That Conference 2018 – Correcting Common Mistakes When Using async/await in .NET

That Conference 2018, Kalahari Resort, Lake Delton, WI
Correcting Common Mistakes When Using async/await in .NET – Brandon Minnick

Day 1, 6 Aug 2018  4:00 PM

Disclaimer: This post contains my own thoughts and notes based on attending That Conference 2018 presentations. Some content maps directly to what was originally presented. Other content is paraphrased or represents my own thoughts and opinions and should not be construed as reflecting the opinion of the speakers.

Brandon Minnick

Developer Advocate, Microsoft

@TheCodeTraveler

https://www.codetraveler.io/THAT-2018-AsyncAwait/

Executive Summary

  • Helps to know underlying IL that’s generated
  • Helps to know core task-based objects that async/await use
  • ConfigureAwait(false) if you don’t need to come back to UI thread
  • .GetAwaiter().GetResult() instead of await to get synchronous execution

Multi Threading

  • e.g. ReadDataFromUrl
    • Our method is async
    • does await on DownloadDataTaskAsync
    • 1st chunk of code is run by the thread that calls this method
    • 2nd thread–calls the Download method
    • After await, execution picks up with 1st thread
  • Typically, first thread, i.e. main thread, is the UI thread
  • UI thread
    • Draws UI on screen
    • Rule-of-thumb for executing on UI thread. Because refresh rate is 60Hz, don’t take more then 1/60 sec (17 ms)

Intermediate Language

  • Compiler generates class for your original method
  • Implements IAsyncStateMachine
  • Every local variable from method becomes private field in class
  • MoveNext method
    • Giant switch statement
    • $PC keeping track of state
    • Also a big try/catch block surrounding everything

Quick Review

  • async keyword adds 100 bytes
  • Every async method becomes a class

Await Every Async Method

  • Non-awaited async methods hide exceptions
  • Ditto for Task.Run — GetAwaiter().GetResult() will ensure we catch exception

Let’s Fix Some Code

  • Can’t await in a constructor
    • ICommand–made for fire and foreget
    • In Execute method, do async lambda with await
    • So from constructor, we can call MyCommand.Execute(null)
  • Never use async void ?
    • Not entirely true
    • Don’t use async void when not on UI thread
    • But on UI thread, if you throw exception from within async void–you’ll catch the exception
  • So from constructor, call Initialize(), which is an async void and can do the await
  • NEVER use .Wait
    • Blocks calling thread entirely
    • Doesn’t unwrap an exception
  • NEVER use .Result
  • Way better method
    • GetAwaiter().GetResult()
    • Does unwrap any exceptions
    • Can get the result from the async method
  • Note: .GetAwaiter().GetResult() is to be used in non-async method
    • In async method, just use await keyword
  • ConfigureAwait(false)
    • By default, you context switch back to UI thread
    • But this is expensive–background thread has to wait for UI thread
    • Context switch takes time
    • When background thread is done, execution stays on that background thread. Does not come back to UI thread
    • But note that you do need to come basck to UI thread if that method does some stuff with the UI, e.g. modify a control
  • ConfigureAwait(true)
    • This is the default, equivalent to not having ConfigureAwait() there at all
  • Don’t call .Result
    • Replace with await that method call to get the result directly
  • Special case, returning await something that returns Task<>
    • You can return the task directly, rather than awaiting it, then doing async on the method
    • Look for “return await”
  • Exception to this rule
    • You should await if you want to catch exception in method that you’re calling
    • So you could argue not to every do that trick

Async/Await Best Practices

  • Never use .Wait or .Result
    • Use await keyword
    • Or .GetAwaiter().GetResult()
  • Fire and forget tasks
    • Use ICommand
    • (or) async void (if you know you’re going to be on main thread)
  • Avoid return await
    • Remove async keyword
    • Except in try/catch blocks or using blocks
  • Utilize ConfigureAwait(false) as much as possible
    • In methods that don’t touch the UI

Xamarin University

  • Various course, e.g. CSC350: Using async await

That Conference 2017 – Concurrent Programming in .NET

That Conference 2017, Kalahari Resort, Lake Delton, WI
Concurrent Programming in .NET – Jason Bock (@JasonBock)

Day 2, 8 Aug 2017

Disclaimer: This post contains my own thoughts and notes based on attending That Conference 2017 presentations. Some content maps directly to what was originally presented. Other content is paraphrased or represents my own thoughts and opinions and should not be construed as reflecting the opinion of the speakers.

