|
interstingly, something like this will result in a NullReferenceException:
string s = null;
string anotherString = s.Trim();
So my best guess is, that whenever using the assignment operator "=" something magic (;)) happens inside the .NET CLR.
There's also a discussion about this topic on google groups:
Topic 1[^]
Topic 2[^]
modified 12-Sep-18 21:01pm.
|
|
|
|
|
The NullReferenceException is because you are calling a method (Trim) on a null reference. There is no magic, you can assign null to references, but you obviously can't call methods or access properties/fields on a null reference.
|
|
|
|
|
The new keyword has to be used when creating a new instance of a string. However, most of the time you don't create any strings directly in your code.
The most common situation is the assignment of a literal string:
string s = "asdf";
In this case, the literal string "asdf" already exists as a constant in the assembly, so you are actually only assigning the reference of an already existing string object.
In most other situations you are calling a method that creates an instance of a string, so the actual creation is not in your code. Example:
string t = s.Substring(0, 1);
The Substring method creates a new instance of the String class and returns the reference, so you are only assigning the return value in your code.
You only use the new keyword when you create a string directly in your code:
string u = new string('*', 42); // creates a string containing 42 asterisks.
Despite everything, the person most likely to be fooling you next is yourself.
|
|
|
|
|
This is the best answer I got. Thank you so much. I need to think the C#-way.
![Rose | [Rose]](https://www.codeproject.com/script/Forums/Images/rose.gif)
// "In the end it's a little boy expressing himself." Yanni
while (I_am_alive)
{
cout<<"I love to do more than just programming.";
}
|
|
|
|
|
Greeeg wrote: This is a quite interesting question, actually. Since String is a class, you'd usually expect that you have to create a new instance when assigning a value to it (at least that's what you have to do with "normal" classes).
A string reference works just like a reference to any other class. You don't use the new keyword when assigning a reference, it's used to create an instance of a class. If the instance already exists, you just assign it's reference to the variable, and that is what happens most of the time when you are using a string.
Greeeg wrote: I guess that the ldstr opcode in MSIL does have a special meaning for the .NET String keyword
When the ldstr instruction is compiled, the string is created as a constant in the assembly. The runtime code created for the ldstr instruction just gets the reference of that string constant.
Despite everything, the person most likely to be fooling you next is yourself.
|
|
|
|
|
Guffa wrote: A string reference works just like a reference to any other class. You don't use the new keyword when assigning a reference, it's used to create an instance of a class. If the instance already exists, you just assign it's reference to the variable, and that is what happens most of the time when you are using a string.
This is not entirely true. string is a special class. It is indeed a reference type but it behaves like a value type.
If it was a reference type like any other the following should return true but it doesn't.
<code>public static bool Test()
{
string s1 = "hello";
string s2 = s1;
s1 = "world";
return s1.Equals(s2);
}</code>
Learn from the mistakes of others, you may not live long enough to make them all yourself.
|
|
|
|
|
BadKarma wrote: This is not entirely true. string is a special class. It is indeed a reference type but it behaves like a value type.
There is nothing special about the string class. It may seem like it behaves like a value type, but actually it doesn't.
BadKarma wrote: If it was a reference type like any other the following should return true but it doesn't.
It's a reference type just like any other, and just as with any other reference type, replacing a reference to an object neither changes the content of the object nor any other reference to the object.
This is what the code is doing:
1. Declare a string reference called "s1".
2. Get the reference to the "hello" literal string.
3. Store the reference in the "s1" variable.
- The s1 variable now contains a reference to "hello".
4. Declare a string reference called "s2".
5. Get the reference from the "s1" variable.
6. Store the reference in the "s2" variable.
- The s2 variable now contains a reference to "hello".
7. Get the reference to the "world" literal string.
8. Store the reference in the "s1" variable.
- The s1 varaible now contains a reference to "world".
9. Compare the contents of the strings that s1 and s2 reference.
- As s1 references "world" and s2 references "hello", the comparison returns false.
Despite everything, the person most likely to be fooling you next is yourself.
|
|
|
|
|
If you must do something, use string.Empty as a temporary value.
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: If you must do something
Christian Graus wrote: use string.Empty as a temporary value
Will this cause the new to be called?
// "In the end it's a little boy expressing himself." Yanni
while (I_am_alive)
{
cout<<"I love to do more than just programming.";
}
|
|
|
|
|
Nope, it simply changes your variable's reference to refer to an empty string.
modified on Wednesday, June 11, 2008 12:56 PM
|
|
|
|
|
S. Senthil Kumar wrote: Nope, it simply changes your variable's reference to refer to an empty string.
Which means if the variable is null, it news a string and assigns it to the variable.
// "In the end it's a little boy expressing himself." Yanni
while (I_am_alive)
{
cout<<"I love to do more than just programming.";
}
|
|
|
|
|
Hamed Mosavi wrote: S. Senthil Kumar wrote:
Nope, it simply changes your variable's reference to refer to an empty string.
Which means if the variable is null, it news a string and assigns it to the variable.
Nope. The empty string already exists as a constant, so there is no new string object created.
Despite everything, the person most likely to be fooling you next is yourself.
|
|
|
|
|
Got it. Thanks.
// "In the end it's a little boy expressing himself." Yanni
while (I_am_alive)
{
cout<<"I love to do more than just programming.";
}
|
|
|
|
|
No, it's an empty string, your string would have a reference to it, that's the point of using it.
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 )
|
|
|
|
|
string.Empty has no advantage over the literal ""; I see no point in its inclusion in the framework.
|
|
|
|
|
I'm trying to do a minor upgrade from v1.0.1 to 1.0.2 for instance.
I change the version code and visual studio recommends that I change the product code. If I do this though I end up with 2 versions of the software installed. If I don't do it I get an error when I attempt to install telling me that the program is already installed.
How do i get the installer to just overwrite the stuff it needs to without forcing the user to manually uninstall and without them ending up with a ne add/remove programs entry everytime they upgrade?
Cheers,
Russell
|
|
|
|
|
This is typically a nightmare, especially with the under powered installation stuff built into VS. You probably need to buy a real installation tool.
However, what I tend to do, is deploy patches that just replace the exe. I write the path to my exe in the registry, so I can easily write an app that finds the path, and copies in a new exe.
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 )
|
|
|
|
|
I might end up resorting to something like that but I'll try leppie's approach for the moment as I'm a bit up against it time wise at the moment.
Cheers,
Russell
|
|
|
|
|
Russell Jones wrote: How do i get the installer to just overwrite the stuff it needs to without forcing the user to manually uninstall and without them ending up with a ne add/remove programs entry everytime they upgrade?
After several attempts to get that working back in the haydays of VS2003, I gave up, chose NSIS, and never looked back.
|
|
|
|
|
Thanks leppie,
I played with NSIS a while back, it's pretty cool as I remember I'll take a look.
Do you get it to create packages as part of the build process?
Russell
|
|
|
|
|
It is quite powerful, and sure it could be run from the commandline, but the learning curve for more advanced stuff is a bit high!
|
|
|
|
|
can it add stuff to the GAC, or should i put all the GAC stuff for my app in an msi and add i that way?
|
|
|
|
|
You just use nsExec I think, so you can do exactly what you do on the commandline, also look at the .NET detecting script example, you will probably need that path. 
|
|
|
|
|
Seeing you strapped for time, have a look at InnoSetup too, also very flexible, and much easier to learn.
|
|
|
|
|
there's a "RemovePreviousVersions" property that you can set to TRUE to take care of that for you.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
