|
This is what happens when you don't use parameterized commands! I know that I've given you code samples before for this. Creating a SQL command by concatenating your params together is arcane and should not be used in .NET. If you use a SqlCommand, for example, then you should add SqlParameters to its SqlCommand.Parameters collection property and use @paramname in your query without quotes. The command will do what's necessary. See the documentation for the SqlParameter (or OleDbParameter, or any of the others) in the .NET Framework SDK.
There are also a lot of other benefits to using parameterized commands, such as easy batch processing when you save the parameters as variables and simply update their Value property, then re-execute the command.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Hi,
I'm trying to populate a TreeView with info I'm gathering via a PeekCompleted callback for a MessageQueue. My PeekCompleted method, which is "void static", does not have any trouble populating a ListBox but the compiler complains that I must use Control.Invoke() from the callback to use TreeNode.Add().
Here are my questions:
1) Why?
2) Should I use Control.Invoke() or do this another way?
All I'm trying to do is build a TreeView using data I collected from a callback method. What are my other options?
TIA,
Matt
|
|
|
|
|
Yes you should use Control.Invoke. This invokes the call on the thread on which the control was created, which is important. You should do this for your ListView as well. Modifying the control from a different thread causes problems in the message queue. The technical details get down into what is encapsulated by the .NET Framework (much of it, anyway): Win32 APIs.
To use Invoke, take a look at this sample code to safely add a TreeNode in the property thread:
internal void SafeAdd(TreeNodeCollection nodes, TreeNode node)
{
if (nodes == null || node == null)
throw new ArgumentNullException(nodes == null ? "nodes" : "node");
if (treeView1.InvokeRequired)
{
Delegate d = new AddTreeNodeHandler(nodes.Add);
treeView1.Invoke(d, new object[] {node});
}
else nodes.Add(node);
}
private delegate int AddTreeNodeHandler(TreeNode node);
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Hi guys,
I am trying to write an application which when installed will keep an watch over the network and generate report about it.
Can any one tell me how can I detect a SOAP call. I want specifically to detect and ensure that a SOAP call has been made and also extract out info like : by which applicaiton(the application and machine name) ,to which server, etc.
Thanks in advance...
regards,
Aryadip.
Cheers !! and have a Funky day !!
|
|
|
|
|
If you're talking about a network filter, you treat it like any other HTTP request or response. Extract the body and check if it uses the SOAP namespace. If so, you can extract additional information from the HTTP headers or from the WS-Routing or WS-Addressing elements, if available in the SOAP header.
If you just want to log information about SOAP calls, you can add a client or server sink to the chain that uses information from the IMessage that's passed to it to log information about the call.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
I made 2 events on System.Windows.Form named as Click and DoubleClick.The sample code for them is
private void Form1_Click(object sender, System.EventArgs e)
{
MessageBox.Show("Click");
}
private void Form1_DoubleClick(object sender, System.EventArgs e)
{
MessageBox.Show("DClick");
}
but when i clicks(whether single or double) only the Form1_Click event occur.I want that on single click Form1_Click occur and on double click Form1_DoubleClick
occur.In other words i want to perform 2 different actions on single and double-click,please give any sort of help for this problem.
mughalali
|
|
|
|
|
I guess you can go about with it in the following manner...
private int clickCount = 0;
private void Form1_MouseDown(object sender, System.Windows.Forms.MouseEventArgs e)
{
clickCount ++;
}
private void Form1_DoubleClick(object sender, System.EventArgs e)
{
if(clickCount>=2)
{
clickCount=0;
MessageBox.Show("double click");
}
}
Cheers !! and have a Funky day !!
|
|
|
|
|
The proper solution is to use a timer, with the timeout value set to SystemInformation.DoubleClickTime. If the timer elapses before the DoubleClick event handler fires (which stops the timer), then execute your code in the timer's callback. This was the common approach that even Microsoft suggested back in the ol' MFC days (even was in a tutorial installed with VS if I remember correctly).