Executive Summary

  • Doing concurrency correctly is very hard when doing it at a lower level
  • Don’t use Thread, ThreadPool anymore
  • Learn and use async/await


Background

  • Concurrency is tough, not as easy as doing things sequentially
  • Games are turn-based
    • But interesting exercise to do chess concurrently
    • Much more complicated, just as a game
  • We’ll just look at typical things that people struggle with


Terms

  • Concurrency – Work briefly and separately on each task, one at a time
  • Parallelism – People all working at the same time, doing separate tasks
  • Asynchronous – One you start a task, you can go off and do something else in the meantime


Recommendations

  • Stop using Thread, ThreadPool directly
  • Understand async/await/Task
  • Use locks wisely
  • Use concurrent and immutable data structures
  • Let someone else worry about concurrency (actors)


Stop Using Thread, ThreadPool

  • This was the way to do concurrency back in .NET 1.0
  • Diagram – Concurrent entities on Windows platform
    • Process – at least one thread
    • Thread – separate thread of execution
    • Job – group of processes
    • Fiber – user-level concurrent tasks
  • But mostly we focus on process and thread
  • Example–multiple threads
    • Use Join to observe at end
  • Problems
    • Stack size
    • # physical cores
  • # Cores
    • May not really be doing things in parallel
    • More threads than cores–context switch and run one at a time on a core
    • Context switches potentially slow things down
    • Don’t just blindly launch a bunch of threads
  • Memory
    • Each thread -> 1 MB memory
  • ThreadPool is an improvement
    • You don’t know when threads are complete
    • Have to use EventWaitHandle, WaitAll
    • Lots of manual works
    • At least threads get reused
  • Existing code that uses Threads works just fine


Async/Await/Tasks

  • Models
    • Asynchronous Programming Model (APM)
    • Event-based Asynchronous Patter (EAP)
    • Task-Based Asynchrony (TAP)
  • Demo – console app
    • AsyncContext.Run – can’t call async method from non-async method
    • AsyncContext – from Stephen Cleary – excellent book
    • This goes away in C# 7.1 – compiler will allow calling async method
    • Reading file
  • Misconception – that you create threads when you call async method
    • No, not true
    • Just because method is asynchronous, it won’t necessarily be on another thread
  • Use async when doing I/O bound stuff
    • Calling ReadLineAsync, you hit IO Completion Point; when done, let caller know
    • When I/O is done, calling thread can continue
    • Asynchronous calls may also not necessarily hit IOCP
  • If you do I/O in non-blocking way on a thread, you can then use thread to do CPU work while the I/O is happening
    • Performance really does matter–e.g. use fewer servers
  • When you call async method, compiler creates asynchronous state machine
    • Eric Lippert’s continuation blog post series
  • Task object has IsCompleted
    • Generated code need to first check state to see if task completed right away
  • Good news is–that you don’t need to write the asynch state machine plumbing code
  • Can use Tasks to run things in parallel
    • Task.Run(() => ..)
    • await Task.WhenAll(t1, t2);
  • Tasks are higher level abstraction than threads
  • Don’t ever do async void
    • Only use it for event handlers
  • Keep in mind that async tasks may actually run synchronously, under the covers


Demo – Evolving Expressions

  • Code uses all cores available


Locks

  • Don’t use them (or try not to use them)
  • If you get them wrong, you can get deadlocks
  • Don’t lock on
    • Strings – You could block other uses of string constant
    • Integer – you don’t lock on the same thing each time, since you lock on boxed integer
  • Just use object to lock on
  • Interlocked.Add/Decrement
    • For incrementing/decrementing integers
    • Faster
  • Tool recommendation: benchmark.net
  • SpinLock
    • Enter / Exit
    • Spins in a while loop
    • Can be slower than lock statement
    • Don’t use SpinLock unless you know it will be faster than other methods


Data Structures

  • List<T> is not thread-safe
  • ConcurrentStack<T>
    • Use TryPop
  • ImmutableStack<T>
    • When you modify, you must capture the new stack, result of the operation
    • Objects within the collection can change


Actors

  • Service Fabric Reliable Actors
  • Benefits
    • Resource isolation
    • Asynchronous communication, but single-threaded
      • The actor itself is single-threaded
      • So in writing the actor, you don’t need to worry about thread safety


Demo – Actors – Using Orleans

  • “Grains” for actors

.NET Basics – Do Work in Background Thread to Keep GUI Responsive

One of the most important things that differentiates a “quick and dirty” application from one that has been designed well is how the application’s user interface behaves during lengthy operations.  The quick-and-dirty approach is to just do all of your work in a button’s Click event handler and not worry about the user interface.  The problem with this is that the GUI will freeze up while the application does whatever work it needs to do.

