It's difficult to tell what is being asked here. This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form. For help clarifying this question so that it can be reopened, visit the help center.
Closed 10 years ago.
I know that few similar topics have been created on StackOverflow (e.g there and there).
I have a well-known problem - user unhandled Exceptions thrown in Visual Studio in Windows 7 64bit are not handled by IDE debugger, so debugger does not break in suitable code line. Because I don't want to catch all exceptions by enabling "Throwing" checkBoxes in Tools->Exceptions... menu, I have tried to use Microsoft article solution.
Applying MS solution caused that, the situation has changed, but VS debugger still does not run correctly.
Currently, when new Exception is thrown, I see system internal error message and then VS debugger will correctly stops on error line BUT only for less than one second and application exits...
Do you have some others solution to resolve this bug? Programming with so called silent exceptions are very uncomfortable...
EDIT: I hope, that now my question is less ranty...
In my case none of registry modifications were needed. According to this topic and Redd answer, I modified my Program.cs file in such manner:
static void Main()
{
try
{
System.Windows.Forms.Application.SetUnhandledExceptionMode(UnhandledExceptionMode.CatchException);
System.Windows.Forms.Application.ThreadException += new System.Threading.ThreadExceptionEventHandler(OnGuiUnhandedException);
AppDomain.CurrentDomain.UnhandledException += OnUnhandledException;
var form = new MainForm();
form.ShowDialog();
}
catch (Exception e)
{
HandleUnhandledException(e);
}
finally
{
// Do stuff
}
}
private static void HandleUnhandledException(Object o)
{
// TODO: Log it!
Exception e = o as Exception;
if (e != null)
{
}
}
private static void OnUnhandledException(Object sender, UnhandledExceptionEventArgs e)
{
HandleUnhandledException(e.ExceptionObject);
}
private static void OnGuiUnhandedException(object sender, System.Threading.ThreadExceptionEventArgs e)
{
HandleUnhandledException(e.Exception);
}
Now my debugger works OK, so it stops on unhandled exceptions and doesn't stop on handled exceptions. It looks like Microsoft solves silent exceptions problem in SP1 for Windows 7 64 bit at OS level, but enforcing correct working of VS still needs some user/programmer actions.
Related
Closed. This question needs debugging details. It is not currently accepting answers.
Edit the question to include desired behavior, a specific problem or error, and the shortest code necessary to reproduce the problem. This will help others answer the question.
Closed 3 years ago.
Improve this question
I have a winforms application that is installed on multiple computers. Most of the time it works perfectly but for a small subset of users the application fails to launch. The only resolution I have found for this issue is to reinstall the application on the users machine.
I have included screenshots below showing the application working after a successful launch and also a screenshot showing what the user sees when the application fails
Normal Launch:
Failed Launch:
When the application fails, the startup form does not get rendered at all. On the users desktop there is nothing visible at all and the program is not outside of any visible area.
If anyone could provide answers or insight into the following questions it would be much appreciated.
What could cause this problem?
Windows or program related?
How could this be fixed?
I have included code snippets from the startup form below
Starting code:
private static void Main()
{
Application.EnableVisualStyles();
Application.SetCompatibleTextRenderingDefault(false);
Application.SetUnhandledExceptionMode(UnhandledExceptionMode.CatchException);
Application.Run(new Timelord());
}
public Timelord()
{
this.InitializeComponent();
this.BringToFront();
this.Focus();
// Displays a date and gets the version of the program
lblDate.Text = DateTime.Now.ToShortDateString();
Version version = System.Reflection.Assembly.GetExecutingAssembly().GetName().Version;
if (ApplicationDeployment.IsNetworkDeployed)
{
lblVersion.Text = string.Format("v{0}", ApplicationDeployment.CurrentDeployment.CurrentVersion.ToString(4));
}
// Loads the comboboxes for selection
this.loadComboUser();
this.loadComboCompany();
this.loadComboTick();
}
I think what is happening is that there is an error being thrown in your Timelord constructor under certain conditions. Since Timelord is the "startup" object for your application, a failure to create its instance properly would cause serious problems. Here is what I would recommend doing to identify those conditions, and to eliminate the issue with the form only being partially created.
