Multithreaded PDF reading in C#

//---------------------------------------------------------------------------------------
// Copyright (c) 2001-2021 by PDFTron Systems Inc. All Rights Reserved.
// Consult legal.txt regarding legal and license information.	 
//---------------------------------------------------------------------------------------

using System;
using System.Drawing;

using pdftron;
using pdftron.Common;
using pdftron.PDF;
using pdftron.SDF;
using System.Threading;
using System.Threading.Tasks;
using System.Collections.Generic;

namespace PDFDrawTestCS
{
	/// <summary>
	//---------------------------------------------------------------------------------------
	// The following sample illustrates how to use PDFDoc locking mechanisms to access the document
	// concurrently.  PDFDoc uses a recursive shared lock model.  Multiple threads can read the document 
	// at the same time, but only one thread can write to the document.  A given thread can acquire 
	// as many locks of the same type as it wants, in a recursive fashion.
	//
	// It's important to note that you cannot upgrade a PDFDoc read lock to a write lock.  Because of
	// the nature of the lock, this would cause a deadlock, and is not allowed.
	//
	// If a thread attempts to acquire a write lock while holding a read lock, an exception will be
	// thrown.  As a side effect, this means certain PDFNet API calls which implicitly acquire a write
	// lock on the document will throw an exception if called while holding a read lock.  The API
	// documents which methods acquire a write lock on the document.
	//
	// For more information on document locking in PDFNet, please consult the following knowledge base
	// article: https://groups.google.com/forum/#!topic/pdfnet-sdk/pDZsqrjCEdg
	//---------------------------------------------------------------------------------------

	class Class1
	{
		private static pdftron.PDFNetLoader pdfNetLoader = pdftron.PDFNetLoader.Instance();
		static Class1() {}
		
		/// <summary>
		/// A static method for rasterizing a given page to a file.
		/// </summary>
		static void DoDraw(PDFDoc doc, Page pg, string output_path)
		{
			//acquire a read lock, since we will be accessing the document.
			doc.LockRead();

			//Acquiring a write lock when holding a read lock is illegal,
			//and would cause an exception to be thrown:
			//doc.Lock();

			//draw the page
			using (PDFDraw draw = new PDFDraw())
			{
				draw.SetDPI(92);
				draw.Export(pg, output_path);
				Console.WriteLine("Rendered page: Result saved in {0}", output_path);
			}

			//release the read lock
			doc.UnlockRead();
		}

		/// <summary>
		/// The main entry point for the application.
		/// </summary>
		static void Main(string[] args)
		{
			// The first step in every application using PDFNet is to initialize the 
			// library and set the path to common PDF resources. The library is usually 
			// initialized only once, but calling Initialize() multiple times is also fine.
			PDFNet.Initialize(PDFTronLicense.Key);				   
			
			// Relative path to the folder containing test files
			string input_path =  "../../../../TestFiles/";
			string output_path = "../../../../TestFiles/Output/";

			//Example - Exporting images of each page in the document in parallel,
			//          and annotating the document in the main thread.
			try  
			{
				// Open the PDF document.
				using (PDFDoc doc = new PDFDoc(input_path + "newsletter.pdf")) 
				{
					// Lock the document, since we are going to annotate it in this thread.
					doc.Lock();

					// Initialize the security handler, in case the PDF is encrypted.
					doc.InitSecurityHandler();

					List<Task> tasks = new List<Task>();

					// Iterate through each page in the document
					for (PageIterator itr=doc.GetPageIterator(); itr.HasNext(); itr.Next()) 
					{
						//note that locking a second time does not cause a deadlock:
						doc.Lock();

						// You can acquire a read lock while holding a write lock:
						doc.LockRead();
						doc.UnlockRead();

						// Choose an output path for this page
						string page_output_path = output_path + "newsletter_" + itr.GetPageNumber() + ".png";

						// Create an asynchronous task to draw the page
						Page pg = itr.Current();
						Task t = new Task(() => DoDraw(doc, pg, page_output_path));
						// Start the Task (although it won't be able to access the document, since we have a write lock)
						t.Start();
						// Add it to our list of Tasks so we can Wait() for it later.
						tasks.Add(t);

						Console.WriteLine("Adding stamp to PDFDoc, page " + itr.GetPageNumber());

						// Create a stamp annotation (See AnnotationTest for more details)
						pdftron.PDF.Annots.RubberStamp stamp = pdftron.PDF.Annots.RubberStamp.Create(doc.GetSDFDoc(), new Rect(30, 30, 300, 200));
						stamp.SetIcon("Approved");
						itr.Current().AnnotPushBack(stamp);

						// Releasing the lock decrements our lock count, 
						// but the thread still holds a write lock on the document:
						doc.Unlock();
					}
	 
					// Now we release the write lock, and the 
					// PDFDraw tasks can begin execution
					doc.Unlock();

					// Wait for the PDFDraw tasks to complete.
					foreach (Task t in tasks)
					{
						t.Wait();
					}
				}
			}
			catch (PDFNetException e) 
			{
				Console.WriteLine(e.Message);
			}
			PDFNet.Terminate();
		}
	}
}