A well designed application, on the other hand, is one that is careful to do as much work as possible in background threads, keeping the GUI responsive and making sure that it makes it obvious to the user that work is going on in the background and adjusts the GUI to disallow any user actions that don’t apply until after the work finishes.

Under .NET 2.0, doing work on a background thread has become a lot easier, with the introduction of the BackgroundWorker class.  You no longer have to worry about cross-threading exceptions and checking a control’s InvokeRequired property.

A Simple Example of Using the BackgroundWorker Class

In this post, I’ll create a simple example of how you might use the BackgroundWorker class to do some work on a background thread and keep your GUI responsive.  We’ll start with a simple example that demonstrates how the GUI can become blocked and then evolve the application to make full use of the capabilities of the BackgroundWorker class.

Here are the basic players. We’ll have a FileReader class/object that reads text from a text file. And a Win Forms form with a button to initiate the file read operation and some GUI elements to display the status/results of the read operation.

Note: All code samples presented here can be found in CodePlex, at threadsafepubsub.codeplex.com

Iteration #1 – The Simplest Possible Solution

Let’s say that we just want to read a text file and return/display the number of lines found in the file. We can just make a call to our FileReader object, which returns the number of lines, and then display that number in our UI. Super simple.

This iteration is implemented in the files Form1.cs and FileReader1.cs.

Here’s what the GUI looks like.  If you click on the Read File button, you get a File Open dialog where you can select a file to read.  The file is read in and then we write out the # lines read, below the button.

001-Iter1Client

So far, so good.  This is how most simple user interfaces are written–you click on a button, which launches a Click callback, which does some work, and then returns to the caller.

Here’s what the FileReader1 class looks like, with a simple ReadTheFile method:

using System;
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Text;

namespace ThreadSafePubSubUI
{
    public class FileReader1
    {
        // Read specified text file & return # lines
        public int ReadTheFile(string fileName)
        {
            int numLines = 0;

            using (StreamReader sr = new StreamReader(fileName))
            {
                string nextLine;
                while ((nextLine = sr.ReadLine()) != null)
                {
                    numLines++;
                }
            }

            return numLines;
        }
    }
}

And here’s the click event handler for the form: the guy that invokes ReadTheFile.

        private void btnSelect_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.CheckFileExists = true;
            ofd.CheckPathExists = true;

            if (ofd.ShowDialog() == DialogResult.OK)
            {
                FileReader1 fr = new FileReader1();
                int numLines = fr.ReadTheFile(ofd.FileName);

                lblResults.Text = string.Format("We read {0} lines", numLines.ToString());
            }
        }

But what if the function that does the work takes a longer amount of time?  It’s pretty common for some action initiated by the user to take a little time.  What happens to the GUI while they are waiting?  We can simulate this by just adding a Thread.Sleep call in the ReadTheFile method.

            Thread.Sleep(3000);     // Simulate lengthy operation

Let’s also add a line in the btnSelect_Click method, to write a “busy” message to the GUI while we are processing.  Here is the updated click event handler:

        private void btnSelect_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.CheckFileExists = true;
            ofd.CheckPathExists = true;

            if (ofd.ShowDialog() == DialogResult.OK)
            {
                lblResults.Text = " ... reading the file ...";
                FileReader1 fr = new FileReader1();
                int numLines = fr.ReadTheFile(ofd.FileName);

                lblResults.Text = string.Format("We read {0} lines", numLines.ToString());
            }
        }

What happens is not good.  Two bad things happen, from a user’s point of view:

  • The user interface is completely unresponsive during the file read operation
  • Our “reading the file” message is not displayed

What happened?  Well, because everything is on one thread, our user interface thread doesn’t respond to mouse clicks until ReadTheFile finishes.  Worse, even though we set the label’s Text property before we call ReadTheFile, the message loop doesn’t get a chance to process that change, and update the text, before we go out to lunch in ReadTheFile.

