We Keep Coding

C# Process.Start Causing AccessViolationException Randomly

I have been tackling this issue for 3 months now.
Error I am Getting in Native Debugging:
"Exception thrown at 0x5A222FC2 (comct123.dll) in FileReader.exe:
0xC0000005: Access violation reading location 0x0000000C."
Normal Debug:
'System.AccessVioliationException' in System.Windows.Forms.dll
My setup is really simple:
public static Form_Interface Interface;
public static void Initialize()
{
Application.SetCompatibleTextRenderingDefault(false);
Interface = new Form_Interface();
Interface.Filesdgv.DataSource = File.SortableBindingList;
Application.Run(Interface);
}
Seems simple enough, right? No.
So basically I have a simple Event that simply opens the file using Process.Start() and no matter what I do it will randomly crash with 'System.AccessVioliationException' in System.Windows.Forms.dll here:
private void Filesdgv_CellContentClick(object sender, DataGridViewCellEventArgs e)
{
Filesdgv.Invoke((Action)(() =>
{
try
{
int rowIndex = e.RowIndex;
if (rowIndex >= 0)
{
int columnIndex = e.ColumnIndex;
File file = (File)((DataGridView)sender).Rows[rowIndex].DataBoundItem;
switch (columnIndex)
{
case 0:
{
Process.Start(file.Location);
}
break;
}
}
}
catch
{
// This fking catch never works anyway.
}
}));
}
private void FileInterface_Load(object sender, EventArgs e)
{
foreach (string oCurrent in Directory.GetFiles(Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.DesktopDirectory), "Files To Check")))
if (oCurrent.EndsWith(".pdf", StringComparison.OrdinalIgnoreCase))
new File(oCurrent.Split('\\').Last(), oCurrent);
}
It doesn't matter if I am opening files/links or anything else, it still behaves in the same way.
The link and file location is a readonly field as well.
I have many other uses for reading row data and it never crashes, even if i spam click 10000 times, It will only crash randomly with Process.Start()
Things I tried:
Using BeginInvoke
Using Invoke
Not Using Invoke/BeginInvoke
Putting File Link into a string before reading it.
Using multiple Try Catch
Recoded on another machine... same results there aswell.
I tried using File.Open (either doesn't open the file or throws same error lmao)
Tried using [HandleProcessCorruptedStateExceptions], still won't catch the exception.
Dosen't matter if i click slow or fast still 1/30 chance it happens.
Tried Putting Task.Run(() => Process.Start()); you'd think that a thread will protect you from an exception? no still crashes...
File Class looks like this:
public class File
{
public static SortableBindingList<File> SortableBindingList = new SortableBindingList<File>(new List<File>());
public readonly string fileName;
public readonly string filePath;
public void AddRow()
{
Client.Interface.Invoke((Action)(() =>
{
lock (SortableBindingList)
if (!SortableBindingList.Contains(this))
SortableBindingList.Add(this);
}));
}
public string FileName
{
get
{
return fileName;
}
}
public string Location
{
get
{
return filePath;
}
}
public File(string fileName, string filePath)
{
this.fileName = fileName;
this.filePath = filePath;
AddRow();
}
}
Initalize() is called in static void Main(string[] args) btw.
There are no other threads running editing stuff or anything like that, the only thread running is the form thread. which waits for user input.
Solutions I am looking for:
Alternative Method to launch files/hyperlinks.
A way to avoid form crashing (try catch style)
Crashes even with static data!:
Other threads running although these were not started by me.

Task.Run(() =>
{
Thread.Sleep(100);
Process.Start("https://www.youtube.com");
});
This has fixed my issues, it seems that when trying to immediately run "process.start" during a click event, the GUI unfocusing + starting a new process the exact same moment causes an Exception. (Microsoft pls fix.)

