|
Let's take this question ... an interesting one ... into the "Twilight Zone:"
public IEnumerable<IGenericInterface> GetStuffX()
{
return (new List<UsingGenericInterface>()) as IEnumerable<IGenericInterface>;
}
public IEnumerable<UsingGenericInterface> GetStuffY()
{
return (new List<IGenericInterface>()) as IEnumerable<UsingGenericInterface>;
}
Why do these compile ? 
best, Bill
"It is the mark of an educated mind to be able to entertain a thought without accepting it." Aristotle
|
|
|
|
|
List<T> implements IEnumerable<T>.
|
|
|
|
|
BillWoodruff wrote: Why do these compile ?
MSDN:
"The as operator is like a cast operation. However, if the conversion is not possible, as returns null instead of raising an exception." link to the "as" operator in msdn[^]
so, what happens is that the compiler doesn't care if your conversion always returns null.
i'm trying to instal the .net sdk in this computer (don't have admin privileges) to disassemble a dll and see what code the compiler generates for List<IGenericInterface> and for List<UsingGenericInterface> but i suppose that the List<IGenericInterface> implements IList<IGenericInterface> and IEnumerable<IGenericInterface> while List<UsingGenericInterface> implements IList<UsingGenericInterface>, IEnumerable<UsingGenericInterface> and, given that UsingGenericInterface implements IGenericInterface, the List<UsingGenericInterface> also implements IEnumerable<IGenericInterface>
if this is the case, then i think it's a design flaw of the compiler, but i wont take precipitated conclusions.
|
|
|
|
|
My apologies, Jasnoch, and Sentenryu; I do know why they compile, and how "as" functions compared to "casting to Type;" in this case I just thought it would be more interesting to "round out" the possibilities listed.
It was not my intent to divert attention from the main focus of your thread.
best, Bill
"It is the mark of an educated mind to be able to entertain a thought without accepting it." Aristotle
|
|
|
|
|
The short answer is because T on IEnumerable<T>[^] is declared covariant, and the one on IList<T> isn't.
I don't really understand why lists and other collection classes don't have covariant type parameters as well; I guess there must be some problems with implementing it.
|
|
|
|
|
Hi, I'm a beginning C# programmer.
I'm playing with a RichTextBox, applying various colors to the lines. I've noticed that there is a property of the RTB called TextLength. In my testing RTB.TextLength is ALWAYS equal to RTB.Text.Length. So why even have such a property? Everyone knows you can get a String.Length, so why the redundancy?
Originally I thought RTB.TextLength might have something to do with formatting characters vs visible characters but this is apparently not the case.
|
|
|
|
|
Oh nevermind, I think I see now. RTB.Text probably pulls out all that text to a separate copy in memory, THEN takes the length. RTB.TextLength probably doesn't have that overhead. Am I right?
|
|
|
|
|
Yes, TextLength is nothing but the length of the Text property and should be used when you want to obtain the length of the text, and not get at the text for any other purpose.
/ravi
|
|
|
|
|
When i read the image it shows the incorrect output in other language not in english. here is the code for read the image..
MODI.Document modDoc = new MODI.Document();
modDoc.Create(@"h:\123.jpg");
modDoc.OCR(MODI.MiLANGUAGES.miLANG_ENGLISH, true, true);
MODI.Image img = (MODI.Image)modDoc.Images[0];
MODI.Layout layout = img.Layout;
String OCRText = layout.Text;
modDoc.Close(false);
for (int i = 0; i < lines.Length; i++)
{
textBox1.Text = textBox1.Text + lines[i];
textBox1.Text += '\n';
}
Plz anybody tell me the solution of this problem .........
|
|
|
|
|
Please do not post the same question twice.
Please delete the other one, before someone answers it.
Ideological Purity is no substitute for being able to stick your thumb down a pipe to stop the water
|
|
|
|
|
When i read the image it shows the incorrect output in other language not in english. here is the code for read the image..