What we need to do to fix this is: do the file read operation on a different thread

The easiest way to do some work on a background thread, keeping the GUI responsive, is to use the BackgroundWorker class.

Iteration #2 – Using the BackgroundWorker Class

You should be doing very little actual work in GUI control event handlers like the Button.Click method.  It’s a good idea to:

  • Move code that does actual work outside of the user interface class
  • Do all work on a background thread.

We want to move code into a separate library or class, rather than having it in our Click event handler, to keep our user interface code separate from our functional code.  This is just a cleaner architecture and makes our code more maintainable, easier to test, and more extensible.

We also want to do as much work as possible on a different thread from the main thread handling the GUI.  If you do your work on the same thread, you risk locking up the user interface.  (As we saw in Iteration #1).

If you’re using the .NET Framework version 2.0 or later, the best way to do work on a background thread is to use the BackgroundWorker class.  This class gives us the ability to do some work on a background thread, provides progress and completed events “out of the box” and also ensures that these callbacks execute on the correct (GUI thread).

What do I mean by “execute on the correct thread”?  Here’s how it works.  To ensure that the GUI stays responsive, we want to do any non-trivial work on a background thread.  This thread can run in parallel to the GUI thread, so the user will still be able to interact with the GUI while the work is being done.

When the work finishes, we likely want to update something in the GUI to indicate this.  (E.g. change the text on a label to indicate that the operation is done).  Our GUI object will be notified by handling an event that the worker object fires.  But since we need to update the GUI, this event handler must be executing on the same thread as the user interface.

This last point is very important.  The core rule in Windows UI programming to remember is: the only thread that can update/change a user interface control is the thread that created it.  (This is true for Windows Forms applications, which use the Single Threaded Apartment model).

The beauty of the BackgroundWorker class is that it automatically handles all of this thread logic:

  • It does work on a background thread
  • It ensures that completed/progress events are fired on the original GUI thread

Let’s change our earlier file-reading example to use the BackgroundWorker.  This example can be found in the threadsafepubsub.codeplex.com project in the Form2/FileWorker2 classes.

Here’s the new Click event handler, where we create the background worker, attach our event handlers, and then tell it to go do some work.

private void btnSelect_Click(object sender, EventArgs e)
        private void btnSelect_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.CheckFileExists = true;
            ofd.CheckPathExists = true;

            if (ofd.ShowDialog() == DialogResult.OK)
            {
                lblResults.Text = " ... reading the file ...";

                // Set up background worker object & hook up handlers
                BackgroundWorker bgWorker;
                bgWorker = new BackgroundWorker();
                bgWorker.DoWork += new DoWorkEventHandler(bgWorker_DoWork);
                bgWorker.RunWorkerCompleted += new RunWorkerCompletedEventHandler(bgWorker_RunWorkerCompleted);

                // Launch background thread to do the work of reading the file.  This will
                // trigger BackgroundWorker.DoWork().  Note that we pass the filename to
                // process as a parameter.
                bgWorker.RunWorkerAsync(ofd.FileName);
            }
        }

We first create an instance of the BackgroundWorker class and then wire up the DoWork and RunWorkerCompleted methods.  DoWork is the event that will fire when we call the RunWorkerAsync method.  And it will run asynchronously, in a background thread, freeing up the user interface.  Because RunWorkerAsync is launched in a background thread, control returns from the btnSelect_Click method quickly, and the UI is responsive, even while the file-read work is going on.

We also hook a handler to the RunWorkerCompleted event, which will fire when our bgWorker_DoWork method has finished doing the work.  This event, however, will execute on the original GUI thread–allowing is to update GUI elements directly within our gbWorker_RunWorkerCompleted handler.

Here’s the body of our DoWork handler.

        void bgWorker_DoWork(object sender, DoWorkEventArgs e)
        {
            FileReader2 fr = new FileReader2();

            // Filename to process was passed to RunWorkerAsync(), so it's available
            // here in DoWorkEventArgs object.
            string sFileToRead = (string)e.Argument;
            e.Result = fr.ReadTheFile(sFileToRead);
        }

Notice that we just use our earlier FileReader class to do the actual work of reading the file.  But there are two additions.