I am assuming based on your comment about the program reading from a database that one or more of the following methods perform data access calls to a database
this.loadComboUser();
this.loadComboCompany();
this.loadComboTick();
You typically want to avoid method calls, ESPECIALLY data access calls in a constructor. There are many reasons for this that I won't list here, but this other stackoverflow article explains some of them.
Is it OK to put a database initialization call in a C# constructor?
To correct these issues, implement an eventhandler for the load event and move all of your Timelord constructor code into the Load event handler. The Form.Load event fires after the constructor is complete but before a form is displayed for the first time.
https://learn.microsoft.com/en-us/dotnet/api/system.windows.forms.form.load?view=netframework-4.7.2
Here is how I would recommend restructuring your Timelord object.
public Timelord()
{
this.InitializeComponent();
}
Private Sub Timelord_Load(sender As Object, e As EventArgs) Handles MyBase.Load
{
Try
{
this.BringToFront();
this.Focus();
// Displays a date and gets the version of the program
lblDate.Text = DateTime.Now.ToShortDateString();
Version version = System.Reflection.Assembly.GetExecutingAssembly().GetName().Version;
if (ApplicationDeployment.IsNetworkDeployed)
{
lblVersion.Text = string.Format("v{0}", ApplicationDeployment.CurrentDeployment.CurrentVersion.ToString(4));
}
// Loads the comboboxes for selection
this.loadComboUser();
this.loadComboCompany();
this.loadComboTick();
}
Catch(Exception ex)
{
MessageBox.Show($"The following error occurred in the Timelord constructor {Environment.NewLine}{ex.Message}")
}
}
Making this change will allow the Timelord constructor to completely create the object, then the Load event will run and load any data into the UI. This way, if an error occurs, you will have at least completely created the Timelord Form and can catch the error.
What could cause this problem?
Your startup object (Timelord()) throwing an error in the constructor, therefore not properly creating object.
Windows or program related?
Program related
How could this be fixed?
Separating your Forms logic so that the only code in the constructor is your instantiation logic.
Implementing Try/Catch blocks to trap errors
I'm currently working through a project that is using HP's Fortify SCA tool to catch security issues in the code base. I'm having a bit of issue determining the best approach to correctly handling JDBC resources.
The code I have at the minute looks like this;
try (Connection conn = new DatabaseService().getConnection();
PreparedStatement ps = conn.prepareStatement(query);) {
ps.setString(1, mString);
try (ResultSet rs = ps.executeQuery();) {
while (rs.next()) {
...Do logic...
}
} catch (SQLException e) {
e.printStackTrace();
}
} catch (SQLException e){
e.printStackTrace();
}
}
The problem is that Fortify will flag this code stating that if an exception were to happen in the nested try statement then the reference to conn and ps will be lost and they won't be properly closed. Is fortify correct to flag this or is it a false positive? From what I understand try-with-resource should always close their resource but perhaps this doesn't always happen when they're nested like this.
I've scoured other related questions and blogs around the internet but I haven't been able to get any definitive proof on this.
The most documented solution that's always safe in this situation is to not use try-with-resource and wrap each resource with a try-catch in both the catch and finally blocks of a broader try-catch statement. However, I'd rather avoid this because it's horribly verbose.
Thanks in advance!
Edit: So I realized I've left something out of the code when I was re-writing it into SO. The original catch blocks had a System.exit(1); statement in them (bad practice I know). That would mean that if an exception was thrown in the nested try-with-resource then Fortify would be correct to say the conn and ps would not be properly closed.
Thanks for the replies, without the System.exit(1); all resources in this situation will be closed properly and I've selected the answer indicating that.
Using try-with-resource is always supported on Java 7 and higher, no matter tooling is sitting on top of it.
So, if this code compiles (meaning you are on Java7+), you can safely ignore any warnings as they are indeed false positives. The auto-closing resource contract is guaranteed for JRE classes.