MODI.Document modDoc = new MODI.Document();
modDoc.Create(@"h:\123.jpg");
modDoc.OCR(MODI.MiLANGUAGES.miLANG_ENGLISH, true, true);
MODI.Image img = (MODI.Image)modDoc.Images[0];
MODI.Layout layout = img.Layout;
String OCRText = layout.Text;
modDoc.Close(false);
for (int i = 0; i < lines.Length; i++)
{
textBox1.Text = textBox1.Text + lines[i];
textBox1.Text += '\n';
}
Plz anybody tell me the solution of this problem .........
|
|
|
|
|
I would like clipping a line but do not know, you can help me?
Such LiangBarsky algorithm used to cut the line is drawn by Bresenham.
|
|
|
|
|
Hello Everyone
I'm trying to create a small project in my laptop. Could someone please tell me once this application is up and running is there a possibility to access this application via bluetooth on to my mobile phone.
Please feel free to correct if my question is not at clear explonation...
Kind regards
Roni
|
|
|
|
|
Define "access"?? That's such a generic term it really has no meaning.
1. Do you want to run the application on the phone? Not going to happen.
2. Do you want something on the phone to communicate with your application? Sure. You just need to define what protocols you're going to use and what you want to transfer over this link.
3. Do you want this application to send some kind of data to the phone? Sure, see number 2.
|
|
|
|
|
Greetings.
I am currently working on a little project for fun; I am trying to have a webcam work in C#, probably by using directshow.
But I am so clueless!
I don't understand the code enough to write my own from scratch so I am seeking someone who masters C# to show Me what I need.
This would work best on MSN; ADEPofArts@gmail.com
Thank You for your help.
|
|
|
|
|
You might get a better response if you post this in this forum:
Collaboration and Beta Testing[^]
The difficult we do right away...
...the impossible takes slightly longer.
|
|
|
|
|
Ah! Didn't see that one. Thanks.
|
|
|
|
|
You should also remove your email from your post so you don't get a ton of spam.
I wasn't, now I am, then I won't be anymore.
|
|
|
|
|
For some reason got on a Tree class kick (tree, btree, bstree, avltree, etc). So I started porting an old MFC version I wrote 10yrs ago and decided instead to redesign it.
Concept I'm going for at this point is I'll have a base Node object, a NodeCollection and a Tree class. The btree will derive from tree, the bstree will derive from btree and the avl tree will derive from bstree.
Right now... everything looks something like:
public class Node<T>
{
public Node(T value)
{
_value = value;
}
public NodeCollection Children
{
get
{
return _children;
}
}
}
public class NodeCollection : Collection<T>
{
}
public class Tree
{
public Node Root
{
get
{
return _root;
}
set
{
_root = value;
}
}
}
Obviously, I snipped out the non important stuff for brevity, but one thing I'm not liking so far is that is that you need to do something like tree.Root.Value, tree.Root.Children[4].Value, etc to get to the real value. I.e. everything works on a Node object instead of the value object. I guess you do need a Node object so you can have somewhere to get the children from. Perhaps if I add some methods to NodeCollection to find Nodes by Value it will make it better?
Any suggestions, I'm not liking something about this design, but I'm not quite sure what it is.
My C++ version sort of hid all that by using the Windows POSITION thingy... which I guess is pretty much the same this as just returning a Node... Hmmm...
|
|
|
|
|
SledgeHammer01 wrote: Perhaps if I add some methods to NodeCollection to find Nodes by Value it will
make it better?
Hmmmm. One thought occurs to me - what happens if the value you want to return is in the root node? You'd have to check that value first, and then call the NodeCollection check. If the check is part of the Node instead, and that knows how to iterate over child nodes then that would simplify the end code somewhat (granted it would make the internal logic slightly more complex). Actually, a simple extension to this would allow you to put it into the Tree class itself, and then that would trigger the Node/NodeCollection checks - this way, the end dev only has to use something like:
var node = myTree.FindNodeByValue(myValue); Thinking this through though, there is a particular smell in place in that you actually have to supply a full copy of value that can be used in an equality test - perhaps you could extend to include an equality test as well (as a Predicate for instance), so that complex objects could be retrieved off one of their properties (useful if you want to search for a particular ID for instance).
As for the retrieving by child position, this code seems very reminiscent of manipulating XML. Again, things could be made simpler by providing methods at the Tree level which would retrieve the relevant child without having to resort to the Root node in external code. As this is all working off a Node object, you could have a method that looks something like this (completely untested, so there may be some kinks in it):
public Node<T> GetNodeByPosition(int position)
{
return this.Root.Children[position];
}The Node class could have a similar method, which would allow the dev to chain results together.