First, because this method is invoked from the BackgroundWorker object, we need to somehow get the name of the file to process.  We knew this filename back in the btnSelect_Click method and we hand it off by passing it as a parameter to RunWorkerAsync and then reading it out of the DoWorkEventArgs parameter.

Similarly, when we finish doing our work (reading the file), we need to make sure the result (# lines read) gets passed back to our RunWorkerCompleted handler.  We do this by setting the Result properly of the DoWorkEventArgs parameter.

Here’s the code for our RunWorkerCompleted event handler:

        void bgWorker_RunWorkerCompleted(object sender, RunWorkerCompletedEventArgs e)
        {
            if (e.Error != null)
            {
                MessageBox.Show(e.Error.Message);
            }
            else
            {
                int numLines = (int)e.Result;
                lblResults.Text = string.Format("We read {0} lines", numLines.ToString());
            }
        }

Here we see the other side of the e.Result handoff–we read the FileReader.ReadTheFile return value out of the RunWorkerCompletedEventArgs parameter.  We also check this parameter to see if an error occurred.

If you now run this example, you’ll see a couple of important things that work better than they did in iteration #1:

  • We now correctly see the “reading the file” label, indicating that work is in progress
  • While the file is being read, we can interact with the GUI normally

You can demonstrate the second part of this by clicking on the “Tell Me a Joke” button.  You’ll get a message box with a clever joke and you can then dismiss the dialog–all while the file read operation is still going on.

Iteration #3 – Application State and Cancel Logic

You might be tempted at this point to think that we’re done and our application has everything that it needs.  But we’re missing a few critical things.  Any time that you do work in a background worker thread, you should also consider:

  • Busy indicator — making it easy for the user to know when work is being done in the background
  • Application state — what can/can’t the user do while the work is in progress?
  • Progress indicator — give the user a visual sense of how much work is left to be done
  • Cancel logic — optionally, give the user a method to cancel the background work

Busy Indicator

Let’s start with the busy indicator.  It’s important to make it obvious to your users that something is happening in the background, and what that something is.

Application State

We have some subtle behavior in our current implementation that is probably not desirable.  Try the following:

  • Click on the Read File button and select a file, to initiate a file read operation
  • Before the read has completed, click on the button again and select a new file

You now have two file read operations running in parallel.  Is this really what we want?  Do we want to prohibit it?  If not, how do we handle the results of two different file read operations, when the operations complete?  How do we avoid mixing up the results?  How do we know which operation the results are coming from?  Is there a chance that the two operations will attempt to work on/with the same data?

For our purposes, let’s agree that we really only want to allow the user to do one operation at a time.  While one operation is in progress, a user cannot initiate another one.  We’ll modify the GUI to enforce this.

Progress Indicator

More than just indicating that some work is going on in the background, it would be nice to indicate how much work we’ve already done and how much work is left to do.  This lets a user judge how long the entire process will take.

Cancel Logic

Whenever you support doing some work on a background thread, you also need to consider whether a user might want to cancel this background activity.  Unless it’s something that happens quite quickly, it’s probably a good idea to allow a user to cancel the operation and return to the original state (no file is being read and they are able to select a new file to be read).

At this point, it’s probably a good idea to do a rough sketch of a state diagram, showing what a user can do and during what state:

Application State Diagram

Notice that we enter the “reading file” state when the user clicks the “Read File” button.  But while in this state, the user cannot press that button again–they either press the “Cancel” button, or we return to the original state when the file read operation completes.

Also note that we should be able to display a joke while in either state.  This confirms what we said earlier–the GUI won’t lock up during the file read operation.

Our Modified Example

Here’s how our file reader example works, after adding a progress indicator, cancel logic, and the ability to keep track of application state.  Here’s the new GUI during a file read operation:

Progress

Note that we now tell the user what file we’re reading and we display a progress indicator, showing how far into the read operation we are.  We also give them a Cancel button, allowing them to Cancel the operation before it completes normally.  Also notice that the Read File button is greyed out—the user can’t initiate another operation until the first one completes.

If the user lets the file read operation complete normally, they’ll see the following:

Success

Notice that when we finish reading the file, returning to the Idle state, we hide all of the progress/cancel widgets.  The Read File button is also enabled again.