Now, if you decide to write you own resource that implements AutoCloseable then it's up to you to make sure that the close() method actually closes the resource =)
The Fortify Java translator may never have been updated with this Java 7+ construct. You should contact Fortify Technical Support and submit the test case. The analysis is incorrect.
Further, you should mark this and other identical findings "Not an Issue" and move on with your life.
Is it possible to modify a solution, and use envdte tools, from a command-line project ?
I have an add-in that modifies a solution. But... the changes are required for over a hundred projects... So I'd like to make a c# program that has the same logic, only it iterates through all solution files.
The add-in starts with
EnvDTE.Solution solution = (EnvDTE.Solution)application.Solution;
where DTE2 application is passed from the add-in...
How can I get the same solution, which then I query for projects...
From a separate program, that will only know the solutionPath ?
Is it possible to open the solution, process it, and close it - to move on to the next solution ?
Microsoft gives this example http://msdn.microsoft.com/en-us/library/envdte._solution.open(v=vs.100).aspx
But I don't know what dte is in the context...
Thank you...
VS 2010
edit: I did what the answer below suggests.
Slightly modified, using the link:
http://msdn.microsoft.com/en-us/library/ms228772(v=vs.100).aspx
Thank you
Yes you can. You just need to activate an instance using the COM CLSID for Visual Studio. An example is below. It actually creates a solution and adds two projects to it but the same initialization applies when opening an existing solution.
A couple of caveats:
Mind the COM threading model. The code created from the console app template is sufficient:
[STAThread]
static void Main()
If you have a powerful VS extension like ReSharper installed, you might be better off suspending it if you don't need it for the VS automation. ReSharper had VS commands that control it.
Console.WriteLine("Opening Visual Studio");
var dte = (DTE)Activator.CreateInstance(Type.GetTypeFromProgID("VisualStudio.DTE.10.0",true),true);
Console.WriteLine("Suspending Resharper");
dte.ExecuteCommand("ReSharper_Suspend");
Console.WriteLine("Working with {0}, {1} edition", dte.FullName, dte.Edition);
dte.SuppressUI = true;
dte.UserControl = false;
foreach (var solution in mySolutionInfoList)
{
try
{
dte.Solution.Create(solution.directory, solution.name);
dte.Solution.AddFromTemplate(csharpTemplatePath, solution.directory + "ClassLibrary1", "ClassLibrary1");
dte.Solution.AddFromTemplate(vcTemplatePath, solution.directory + "Win32Dll", "Win32Dll");
Directory.CreateDirectory(solution.directory); // ensure directory exists. Otherwise, user will be asked for save location, regardless of SupressUI value
dte.Solution.Close(true);
Console.WriteLine();
}
catch (Exception e)
{
Console.Error.WriteLine(e);
}
}
Console.WriteLine("Resuming Resharper");
dte.ExecuteCommand("ReSharper_Resume");
try
{
dte.Quit();
}
catch (Exception e)
{
Console.Error.WriteLine(e);
}
I'm working off of the following example to implement editing of a cell in my grid when the cell is clicked:
http://demos.telerik.com/aspnet-ajax/grid/examples/dataediting/editondblclick/defaultcs.aspx
I'd like it to work just like in the example, but based on a single-click. I can't get it to work as I keep getting the following error buried away in Telerik.Web.UI.WebResource:
0x800a139e - Microsoft JScript runtime error: Sys.WebForms.PageRequestManagerServerErrorException: The string was not recognized as a valid format.
If anyone can lend any assistance, I will you owe you my first-born, as I am pulling my hair out trying to get this to work.