Anyway, these are just some thoughts off the top of my head - I'd really have to sit down and work through this some more.
[Edit]Not sure why your post was 1-voted. I don't see anything inherently wrong with the post.
modified 5-Jan-12 8:50am.
|
|
|
|
|
I was just a little uneasy about exposing the Node objects... but the more I think about it, maybe thats OK. In my MFC version, I was exposition a POSITION which was just a pointer to the node structure, but the node structure was hidden from the user. The user navigated through the tree by the various POSITION based methods. GetLeft(POSITION pos), GetRight(POSITION pos), GetValue(POSITION pos), etc. I guess in the C# world, just returning the Node object is considered better then trying to hide it from the user. It just seems a bit odd to me that the tree is working based on Node objects rather then the value objects. What I mean is...
if I have List<int> lst;
lst works based on ints, there is no intermediate structure hiding things.
if I have Dictionary<int, int=""> dict;
dict itself stores a KeyValuePair<int, int="">, but the way they designed it, you are usually just accessing it by dict[6] = whatever; The only time the KeyValuePair object comes into play is in the foreach loop I think.
|
|
|
|
|
Pete O'Hanlon wrote: Hmmmm. One thought occurs to me - what happens if the value you want to return
is in the root node? You'd have to check that value first
This part just made sense when I was looking over the code again . Hmm... I wasn't thinking about having a method to find a node in the ENTIRE tree, just the current level, but then that would be funky... you'd have to write code like:
if (node.Value == blah || node.Children.FindNodeByValue(blah))
{
}
but it may not be that useful to search the entire tree because you might have a lot of items and just want to search the current branch... now I'm maybe thinking of having flags (node only, children only, all, recursive) to specifying how to search.
Pete O'Hanlon wrote: [Edit]Not sure why your post was 1-voted. I don't see anything inherently wrong
with the post.
Lol, Yeah, I stopped paying attention to votes on here. I've told Chris MANY times that the voting on here is stupid .
|
|
|
|
|
Certainly agree that whoever down-voted this post is an idiot.
I've been working with TreeViews a long time, and one question your "spec" brings to my mind is: do you define a Tree as one "root" node, from which "all springs forth," or, is a TreeView a "container" that has a "root level" collection of TreeNodes. From a classic "computer science" point of view: probably the first definition, yes. In the WinForms MS provided TreeView, and the third-party TreeView I use (a tool which makes the WinForms standard TV look, and perform, like sandpile mudspatter, imho, from Lidor Systems), the second answer would apply.
This may be a "moot" point to raise here: after all you could fill any container with a collection of "views of " TreeViews based on the "one root" model.
At my suggestion, Lidor created a dynamically updated "flat list" collection of Nodes in later versions of their TreeView, which I find quite handy, and, of course, you can write an extension to create a "flat list" of a WinForms TV (excellent examples of this based on ideas by Eric Lippert and Wes Dyer on StackOverFlow that use a stack to make creating such a flat list very efficient).
Some other stray thoughts on things I would like a TreeView able to allow:
1. the ability to declare a strongly typed TreeView: something like:
StrongTypedTreeView myStrongTV = new StrongTypedTreeView(NodeValueType: typeof(double)); 2. the ability to declare either a strongly typed 'Tag property, or a strongly typed Collection that functioned like the typical 'Tag property does now. The idea being to reduce casting of the Tag back into whatever object form you want it in.
best, Bill
"It is the mark of an educated mind to be able to entertain a thought without accepting it." Aristotle
|
|
|
|
|
I haven't written a line of C# in anger for a while but would something like this be a bit cleaner
public class Node<T> :Tree<T>
{
public T Value { get; set; }
}
public class Tree<T>
{
public List<Node<T>> Children { get; set; }
}
"You get that on the big jobs."
|
|
|
|
|
+5 Thanks, Rob: that's a very "evocative" design idea that has me headed to VS Studio right now to try it out, since it seems (to me naive brain) that architecture could be the key to a "strongly typed" TreeView.
best, Bill
"It is the mark of an educated mind to be able to entertain a thought without accepting it." Aristotle
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