Unsure of type to use in an event subscription method

I was trying to create a general function, that should describe some threads to the windows event log to different internal methods. So i wrote:
void SubScribeToEventLogEvent(string EventLog,int EventID,string Source, Action named ="myMethod")
{
string query = "*[System/Level=8] and [System/EventID=" +EventID.ToString()+"]";
EventLogQuery subscriptionQuery = new EventLogQuery(EventLog, PathType.LogName, "*[System/Level=8]");
EventLogWatcher watcher = new EventLogWatcher(subscriptionQuery);
watcher.EventRecordWritten += new EventHandler<EventRecordWrittenEventArgs>(named);
watcher.Enabled = true;
}
Then I have a few functions triggered for specific events
private void myMethod(object obj,EventRecordWrittenEventArgs arg)
{
//do something
if (arg.EventRecord != null)
{
MessageBox.Show(arg.EventRecord.Id.ToString() + arg.EventRecord.FormatDescription());
Console.WriteLine("Received event {0} from the subscription.", arg.EventRecord.Id);
Console.WriteLine("Description: {0}", arg.EventRecord.FormatDescription());
}
else
{
Console.WriteLine("The event instance was null.");
}
}
So that in my main I could simply write
SubScribeToEventLogEvent(EventlogName,200,myMethod);
However it doesn't seam to work, I am coding in .net 4.0 and I believed that Action is a reserved method word for doing such things. But I get an error (red line already under named in the first line of the code):
SubScribeToEventLogEvent(string EventLog,int EventID,string Source, Action named ="myMethod")
apparently it cann't be assigned a string value (but what else?), and if I remove the string value its still not recognized in the Eventhandler function.

Async XML Reading in Windows Phone 7

So I have a Win Phone app that is finding a list of taxi companies and pulling their name and address from Bing successfully and populating a listbox that is being displayed to users. Now what I want to do is, to search for each of these terms on Bing, find the number of hits each search term returns and rank them accordingly (a loose sort of popularity ranking)
void findBestResult(object sender, DownloadStringCompletedEventArgs e)
{
string s = e.Result;
XmlReader reader = XmlReader.Create(new MemoryStream(System.Text.UTF8Encoding.UTF8.GetBytes(s)));
String name = "";
String rName = "";
String phone = "";
List<TaxiCompany> taxiCoList = new List<TaxiCompany>();
while (reader.Read())
{
if (reader.NodeType == XmlNodeType.Element)
{
if (reader.Name.Equals("pho:Title"))
{
name = reader.ReadInnerXml();
rName = name.Replace("&","&");
}
if (reader.Name.Equals("pho:PhoneNumber"))
{
phone = reader.ReadInnerXml();
}
if (phone != "")
{
string baseURL = "http://api.search.live.net/xml.aspx?Appid=<MyAppID>&query=%22" + name + "%22&sources=web";
WebClient c = new WebClient();
c.DownloadStringAsync(new Uri(baseURL));
c.DownloadStringCompleted += new DownloadStringCompletedEventHandler(findTotalResults);
taxiCoList.Add (new TaxiCompany(rName, phone, gResults));
}
phone = "";
gResults ="";
}
TaxiCompanyDisplayList.ItemsSource = taxiCoList;
}
}
So that bit of code finds the taxi company and launches an asynchronous task to find the number of search results ( gResults ) to create each teaxicompany object.
//Parses search XML result to find number of results
void findTotalResults(object sender, DownloadStringCompletedEventArgs e)
{
lock (this)
{
string s = e.Result;
XmlReader reader = XmlReader.Create(new MemoryStream(System.Text.UTF8Encoding.UTF8.GetBytes(s)));
while (reader.Read())
{
if (reader.NodeType == XmlNodeType.Element)
{
if (reader.Name.Equals("web:Total"))
{
gResults = reader.ReadInnerXml();
}
}
}
}
}
The above snipped finds the number of search results on bing, but the problem is since it launches async there is no way to correlate the gResults obtained in the 2nd method with the right company in method 1. Is there any way to either:
1.) Pass the name and phone variables into the 2nd method to create the taxi object there
2.) Pass back the gResults variable and only then create the corresponding taxicompany object?