Thank you
Initially, the error was here but it didn't seem essential:
protected void detailsGrid_ItemCreated(object sender, GridItemEventArgs e)
{
if (e.Item is GridDataItem && e.Item.IsInEditMode)
{
((e.Item as GridDataItem)["detailsGridMonthOneCol"].Controls[0] as RadNumericTextBox).Width = Unit.Pixel(50); // ArgumentOutOfRangeException - Specified argument was out of the range of valid values
}
}
detailsGridMonthOneCol is the name of the column I double-clicked. This didn't seem essential, so I commented it out and that's when I got the following error:
Unhandled exception at line 15, column 16485 in http://localhost:63919/Telerik.Web.UI.WebResource.axd?_TSM_HiddenField_=RadScriptManager1_TSM&compress=1&_TSM_CombinedScripts_=;;System.Web.Extensions,+Version=4.0.0.0,+Culture=neutral,+PublicKeyToken=31bf3856ad364e35:en-US:10a773fc-9022-49ec-acd6-8830962d8cbb:ea597d4b:b25378d2;Telerik.Web.UI,+Version=2012.2.815.40,+Culture=neutral,+PublicKeyToken=121fae78165ba3d4:en-US:bd12f06c-2391-4523-868e-0017245d9792:16e4e7cd:ed16cbdc:f7645509:24ee1bba:e330518b:1e771326:8e6f0d33:6a6d718d:58366029:4b09f651:a2c5be80:874f8ea2:c172ae1e:f46195d3:9cdfc6e7:2003d0b8:c8618e41:e4f8f289
0x800a139e - Microsoft JScript runtime error: Sys.WebForms.PageRequestManagerServerErrorException: The string was not recognized as a valid format.
The code is buried away but here's where the exception gets thrown:
var e=this._get_eventHandlerList().getHandler("endRequest"),b=false;if(e){var c=new Sys.WebForms.EndRequestEventArgs(a,f?f.dataItems:{},d);e(this,c);b=c.get_errorHandled()}if(a&&!b)throw a}
In your Script Manager add a handler to the OnAsyncPostBackError="myScriptManager_AsyncPostBackError" and in code behind just put one breakpoint on the open curly brace of the method.
protected void myScriptManager_AsyncPostBackError(object sender, AsyncPostBackErrorEventArgs e)
{ // breakpoint this line.
}
doing this, probaly, this breakpoint will be hit and you could debug your code, and inspect who was thwrowing the exception.
This can help, but, the only way to help you, in fact, is if you provide the full source code. I suggest you to create another project, isolate the code that you want to work, and publish this code on github, ftp, etc.
Please, post your code and i will help.
The code is not really buried away. Javascript is showing you this error. However. the error is happening on the server side (Sys.WebForms.PageRequestManagerServerErrorException)
Check the Event Viewer (start => Run => eventvwr) it will show you more details of the error.
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 11 years ago.
Grails can be a bit of a bear to debug with its long stack dumps. Getting to the source of the problem can be tricky. I've gotten burned a few times in the BootStrap.groovy doing "def foo = new Foo(a: a, b: b).save()", for example. What are your favorite tricks for debugging Grails apps?
Some general tips:
Clear stacktrace.log, do grails run-app, then open stacktrace.log in a viewer (I prefer less stacktrace.log on linux)... once in your viewer, search for .groovy and .gsp... that generally brings you to what you actually care about.
When a stacktrace refers to a line number in a GSP file, you should open that view in a browser with ?showSource in the query string, i.e. http://localhost:8080/myProject/myController/myAction?showSource... this shows the compiled GSP source, and all GSP line numbers in the stacktrace refer to the compiled GSP, not the actual GSP source
Always, always, always surround your saves with at least some minimal error handling.
Example:
try {
if(!someDomainObject.save()) {
throw new Exception ("Save failed")
}
} catch(Exception e) {
println e.toString()
// This will at least tell you what is wrong with
// the instance you are trying to save
someDomainObject.errors.allErrors.each {error ->
println error.toString()
}
}
Beyond that, a lot of it just comes down to recognizing stacktraces and error messages... a lot of the time, Grails is incredibly unhelpful in the error messages it gives you, but you can learn to recognize patterns, like the following:
Some of the hardest errors to make sense of are because you didn't run grails clean or grails upgrade... to avoid these problems, I always use the following on the command line to run grails: grails clean; yes | grails upgrade; grails run-app
If the error has to do with duplicate definitions of a class, make sure that you declare the package the class belongs to at the top of the class's file
If the error has to do with schema metadata, connection, socket, or anything like that, make sure your database connector is in lib/, make sure your permissions are correct both in DataSource.groovy and in the database for username, password, and host, and make sure that you know the ins and outs of your connector's version (i.e. mysql connector version 5.1.X has a weird issue with aliases that may require you to set useOldAliasMetadataBehavior=true on the url in DataSource.groovy)
And so on. There are a lot of patterns to learn to recognize.