If the user cancels the file read operation, they’ll see the following:

Cancelled

Again, all of the progress/cancel widgets are gone, since we’re back in the Idle state.  And the Read File button is available again.  But this time, we tell the user that they cancelled the operation.

The code for this iteration can be found in threadsafepubsub.codeplex.com, as Form3.cs and FileReader3.cs.

We added a couple of things at the top of the class–an enumeration to keep track of our state, and a class-level BackgroundWorker instance.  (We move this variable into class scope because our Cancel button will need access to the BackgroundWorker object.

    private enum AppStates { Idle, ReadingFile };

    private BackgroundWorker _worker;

Here’s our new Form3 constructor, where we now call a method to set the initial application state:

        public Form3()
        {
            InitializeComponent();

            // Set up initial state
            SetAppState(AppStates.Idle, null);
        }

Here’s the actual code for the new SetAppState function, as well as a helper function that sets visibility for several controls.

        // Set new application state, handling button sensitivity, labels, etc.
        private void SetAppState(AppStates newState, string filename)
        {
            switch (newState)
            {
                case AppStates.Idle:
                    // Hide progress widgets
                    SetFileReadWidgetsVisible(false);
                    btnSelect.Enabled = true;
                    break;

                case AppStates.ReadingFile:
                    // Display progress widgets & file info
                    SetFileReadWidgetsVisible(true);
                    lblProgress.Text = string.Format("Reading file: {0}", filename);
                    pbProgress.Value = 0;
                    lblResults.Text = "";
                    btnSelect.Enabled = false;
                    break;
            }
        }

        private void SetFileReadWidgetsVisible(bool visible)
        {
            lblProgress.Visible = visible;
            pbProgress.Visible = visible;
            btnCancel.Visible = visible;
        }

We’re basically just changing the visibility of the various progress widgets in the StatusStrip at the bottom of the form.  We also handle enabling/disabling the File Read button here.

The Click event handler for our File Read button is also slightly different. We add a line that sets the application state to indicate that a file is being read, we attach a handler to track progress, and we add an exception handler to ensure that the application state is set back to idle if anything goes wrong.

        private void btnSelect_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.CheckFileExists = true;
            ofd.CheckPathExists = true;

            if (ofd.ShowDialog() == DialogResult.OK)
            {
                FileInfo fi = new FileInfo(ofd.FileName);
                SetAppState(AppStates.ReadingFile, fi.Name);

                try
                {
                    // Set up background worker object & hook up handlers
                    _worker = new BackgroundWorker();
                    _worker.DoWork += new DoWorkEventHandler(bgWorker_DoWork);
                    _worker.RunWorkerCompleted += new RunWorkerCompletedEventHandler(bgWorker_RunWorkerCompleted);
                    _worker.WorkerReportsProgress = true;
                    _worker.WorkerSupportsCancellation = true;
                    _worker.ProgressChanged += new ProgressChangedEventHandler(bgWorker_ProgressChanged);

                    // Launch background thread to do the work of reading the file.  This will
                    // trigger BackgroundWorker.DoWork().  Note that we pass the filename to
                    // process as a parameter.
                    _worker.RunWorkerAsync(ofd.FileName);
                }
                catch
                {
                    SetAppState(AppStates.Idle, null);
                    throw;
                }
            }
        }

Note also that we have to explicitly tell the BackgroundWorker that it should support both progress and cancellation functionality.

We also now have a new event handler for the ProgressChanged event, which looks like this:

        // Get info on progress of file-read operation (% complete)
        void bgWorker_ProgressChanged(object sender, ProgressChangedEventArgs e)
        {
            // Just update progress bar with % complete
            pbProgress.Value = e.ProgressPercentage;
        }

This one is pretty simple—we just set the value of the progress bar, which runs from 0 to 100, to the reported % complete value.

Our DoWork handler has just a few changes.  Here is the new version:

        // Do work--runs on a background thread
        void bgWorker_DoWork(object sender, DoWorkEventArgs e)
        {
            // Note about exceptions:  If an exception originates anywhere in
            // this method, or methods that it calls, the BackgroundWorker will
            // automatically populate the Error property of the RunWorkerCompletedEventArgs
            // parameter that gets passed into the RunWorkerCompleted event handler.
            // So we can handle the exception in that method.