Right well there is a lot do here.
Getting some small helper code
First off I want to point you to a couple of blog posts called Simple Asynchronous Operation Runner Part 1 and Part 2. I'm not suggesting you actually read them (although you're welcome too but I've been told they're not easy reading). What you actually need is a couple of code blocks from them to put in your application.
First from Part 1 copy the code from the "AsyncOperationService" box, place it in new class file in your project called "AsyncOperationService.cs".
Second you'll need the "DownloadString" function from Part 2. You could put that anywhere but I recommend you create a static public class called "WebClientUtils" and put it in there.
Outline of solution
We're going to create a class (TaxiCompanyFinder) that has a single method which fires off the asynchronous job to get the results you are after and then has an event that is raised when the job is done.
So lets get started. You have a TaxiCompany class, I'll invent my own here so that the example is as complete as possible:-
public class TaxiCompany
{
public string Name { get; set; }
public string Phone { get; set; }
public int Total { get; set; }
}
We also need an EventArgs for the completed event that carries the completed List<TaxiCompany> and also an Error property that will return any exception that may have occured. That looks like this:-
public class FindCompaniesCompletedEventArgs : EventArgs
{
private List<TaxiCompany> _results;
public List<TaxiCompany> Results
{
get
{
if (Error != null)
throw Error;
return _results;
}
}
public Exception Error { get; private set; }
public FindCompaniesCompletedEventArgs(List<TaxiCompany> results)
{
_results = results;
}
public FindCompaniesCompletedEventArgs(Exception error)
{
Error = error;
}
}
Now we can make a start with some bare bones for the TaxiCompanyFinder class:-
public class TaxiCompanyFinder
{
protected void OnFindCompaniesCompleted(FindCompaniesCompletedEventArgs e)
{
Deployment.Current.Dispatcher.BeginInvoke(() => FindCompaniesCompleted(this, e));
}
public event EventHandler<FindCompaniesCompletedEventArgs> FindCompaniesCompleted = delegate {};
public void FindCompaniesAsync()
{
// The real work here
}
}
This is pretty straight forward so far. You'll note the use of BeginInvoke on the dispatcher, since there are going to be a series of async actions involved we want to make sure that when the event is actually raised it runs on the UI thread making it easier to consume this class.
Separating XML parsing
One of the problems your original code has is that it mixes enumerating XML with trying to do other functions as well, its all a bit spagetti. First function that I indentified is the parsing of the XML to get the name and phone number. Add this function to the class:-
IEnumerable<TaxiCompany> CreateCompaniesFromXml(string xml)
{
XmlReader reader = XmlReader.Create(new StringReader(xml));
TaxiCompany result = new TaxiCompany();
while (reader.Read())
{
if (reader.NodeType == XmlNodeType.Element)
{
if (reader.Name.Equals("pho:Title"))
{
result.Name = reader.ReadElementContentAsString();
}
if (reader.Name.Equals("pho:PhoneNumber"))
{
result.Phone = reader.ReadElementContentAsString();
}
if (result.Phone != null)
{
yield return result;
result = new TaxiCompany();
}
}
}
}
Note that this function yields a set of TaxiCompany instances from the xml without trying to do anything else. Also the use of ReadElementContentAsString which makes for tidier reading. In addition the consuming of the xml string is much smoother.
For similar reasons add this function to the class:-
private int GetTotalFromXml(string xml)
{
XmlReader reader = XmlReader.Create(new StringReader(xml));
while (reader.Read())
{
if (reader.NodeType == XmlNodeType.Element)
{
if (reader.Name.Equals("web:Total"))
{
return reader.ReadElementContentAsInt();
}
}
}
return 0;
}
The core function
Add the following function to the class, this is the function that does all the real async work:-
private IEnumerable<AsyncOperation> FindCompanies(Uri initialUri)
{
var results = new List<TaxiCompany>();
string baseURL = "http://api.search.live.net/xml.aspx?Appid=<MyAppID>&query=%22{0}%22&sources=web";
string xml = null;
yield return WebClientUtils.DownloadString(initialUri, (r) => xml = r);
foreach(var result in CreateCompaniesFromXml(xml))
{
Uri uri = new Uri(String.Format(baseURL, result.Name), UriKind.Absolute);
yield return WebClientUtils.DownloadString(uri, r => result.Total = GetTotalFromXml(r));
results.Add(result);
}
OnFindCompaniesCompleted(new FindCompaniesCompletedEventArgs(results));
}
It actually looks pretty straight forward, almost like synchonous code which is the point. It fetchs the initial xml containing the set you need, creates the set of TaxiCompany objects. It the foreaches through the set adding the Total value of each. Finally the completed event is fired with the full set of companies.
We just need to fill in the FindCompaniesAsync method:-
public void FindCompaniesAsync()
{
Uri initialUri = new Uri("ConstructUriHere", UriKind.Absolute);
FindCompanies(initialUri).Run((e) =>
{
if (e != null)
OnFindCompaniesCompleted(new FindCompaniesCompletedEventArgs(e));
});
}
I don't know what the initial Uri is or whether you need to paramatise in some way but you would just need to tweak this function. The real magic happens in the Run extension method, this jogs through all the async operations, if any return an exception then the completed event fires with Error property set.
Using the class
Now in you can consume this class like this:
var finder = new TaxiCompanyFinder();
finder.FindCompaniesCompleted += (s, args) =>
{
if (args.Error == null)
{
TaxiCompanyDisplayList.ItemsSource = args.Results;
}
else
{
// Do something sensible with args.Error
}
}
finder.FindCompaniesAsync();
You might also consider using
TaxiCompanyDisplayList.ItemsSource = args.Results.OrderByDescending(tc => tc.Total);
if you want to get the company with the highest total at the top of the list.

