I was looking for an easy and consistent way of creating properties for classes. I sometimes find that it can be a long drawn out process creating 5 to 100 properties for database classes. Here is a quick example of what I did.
You will need the following references to make this example work.
- System.Diagnostics;
- System.Reflection;
- System.Collections;
public class ErrorLogRecord
{
Hashtable _hsh = new Hashtable();
public DateTime DateOfOccurance
{
get { return (DateTime?)_hsh[GetName()] ?? DateTime.Now; }
set { _hsh[GetName()] = value; }
}
public String ErrorText
{
get { return (String)_hsh[GetName()]; }
set { _hsh[GetName()] = value; }
}
public String GetName()
{
StackTrace stackTrace = new StackTrace();
StackFrame stackFrame = stackTrace.GetFrame(1);
MethodBase methodBase = stackFrame.GetMethod();
return methodBase.Name.Replace("set_", "").Replace("get_", "");
}
}
While this method is probably not perfect and could be argued there is a better way, this seems to satisfy the need. I would welcome alternatives.
Alternate Reading: Shortcut To Creating Properties in C-Sharp (revisited)
What’s wrong with
public class SomeClass
{
public string SomeStringProperty { get; set; }
}
The Compiler automatically creates the backing field for you and you get fully working property.
Plus you can still do
public class SomeClass
{
public string SomeStringProperty { get; private set; }
}
Which makes the setter only visible to the internals of the class 🙂
There is nothing wrong with it. It is a great shortcut, which I have used for simple classes.
However; I find that there are many times that I need a dictionary, hashtable or other to store the properties for further manipulation. As an example it there are plenty of examples how to turn a dictionary or hastable into an array that can be passed into a DataRow or other objects with a single line of code. Using a single line of code to take my properties and put them into an array and then into DataRow or other is very appealing to me.
I had met another developer that has a similar approach to mine and had the same outcome, but it was slightly different again. When you are presented with a wall, you find ways to get over it and this was my way and that doesn’t mean it is the only way. I always have my ear to the ground for other solutions.
I think it just seems a bit of a faff to me, I would probably write my classes normally, and if I really need to turn the class into a dictionary or the like I would probably write an extension method to convert it to one. Probably a generic extension method that uses reflection to get the values and property names out. Saves all the casting, reflection name lookups and boxing until it’s needed.
Something like: https://gist.github.com/1264422
Of course, that doesn’t make what I do any more right than yours, and I’ll keep you solution in mind in case I come across something in my travels that will benefit from it.
Thanks for the Link to your example code. I do like that.
I don’t believe anyone is every completely right as there are many ways for doing the same thing. Sometimes it is a matter of opinion. I am going to convert one of my classes to do what you are suggesting and see what happens.
One of the things I like to do is benchmark tests to see how things look in memory usage, CPU and overall speed. I will find this to be an interesting experiment.
This is a great example of collaborating using social media and blogs. Thanks for sharing.