To add to Chris King's suggestion on save, I wrote a reusable closure:
Closure saveClosure = { domainObj ->
if(domainObj.save())
println "Domain Object $domainObj Saved"
else
{
println "Errors Found During Save of $domainObj!"
println domainObj.errors.allErrors.each {
println it.defaultMessage
}
}
}
Then you can just use it everywhere and it will take care of error reporting:
def book = new Book(authorName:"Mark Twain")
saveClosure(book)
Additionally, I use the debug plugin - it allows extra logging, and I added tag to the bottom of my main - that gives me a view of all the variables in session / request.
Runtime Logging plugin allows to enable logging at runtime.
While writing this answer, P6SPY plugin also seems like it could be useful - it logs all statements your app makes against the database by acting as a proxy.
Grails Console is also useful. I use it to interactively poke around and experiment with some code, which also comes in handy during debugging.
And of course, being able to step through Debugger is sweet. I switched to IntelliJ IDEA since it has the best Grails / Groovy support.
I once asked an experienced groovy developer about how he effectively debugged his applications. His answer:
I write tests!
And he has a very good point: If your code has sufficient unit and integration tests, you will hardly ever need to debug anything. Plus you get to say smug things like that to your fellow developers...
For Grails:
Unit Testing
Functional Testing
Really excellent grails app testing developerWorks article
To log exceptions with GrailsUtil.
try{
...
}catch (Exception e){
log.error("some message", GrailsUtil.sanitize(e))
...
}
More info about sanitize.
I'm not sure if this can be done out-of-the-box, but in webapps I find it useful to have a "who am I?" facility in the various view files.
The idea is to emit a message into the rendered HTML, to identify the fragment. This is especially true when I am encountering an app for the first time.
In Grails, I do this with a custom tag. For example, consider list.gsp for a Student:
<g:debug msg="student list" />
Here is the code:
class MiscTagLib {
def debug = { map ->
if (grailsApplication.config.grails.views.debug.mode == true) {
def msg = map['msg']
out << "<h2>${msg}</h2><br/>"
}
}
}
The key is that you can leave those tags in there, if desired, as they only appear in when the mode is enabled in Config.groovy:
grails.views.debug.mode=true
adding this code To the Bootsrap.groovy:init will overwrite the save method and execute some other code as well, printing out error messages in this case.
class BootStrap {
def grailsApplication
def init = {servletContext ->
grailsApplication.domainClasses.each { clazz ->
clazz.clazz.get(-1)
def gormSave = clazz.metaClass.getMetaMethod('save')
clazz.metaClass.save = {->
def savedInstance = gormSave.invoke(delegate)
if (!savedInstance) {
delegate.errors.each {
println it
}
}
savedInstance
}
def gormSaveMap = clazz.metaClass.getMetaMethod('save', Map)
clazz.metaClass.save = { Map m ->
def savedInstance = gormSaveMap.invoke(delegate, m)
if (!savedInstance) {
delegate.errors.each {
println it
}
}
savedInstance
}
def gormSaveBoolean = clazz.metaClass.getMetaMethod('save', Boolean)
clazz.metaClass.save = { Boolean b ->
def savedInstance = gormSaveBoolean.invoke(delegate, b)
if (!savedInstance) {
delegate.errors.each {
println it
}
}
savedInstance
}
}
...
}
hope that helps someone :)
(i know its not really DRY)
ref: http://grails.1312388.n4.nabble.com/How-to-override-save-method-on-domain-class-td3021424.html
Looking at the source code! This has saved me so many times now! And now that the code is hosted at GitHub it's easier than ever. Just press "t" and start typing to find the class that you're looking for!
http://github.com/grails/grails-core
Here's some tricks collected by #groovymag from Grails people in twitter:
http://blog.groovymag.com/2009/02/groovygrails-debugging/
For simple applications I use println statement.It is very very easy trick.For complex applications use debug mode in intellij idea.