Using a counter as the other person suggested could lead to problems keeping the counter in sync with other events.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Hi,
I am monitoring 2 MessageQueues (MSMQ) via the PeekCompleted callback.
It seems to be working but I get a lot of duplicate msgs. By duplicate I mean I receive 2+ messages with exact same System.Messaging.Message.Id.
Is this normal? Doesn't seem efficient to me and I wonding if I'm doing something wrong?
TIA,
Matt
|
|
|
|
|
hi, I want my text box to be displayed in a multi line style
and not a long line (even when the focus is not on the cell)
how do i do this (i already word wraped and multilined it)?
|
|
|
|
|
First of all, realize that the TextBox isn't used to display the data - only to edit it. That being said, you need to extend DataGridColumnStyle (or you could even extend DataGridTextBoxColumn and just modify the TextBox to be multi-line when editing) and override Paint method to draw your text on multiple lines. The Graphics class and supporting classes (like StringFormat) provide easy support for doing this.
You'll also need to override GetMinimumHeight to provide the height (you should cap it at some reasonable limit) you'll need to draw your multi-line text (which you can determine using Graphics.MeasureString).
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Thanks for the pointers , I'll try them out.
|
|
|
|
|
I really hope someone can help me with this, I've looked everywhere.
I have a class that I want to save to a file, and it was working fine using a static method within the actual class:
static public void save(string fileName, CMyClass c)
{
IFormatter formatter = new BinaryFormatter();
Stream stream = new FileStream(fileName, FileMode.Create, FileAccess.Write, FileShare.None);
formatter.Serialize(stream, c);
stream.Close();
}
However, I want to use this class remotely, but when I call this method, using a transparent proxy, I get an exception when formatter.Serialize runs.
It says "Attempted to call a method declared on type System.Runtime.Remoting.Messaging.IMethodMessage on an object which exposes CMyClass."
I tried writing the save method as non static, and passing the this pointer to Serialize, but I get another error, which doesn't surprize me.
Does anyone know a simple solution for saving my class?
|
|
|
|
|
CMyClass must be remotable, first of all. This means it must derive from MarshalByRefObject or MarshalByValueComponent and be attributed with the SerializableAttribute (optionally implementing ISerializable if you want to explicitly control serialization).
A few things about design, though. First, never prefix classes with "C". There's a naming convention for every framework - like the .NET Framework - for a reason: to provide consistent naming through the framework. Microsoft has done a pretty good job of this throughout the .NET FCL (Framework Class Library) with only a few exceptions. See Naming Guidelines[^] in the .NET Framework SDK for more information.
Also, if an instance of CMyClass is what you're saving, then your method should be an instance method (not static). What you're doing works, but it's not a very good object-oriented design because the instance is what you're saving. Statics are usually reserved for properties, methods, and fields (like constants) that apply to the class, not an instance of the class.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Firstly, thanks for the information about naming conventions for the framework, I guess I'm just stuck in my ways because of many years of coding in C++ and using a C in front of classes.
MyClass does inherit from MarshalByRefObject, and it is attributed with the SerializableAttribute, sorry I missed them.
I do want to use an instance method to save the instance. Like so:
public void save(string fileName){
try
{
IFormatter formatter = new BinaryFormatter();
Stream stream = new FileStream(fileName, FileMode.Create, FileAccess.Write, FileShare.None);
formatter.Serialize(stream, this);
stream.Close();
}
catch(Exception e){
MessageBox.Show(e.Message);
}
}
Which would be ideal, however when this is run an exception is thrown with the following message:
"The type System.Runtime.Remoting.ServerIdentity in Assembly mscorlib, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089 is not marked as serializable."
Everything stored in MyClass is marked as serializable, it saves fine when using my static method, but not when called remotely. All other remote calls in MyClass work ok.
|
|
|
|
|
10+ years of C/C++, about 5 with Perl, 6 with Java and now about 3 with .NET: each uses a different naming convention. It's just something you have to get used to.
It's not that it wouldn't save when using a static method, just that it's not a very good OO design.