You can pass any object as "UserState" as part of making your asynchronous call, which will then become available in the async callback. So in your first block of code, change:
c.DownloadStringAsync(new Uri(baseURL));
c.DownloadStringCompleted += new DownloadStringCompletedEventHandler(findTotalResults);
to:
TaxiCompany t = new TaxiCompany(rName, phone);
c.DownloadStringAsync(new Uri(baseURL), t);
c.DownloadStringCompleted += new DownloadStringCompletedEventHandler(findTotalResults);
Which should then allow you to do this:
void findTotalResults(object sender, DownloadStringCompletedEventArgs e)
{
lock (this)
{
TaxiCompany t = e.UserState;
string s = e.Result;
...
}
}
I haven't tested this code per-se, but the general idea of passing objects to async callbacks using the eventarg's UserState should work regardless.
Have a look at the AsyncCompletedEventArgs.UserState definition on MSDN for further information.

What's wrong with my application ---- Size was 0, but I expected 46806 !

I'm a C# programmer.
Now, I'm using the ICSharpCode.SharpZipLib.dll to create a zip file in my current project. But it occurs to me that when I click the button at the SECOND TIME to execute a function to create a zip file, the application will throw an exception, friendly and seriously told me that "Size was zero, but I expected 46086".
I'm so confused that I want to know why? When I click the button at the first time, I can do it successfully without any error.
My related codes are as follows:
internal void ThreadProc()
{
try
{
ZipHelper.CreateZip(backupZipFile, Constants.HomeConstant, true);
// do other things
}
}
The CreateZip() function's realization is as follows:
public static void CreateZip(string zipFileName, string sourceDirectory, bool recurse)
{
FastZip zip = new FastZip();
if (File.Exists(zipFileName))
{
File.Delete(zipFileName);
}
zip.CreateZip(zipFileName, sourceDirectory, true, "");
}
Now, I will show you the recursive calling process:
Call method "UpdateAppAsync" in "ActiveCheckManager" class
public void UpdateAppAsync(string masterConfig)
{
this.masterConf = masterConfig;
Thread actualThread = new Thread(new ThreadStart(UpdateApp));
actualThread.IsBackground = true;
actualThread.CurrentCulture = Thread.CurrentThread.CurrentCulture;
actualThread.CurrentUICulture = Thread.CurrentThread.CurrentUICulture;
actualThread.Start();
}
Call the UpdateApp function asynchronously, in the UpdateApp method, it will only call the UpdateDetail function simply.
private void UpdateDetail(string masterConfig, string category)
{
IUpdate worker = new HP.ActiveCheckLocalMode.UpdateEngine.UpdateManager();
worker.UpdateApp(masterConf);
}
The worker.UpdateApp will call UpdateDetail(string, UpdateCategory) only.
private void UpdateDetail(string masterConfig, UpdateCategory cat)
{
UpdateThread updateThread = new UpdateThread(this, cat);
updateThread.MasterConfig = masterConfig;
updateThread.ThreadProc();
}
That is the calling process. When I click the update button second time, it will throw an exception, can you help me? Thank you very much.