            FileReader3 fr = new FileReader3();

            // Filename to process was passed to RunWorkerAsync(), so it's available
            // here in DoWorkEventArgs object.
            BackgroundWorker bw = sender as BackgroundWorker;
            string sFileToRead = (string)e.Argument;

            e.Result = fr.ReadTheFile(bw, sFileToRead);

            // If operation was cancelled (triggered by CancellationPending),
            // we bailed out of ReadTheFile() early.  But still need to set
            // Cancel flag, because RunWorkerCompleted event will still fire.
            if (bw.CancellationPending)
                e.Cancel = true;
        }

I added a note to remind us that exceptions originating in this chunk of code (or on this thread) are automatically made available to us in the RunWorkerCompleted handler.

Notice also that we’re now passing the BackgroundWorker object into the ReadTheFile method.  We do this because we need access to it, within this message, to check for user cancellation and to report progress.

Finally, we see a piece of the cancellation infrastructure here.  Below is another code chunk to help us understand how the cancel operation works—the click handler for the Cancel button.

        private void btnCancel_Click(object sender, EventArgs e)
        {
            _worker.CancelAsync();
        }

This is pretty simple.  When the user clicks the Cancel button, we tell the BackgroundWorker object to initiate a cancel operation.  Here’s a summary of the entire cancel operation (what happens when):

  • User clicks Cancel button
  • btnCancel_Click handler invokes BackgroundWorker.CancelAsync on active worker object
  • Method doing actual work (reading file) periodically checks BackgroundWorker.CancellationPending and aborts if it sees this property set
  • Control returns to bgWorker_DoWork method
  • DoWork method checks CancellationPending property and sets DoWorkEventArgs.Cancel to true if operation was cancelled
  • BackgroundWorker.RunWorkerCompleted fires
  • We can check RunWorkerCompletedEventArgs.Cancelled, in our RunWorkerCompleted handler, to detect whether operation was cancelled

This is a little involved, but if you walk through the code, you’ll see how things work.

Finally, here is our RunWorkerCompleted event handler:

        void bgWorker_RunWorkerCompleted(object sender, RunWorkerCompletedEventArgs e)
        {
            try
            {
                if (e.Error != null)
                {
                    MessageBox.Show(e.Error.Message, "Error During File Read");
                }
                else if (e.Cancelled)
                {
                    lblResults.Text = "** Cancelled **";
                }
                else
                {
                    int numLines = (int)e.Result;
                    lblResults.Text = string.Format("We read {0} lines", numLines.ToString());
                }
            }
            finally
            {
                // State now goes back to idle
                SetAppState(AppStates.Idle, null);
            }
        }

There are just a couple of new things here.  We now check the Cancelled property and display a message if the operation was cancelled.  We also add a finally block, where we ensure that we transition back to the Idle state, whether things completed normally, the user cancelled, or there was an error.

I have one final block of code to share—the ReadTheFile method that does the actual work:

        public int ReadTheFile(BackgroundWorker bw, string fileName)
        {
            int numLines = 0;
            FileInfo fi = new FileInfo(fileName);
            long totalBytes = fi.Length;
            long bytesRead = 0;

            using (StreamReader sr = new StreamReader(fileName))
            {
                // Note: When BackgroundWorker has CancellationPending set, we bail
                // out and fall back to the _DoWork method that called us.
                string nextLine;
                while (((nextLine = sr.ReadLine()) != null) &&
                       !bw.CancellationPending)
                {
                    bytesRead += sr.CurrentEncoding.GetByteCount(nextLine);
                    numLines++;
                    int pctComplete = (int)(((double)bytesRead / (double)totalBytes)* 100);
                    bw.ReportProgress(pctComplete);
                    Thread.Sleep(10);  // ms
                }
            }

            return numLines;
        }

We’ve basically added two things here: support for cancellation and for progress reporting.

To support user-initiated cancellation, we just check to see if the operation has been cancelled, after each line in the file that we read.  The frequency with which you check for cancellation is important.  If you don’t check often enough, the application will appear to not be responding to the cancel request and the user may become frustrated.

We report progress (% complete) by invoking the ReportProgress method on the background worker.  We do this after calculating the actual progress, in terms of # bytes read so far.