As far as the error goes (which would've been helpful), some member of your class uses the ServerIdentity Type, which isn't serializable. I suggest implementing ISerializable and serializing only those fields you want to save. There's no documentation for this Type so without digging through the IL I couldn't tell you want it does (though the name gives me a few ideas), but chances are that whatever field Type defines such a property is probably something you don't need to save.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Thanks for your help Heath.
Through debugging I found out that ServerIdentity seems to be used by MarshalByRef, so when my code trys to serialize it, it won't allow it. From the .net help files I found this:
"CAUTION As a general design pattern, it would be unusual for a class to be both marked with the serializable attribute and extend MarshalByRefObject. Developers should think carefully about the possible serialization and remoting scenarios when combining these two characteristics..... For more details about serialization of classes that extend MarshalByRefObject see RemotingSurrogateSelector"
There is not a great deal of information about using RemotingSurrogateSelector, on the internet, or the .net help. So I have decided to do it a different way.
The class 'MyClass' is actually called SupremeHome and is the main class in my home automation program I'm writing, it holds collections of various objects. Each object I want to save will now implement ICloneable. SupremeHome will have a (non static) method called Save that will create a new SupremeHome object based on itself, using ICloneable.Clone(), then write the whole thing to disk with the given file name. I have tested it a bit and it works ok.
|
|
|
|
|
A RemotingSurrogateSelector selects surrogates - or class that serialize other classes that aren't serializable themselves. So, if some class isn't marked with the SerializationAttribute (and hence, not serializable regardless of whether or not is extends MarshalByRefObject), then you add an ISerializationSurrogate implementation associated with a Type you want to serialize.
What you're doing right now - though it may work - is very inefficient. As I mentioned before, you should just implement ISerializable (forget what MSDN says in this case - Microsoft does it throughout the FCL and I use it throughout a very large N-tier application I architected at work that uses .NET Remoting and a heck of a lot of sinks) and not serialize what you don't need to pass to the other application domain.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
I see what you mean about inefficiency, and as implementing ISerializable requires about the same amount of coding, I am now doing it that way. Thanks again for your help.
|
|
|
|
|
According to msdn and all .NET 1.1 docs, the ToolBar.BackColor property is intended to be public. Every attempt to access it in code, however, indicates that it was accidentally implemented as private. I have tried deriving new subclass from ToolBar to override the property and have tried casting up the chain to Control.BackColor. Neither route has worked. Any ideas on how to change the color of a ToolBar?
|
|
|
|
|
It's not provide, but it is hidden from the code editor and property grid using the BrowsableAttribute and the EditorBrowsableAttribute. Why? Because it does nothing. This is actually common throughout the System.Windows.Forms members. You cannot explicitly set the background color of a ToolBar because the Toolbar common control that the ToolBar class encapsulates does not support background colors. It uses the Control color defined in your display settings. The same goes for a background image.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
Hello,
I am trying to figure out a solution for hosting an application in IE, but runs on the client machine (to access printer configuration). I want to use C#, b/c the client has the .NET runtime, otherwise the answer would be, develop an ActiveX control (ATL).
Does anyone know how to write a C# app (winform) that is hosted in Internet Explorer?
For example, the user browses to http://someserver/somepage.aspx , which then downloads the C# app, and is hosted in IE. The app then has control to access printers on the local machine and local network (the same way windows does)
Thanks. 
R.Bischoff
.NET, Kommst du mit?
|
|
|
|
|
|
Thanks for your helpful information.
How would you access the PrintDialog control from the control you are hosting in IE ?
Thanks.
R.Bischoff
.NET, Kommst du mit?
|
|
|
|
|
Same as you would using any other form. It's just a control. If you're trying to access the print dialog for IE, that's not so simple and involves dispatch interfaces and client-side scripting.
You must also have the right code access permissions, which are documented in the class documentation for the classes you want to use.
My article does discuss code access security a little, so you need to author your control with this in mind. You shouldn't just assume you'll have FullTrust permissions - that's poor design.
Microsoft MVP, Visual C#
My Articles
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