Has the first task thread finished before you start the second time?
I would imagine that File.Delete() and some items in the SharpZipLib to not respond nicelly to multithreadingly zip the same folder simultaneously to the same file.
Promote that " UpdateThread updateThread " as a private member of the "ActiveCheckManager" class, then check if it is already running from a previous click before creating a new thread.

Unique EventId generation

I'm using the Windows Event Log to record some events. Events within the Windows Event Log can be assigned a handful of properties. One of which, is an EventID.
Now I want to use the EventId to try and group related errors. I could just pick a number for each call to the logging method I do, but that seems a little tedious.
I want the system to do this automatically. It would choose an eventId that is "unique" to the position in the code where the logging event occurred. Now, there's only 65536 unique event IDs, so there are likely to be collisions but they should be rare enough to make the EventId a useful way to group errors.
One strategy would be to take the hashcode of the stacktrace but that would mean that the first and second calls in the following code would have generate the same event ID.
public void TestLog()
{
LogSomething("Moo");
// Do some stuff and then a 100 lines later..
LogSomething("Moo");
}
I thought of walking up the call stack using the StackFrame class which has a GetFileLineNumber method. The problem with this strategy is that it will only work when built with debug symbols on. I need it to work in production code too.
Does anyone have any ideas?

Here is some code you can use to generate an EventID with the properties I describe in my question:
public static int GenerateEventId()
{
StackTrace trace = new StackTrace();
StringBuilder builder = new StringBuilder();
builder.Append(Environment.StackTrace);
foreach (StackFrame frame in trace.GetFrames())
{
builder.Append(frame.GetILOffset());
builder.Append(",");
}
return builder.ToString().GetHashCode() & 0xFFFF;
}
The frame.GetILOffset() method call gives the position within that particular frame at the time of execution.
I concatenate these offsets with the entire stacktrace to give a unique string for the current position within the program.
Finally, since there are only 65536 unique event IDs I logical AND the hashcode against 0xFFFF to extract least significant 16-bits. This value then becomes the EventId.

The IL offset number is available without debug symbols. Combined with the stack information and hashed, I think that would do the trick.
Here's an article that, in part, covers retrieving the IL offset (for the purpose of logging it for an offline match to PDB files--different problem but I think it'll show you what you need):
http://timstall.dotnetdevelopersjournal.com/getting_file_and_line_numbers_without_deploying_the_pdb_file.htm

Create a hash using the ILOffset of the last but one stack frame instead of the line number (i.e. the stack frame of your TestLog method above).

