.Key}: {kvp.Value}");
    }
    
    // Get value of BlackIs1.
    var blackIs1 = psi.ImageFilterDecodeParams.GetBool(PdfName.Std.BlackIs1, null);
    Console.WriteLine($"BlackIs1: {blackIs1}");
}
                
// Dump properties of PdfImage dictionary.
Console.WriteLine();
Console.WriteLine("Properties of PdfImage dictionary:");
foreach (KeyValuePair<PdfName, IPdfObject> kvp in pi.PdfDict.Dict)
{
    Console.WriteLine($"{kvp.Key}: {kvp.Value}");
}
                
// Get color space and bits per component.
var cs = pi.Get<IPdfObject>(PdfName.Std.ColorSpace);
Console.WriteLine($"ColorSpace: {cs.GetType().Name} {cs}");
var bpc = pi.Get<IPdfObject>(PdfName.Std.BitsPerComponent);
Console.WriteLine($"BitsPerComponent: {bpc?.GetType().Name} {bpc}");