|
A little background.
I'm playing with BusinessObjects and Factory(s) which contain collections of BusinessObjects.
My business objects all derive from a BusinessObjectBase class.
I have an IBusinessObjectBase interface, which the BusinessObjectBase implements.
My factory objects all derive from a FactoryBase class.
The FactoryBase class is generic (FactoryBase<T> where T : IBusinessObjectBase)
I have an IFactoryBase interface, which the FactoryBase impliments.
public interface IFactoryBase<T> where T : IBusinessObjectBase
Everything works fine (I'm not sure if the model is a little simplistic maybe)
Now.
I'm wanting to create a status bar, which displays a few properties from the current BusinessObject, and the currenct Factory (Like data errors, read only info etc.).
So I created a test to see how to go about this, but encountered a problem.
customerFactory cFact = new customerFactory();
if (cFact is IFactoryBase<IBusinessObjectBase> ) TestContext.WriteLine("IFactoryBase supported");
if (cFact is IComponent) TestContext.WriteLine("IComponent supported");
if (((IFactoryBase<IBusinessObjectBase> )cFact).IsClean) TestContext.WriteLine("Factory is clean");
customerFactory is derived from FactoryBase, and does indeed support the IFactoryBase interface.
public class customerFactory : FactoryBase<customer>
The customer object is derived from BusinessObjectBase, and supports IBusinessObjectBase
The first line, fails even though my Factory does support the interface.
The second line succeeds (IComponent is indeed supported)
And the third line fails to cast.
Many thanks.
|
|
|
|
|
stovesy wrote: My factory objects all derive from a FactoryBase class.
The FactoryBase class is generic (FactoryBase<t> where T : IBusinessObjectBase)
I have an IFactoryBase interface, which the FactoryBase impliments.
public interface IFactoryBase<t> where T : IBusinessObjectBase
You cant just implement the interface and expect it to magically become one!
FactoryBase should be declared as:
class FactoryBase<t> : IFactoryBase<t> where T : IBusinessObjectBase ...</t></t>
|
|
|
|
|
As stated above, the FactoryBase class, does implement the IFactoryBase interface, I've just not show that in code.
Cheers.
|
|
|
|
|
stovesy wrote: I've just not show that in code.
Then dont complain at my answer. If it indeed is a FactoryBase derived object it has to be IFactoryBase then, and you have made a mistake somewhere.
|
|
|
|
|
You misunderstand - I'm not complaining about your answer, just clarifying things.
Thanks for your interest.
|
|
|
|
|
stovesy wrote: IFactoryBase<ibusinessobjectbase>
IFactoryBase<customer> is needed then.
|
|
|
|
|
We're getting to the crux of the matter.
What you suggest does work, but I don't know the type of the generic class, only that it derives from IBusinessObjectBase
So when I try ...(customerFactory is IFactoryBase<ibusinessobjectbase>).... it fails.
Thanks for your time
|
|
|
|
|
stovesy wrote: We're getting to the crux of the matter.
Indeed 
stovesy wrote: What you suggest does work, but I don't know the type of the generic class, only that it derives from IBusinessObjectBase
So when I try ...(customerFactory is IFactoryBase).... it fails.
Why not make IFactoryBase non-generic, and expose the non-generic properties/methods you need to access? Else create a non-generic IFactoryBaseBase interface
|
|
|
|
|
I have changed as you suggest, and it does work.
I still wonder why I can't do the generics thing - seems more elegant.
One for the back-burner maybe.
Thanks for your help.
|
|
|
|
|
Generics having full covariance is a common expectation, and I find the easiest way to explain it is to give a counter-example.
Say you have IBusinessObject< T > { IList< T > ListOfChildren; }.
So with our IBusinessObject< Customer > foo, we can .ListOfChildren and get a IList< Customer >.
Good so far.
Then if we could cast foo to IBusinessObject< object > - if we did .ListOfChildren we could have a IList< object >.
Still everything is peachy until we take that IList< object > and break type-safety by doing .Add(someRandomObject). We obviously can't actually Add any object to that IList, it must be a Customer. So IBusinessObject< object > is not an equivalent type.
IIRC the compiler can be nice to us and provides covariance in the case of a "read-only" generic type, identified by the type parameter being used for read-only properties, and return values only. IEnumerable< T > is probably one of these.
So in summary, IFactoryBase< Customer > cannot be cast to IFactoryBase< IBusinessObjectBase >, they are different types - it is specialized in dealing with Customers, not any IBusinessObjectBase. You might be able to get away with it if IFactoryBase< T > only uses T on the return side of it's members (like IEnumerable< T >), but I'm not sure if this behaviour is fully documented/guaranteed. (Check for me and reply?  )
|
|
|
|
|
Thanks Mark.