*Important: This post focuses at solving the root cause of what it appears your problem is instead of providing a solution you specifically asked for. I realize this post is old, but felt it important to contribute. *
My team had a similar issue, and we changed the way we managed our logging which has reduced production support and bug patching times significantly. Pragmatically this works in most enterprise apps my team works on:
Prefix log messages with the "class name"."function name".
For true errors, output the captured Exception to the event logger.
Focus on having clear messages as part of the peer code review as opposed to event id's.
Use a unique event id for each function, just go top to bottom and key them.
when it becomes impractical to code each function a different event ID, each class should just just have a unique one (collisions be damned).
Utilize Event categories to reduce event id reliance when filtering the log
Of course it matters how big your apps are and how sensitive the data is. Most of ours are around 10k to 500k lines of code with minimally sensitive information. It may feel oversimplified, but from a KISS standpoint it pragmatically works.
That being said, using an abstract Event Log class to simplify the process makes it easy to utilize, although cleanup my be unpleasant. For Example:
MyClass.cs (using the wrapper)
class MyClass
{
// hardcoded, but should be from configuration vars
private string AppName = "MyApp";
private string AppVersion = "1.0.0.0";
private string ClassName = "MyClass";
private string LogName = "MyApp Log";
EventLogAdapter oEventLogAdapter;
EventLogEntryType oEventLogEntryType;
public MyClass(){
this.oEventLogAdapter = new EventLogAdapter(
this.AppName
, this.LogName
, this.AppName
, this.AppVersion
, this.ClassName
);
}
private bool MyFunction() {
bool result = false;
this.oEventLogAdapter.SetMethodInformation("MyFunction", 100);
try {
// do stuff
this.oEventLogAdapter.WriteEntry("Something important found out...", EventLogEntryType.Information);
} catch (Exception oException) {
this.oEventLogAdapter.WriteEntry("Error: " + oException.ToString(), EventLogEntryType.Error);
}
return result;
}
}
EventLogAdapter.cs
class EventLogAdapter
{
//vars
private string _EventProgram = "";
private string _EventSource = "";
private string _ProgramName = "";
private string _ProgramVersion = "";
private string _EventClass = "";
private string _EventMethod = "";
private int _EventCode = 1;
private bool _Initialized = false;
private System.Diagnostics.EventLog oEventLog = new EventLog();
// methods
public EventLogAdapter() { }
public EventLogAdapter(
string EventProgram
, string EventSource
, string ProgramName
, string ProgramVersion
, string EventClass
) {
this.SetEventProgram(EventProgram);
this.SetEventSource(EventSource);
this.SetProgramName(ProgramName);
this.SetProgramVersion(ProgramVersion);
this.SetEventClass(EventClass);
this.InitializeEventLog();
}
public void InitializeEventLog() {
try {
if(
!String.IsNullOrEmpty(this._EventSource)
&& !String.IsNullOrEmpty(this._EventProgram)
){
if (!System.Diagnostics.EventLog.SourceExists(this._EventSource)) {
System.Diagnostics.EventLog.CreateEventSource(
this._EventSource
, this._EventProgram
);
}
this.oEventLog.Source = this._EventSource;
this.oEventLog.Log = this._EventProgram;
this._Initialized = true;
}
} catch { }
}
public void WriteEntry(string Message, System.Diagnostics.EventLogEntryType EventEntryType) {
try {
string _message =
"[" + this._ProgramName + " " + this._ProgramVersion + "]"
+ "." + this._EventClass + "." + this._EventMethod + "():\n"
+ Message;
this.oEventLog.WriteEntry(
Message
, EventEntryType
, this._EventCode
);
} catch { }
}
public void SetMethodInformation(
string EventMethod
,int EventCode
) {
this.SetEventMethod(EventMethod);
this.SetEventCode(EventCode);
}
public string GetEventProgram() { return this._EventProgram; }
public string GetEventSource() { return this._EventSource; }
public string GetProgramName() { return this._ProgramName; }
public string GetProgramVersion() { return this._ProgramVersion; }
public string GetEventClass() { return this._EventClass; }
public string GetEventMethod() { return this._EventMethod; }
public int GetEventCode() { return this._EventCode; }
public void SetEventProgram(string EventProgram) { this._EventProgram = EventProgram; }
public void SetEventSource(string EventSource) { this._EventSource = EventSource; }
public void SetProgramName(string ProgramName) { this._ProgramName = ProgramName; }
public void SetProgramVersion(string ProgramVersion) { this._ProgramVersion = ProgramVersion; }
public void SetEventClass(string EventClass) { this._EventClass = EventClass; }
public void SetEventMethod(string EventMethod) { this._EventMethod = EventMethod; }
public void SetEventCode(int EventCode) { this._EventCode = EventCode; }
}

Thanks for the idea of hashing the call stack, I was going to ask that very same question of how to pick an eventId.
I recommend putting a static variable in LogSomething that increments each time it is called.

Now I want to use the EventId to try
and group related errors.
You have filters in event viewer so why (Go to find ? You have 65536 unique event IDs too.
Or rather use log4net or something ??
just my ideas....