The way of looking at it that sits well in my brain - I was seeing a generic type definition as two types: clear now that this is not the case.
You do learn something every day.
|
|
|
|
|
Hi,
I have developed a winform application which monitors files using FileSystemWatcher. This works fine.
Now I have to develop it in Windows service.
Can you see why the following line does not compile in windows service please?
//in winform I am using this which works fine...
_watcher.SynchronizingObject = this;
//this line does not compile in windows service. Please note I changed the word this to the service class name i.e. FileMonitor
_watcher.SynchronizingObject = FileMonitor;
The error is: FileMonitor.FileMonitor is a type but it's used as a variable.
|
|
|
|
|
arkiboys wrote: //in winform I am using this which works fine...
_watcher.SynchronizingObject = this;
//this line does not compile in windows service. Please note I changed the word this to the service class name i.e. FileMonitor
SynchronizingObject is not needed outside a WinForm project. It is simply there to prevent cross threaded calls. You can safely ignore that in a Window Service.
|
|
|
|
|
you need to assign the actual instantiated variable and not the class.
IE
FileMonitor FM = new FileMonitor();
_watcher.SynchronizingObject = fileMonitor;
Regards
Michael
|
|
|
|
|
I wanted to use the Regex.Split function to split an incoming string up. The problem is that the string parts are separated by a pipe | but the pipe is used inside the Split function to split different separators (e.g. if you want to split on dash and space you would write it as " |-"). So when I use "|" the string splits into it's characters. I tried using "\|" but this is considered an Unrecognized Escape Character and doesn't compile.
Does anyone know if this is possible in another way?
|
|
|
|
|
You have to encode the characters in the expression twice. First you have to encode the pipe to put it in the regular expression, then you have to encode the regular expression to put it in a literal string in the code.
Use \\| in a regular literal string, or \| in an @ delimited string.
Example:
string pattern = "\\|";
or
string pattern = @"\|";
Despite everything, the person most likely to be fooling you next is yourself.
|
|
|
|
|
Thanks, that solved the problem
|
|
|
|
|
public string deleteDB(string strSelect)
{
try
{
cmd.Connection = objConn;
cmd.CommandText = strSelect;
cmd.ExecuteScalar();
}
catch (SqlException eException)
{
Console.Write("Error in connection : " + eException.Message);
return -10;
//return strSelect;
}
}
|
|
|
|
|
Well, did you read the error message ? Your function returns a string, and you're tring to return -10. Which is both stupid ( what does -10 mean ? ) and illegal ( an int is not a number ). The logical thing to return here is probably a bool. Why bother returning the string that was passed in ?
Where does strSelect come from ? This looks silly to me from the get go. You're just taking SQL as a string and calling it, this should not exist at all, but if it did, it should have a generic name, there is not guarentee it will delete anything.
There is so much wrong with this, you should probably go back to basics and do some reading. But, to fix the problem at hand, read your error messages and believe them. You said you were returning a string. The compiler won't turn an int to a string, so return a string, or change the return type. And, return something when the code doesn't blow up, too.
Christian Graus
Please read this if you don't understand the answer I've given you
"also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
|
|
|
|
|
Christian Graus wrote: an int is not a number
You mean string 
|
|
|
|
|
*sigh* yes, I did...
Christian Graus
Please read this if you don't understand the answer I've given you
"also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
|
|
|
|
|
your trying to return an int in a method that needs to return a string... youl have to return "-10" in your catch block...
and use PRE tags when posting code please
Harvey Saayman - South Africa
Junior Developer
.Net, C#, SQL
you.suck = (you.passion != Programming)
|
|
|
|
|
-10 is a int not a string
|
|
|
|
|
Hello everyone,
I need some help here, maybe lots.
Want a quick way to upload images to tinypic dot com from within windows.
Its a free image hosting site. Hope I am not breaking any rules here by typing the site name.
I can think of doing this with right-click context menu.
I am a newbie in programming. This is what I have come up with so far.
Regedit : HKEY_CLASSES_ROOT\jpegfile\shell
Can add a KEY as "Upload to TinyPic". Right click "Upload to TinyPic" and add another key "command".
Right click any jpg/jpeg image file and you'll see the option "Upload to TinyPic" in the context menu.
Now what command to use so that whenever I click on that, the pic will get uploaded to the site and the uploaded site opens up showing the links or something like that.
A friend of mine told me, it can be done thru programming. So thats the reason, I am here. I don't know any programming, so give me codes and stuffs to do that. Just let me know something easy a normal person like me can use.
Kindly help me out.
Thanks a lot.
|
|
|
|
|
You're not in Kansas anymore. I recommend asking your friend to write this for you. No-one here is going to do it. Your command would be the path to the program you write to upload the files.
Christian Graus
Please read this if you don't understand the answer I've given you
"also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
