I guess one of the most important issues in this topic is building an exception handling architecture in your application. This is interesting for many reasons. And the main reason, I think, is an apparent simplicity, which you don’t always know what to do with. All the basic constructs such as IEnumerable, IDisposable, IObservable, etc. have this property and use it everywhere. On the one hand, their simplicity tempts to use these constructs in different situations. On the other hand, they are full of traps which you might not get out. It is possible that looking at the amount of information we will cover you’ve got a question: what is so special about exceptional situations?

However, to make conclusions about building the architecture of exception classes we should learn some details about their classification. Because before building a system of types that would be clear for the user of code, a programmer should determine when to choose the type of error and when to catch or skip exceptions. So, let’s classify the exceptional situations (not the types of exceptions) based on various features.

Introduction

It’s time to talk about exceptions or, rather, exceptional situations. Before we start, let’s look at the definition. What is an exceptional situation?

This is a situation that makes the execution of current or subsequent code incorrect. I mean different from how it was designed or intended. Such a situation compromises the integrity of an application or its part, e.g. an object. It brings the application into an extraordinary or exceptional state.

But why do we need to define this terminology? Because it will keep us in some boundaries. If we don’t follow the terminology, we can get too far from a designed concept which may result in many ambiguous situations. Let’s see some practical examples:

 struct Number
 {
     public static Number Parse(string source)
     {
         // ...
         if(!parsed)
         {
             throw new ParsingException();
         }
         // ...
     }

     public static bool TryParse(string source, out Number result)
     {
        // ..
        return parsed;
     }
 }

This example seems a little strange, and it is for a reason. I made this code slightly artificial to show the importance of problems appearing in it. First, let’s look at the Parse method. Why should it throw an exception?

Multithreading

Now let’s talk about thin ice. In the previous sections about IDisposable we touched one very important concept that underlies not only the design principles of Disposable types but any type in general. This is the object’s integrity concept. It means that at any given moment of time an object is in a strictly determined state and any action with this object turns its state into one of the variants that were pre-determined while designing a type of this object. In other words, no action with the object should turn it into an undefined state. This results in a problem with the types designed in the above examples. They are not thread-safe. There is a chance the public methods of these types will be called when the destruction of an object is in progress. Let’s solve this problem and decide whether we should solve it at all.

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.

Memory<T> and ReadOnlyMemory<T>

There are two visual differences between Memory<T> and Span<T>. The first one is that Memory<T> type doesn’t contain ref modifier in the header of the type. In other words, the Memory<T> type can be allocated both on the stack while being either a local variable, or a method parameter, or its returned value and on the heap, referencing some data in memory from there. However, this small difference creates a huge distinction in the behavior and capabilities of Memory<T> compared to Span<T>. Unlike Span<T> that is an instrument for some methods to use some data buffer, the Memory<T> type is designed to store information about the buffer, but not to handle it. Thus, there is the difference in API.

• Memory<T> doesn’t have methods to access the data that it is responsible for. Instead, it has the Span property and the Slice method that return an instance of the Span type.
• Additionally, Memory<T> contains the Pin() method used for scenarios when a stored buffer data should be passed to unsafe code. If this method is called when memory is allocated in .NET, the buffer will be pinned and will not move when GC is active. This method will return an instance of the MemoryHandle structure, which encapsulates GCHandle to indicate a segment of a lifetime and to pin array buffer in memory.

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.

Span<T> usage examples

A human by nature cannot fully understand the purpose of a certain instrument until he or she gets some experience. So, let’s turn to some examples.

ValueStringBuilder

One of the most interesting examples in respect to algorithms is the ValueStringBuilder type. However, it is buried deep inside mscorlib and marked with the internal modifier as many other very interesting data types. This means we would not find this remarkable instrument for optimization if we haven’t researched the mscorlib source code.

What is the main disadvantage of the StringBuilder system type? Its main drawback is the type and its basis — it is a reference type and is based on char[], i.e. a character array. At least, this means two things: we use the heap (though not much) anyway and increase the chances to miss the CPU cash.

Another issue with StringBuilder that I faced is the construction of small strings, that is when the resulting string must be short e.g. less than 100 characters. Short formatting raises issues on performance.

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.

SafeHandle / CriticalHandle / SafeBuffer / derived types

I feel I’m going to open the Pandora’s box for you. Let’s talk about special types: SafeHandle, CriticalHandle and their derived types.

This is the last thing about the pattern of a type that gives access to an unmanaged resource. But first, let’s list everything we usually get from unmanaged world:

The first and obvious thing is handles. This may be an meaningless word for a .NET developer, but it is a very important component of the operating system world. A handle is a 32- or 64-bit number by nature. It designates an opened session of interaction with an operating system. For example, when you open a file you get a handle from the WinApi function. Then you can work with it and do Seek, Read or Write operations. Or, you may open a socket for network access. Again an operating system will pass you a handle. In .NET handles are stored as IntPtr type;

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.

Starting from .NET Core 2.0 and .NET Framework 4.5 we can use new data types: Span and Memory. To use them, you just need to install the System.Memory nuget package:

PM> Install-Package System.Memory

These data types are notable because the CLR team has done a great job to implement their special support inside the code of .NET Core 2.1+ JIT compiler by embedding these data types right into the core. What kind of data types are these and why are they worth a whole chapter?

If we talk about problems that made these types appear, I should name three of them. The first one is unmanaged code.

Both the language and the platform have existed for many years along with means to work with unmanaged code. So, why release another API to work with unmanaged code if the former basically existed for many years? To answer this question, we should understand what we lacked before.

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.

The Object base type and implementation of interfaces. Boxing

It seems we came through hell and high water and can nail any interview, even the one for .NET CLR team. However, let's not rush to microsoft.com and search for vacancies. Now, we need to understand how value types inherit an object if they contain neither a reference to SyncBlockIndex, not a pointer to a virtual methods table. This will completely explain our system of types and all pieces of a puzzle will find their places. However, we will need more than one sentence.

Now, let's remember again how value types are allocated in memory. They get the place in memory right where they are. Reference types get allocation on the heap of small and large objects. They always give a reference to the place on the heap where the object is. Each value type has such methods as ToString, Equals and GetHashCode. They are virtual and overridable, but don’t allow to inherit a value type by overriding methods. If value types used overridable methods, they would need a virtual methods table to route calls. This would lead to the problems of passing structures to unmanaged world: extra fields would go there. As a result, there are descriptions of value type methods somewhere, but you cannot access them directly via a virtual methods table.

This may bring the idea that the lack of inheritance is artificial

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.

First, let’s talk about Reference Types and Value Types. I think people don’t really understand the differences and benefits of both. They usually say reference types store content on the heap and value types store content on the stack, which is wrong.

Let’s discuss the real differences:

• A value type: its value is an entire structure. The value of a reference type is a reference to an object. – A structure in memory: value types contain only the data you indicated. Reference types also contain two system fields. The first one stores 'SyncBlockIndex', the second one stores the information about a type, including the information about a Virtual Methods Table (VMT).
• Reference types can have methods that are overridden when inherited. Value types cannot be inherited.
• You should allocate space on the heap for an instance of a reference type. A value type can be allocated on the stack, or it becomes the part of a reference type. This sufficiently increases the performance of some algorithms.

However, there are common features:

• Both subclasses can inherit the object type and become its representatives.

Let’s look closer at each feature.

This chapter was translated from Russian jointly by author and by professional translators. You can help us with translation from Russian or English into any other language, primarily into Chinese or German.

Also, if you want thank us, the best way you can do that is to give us a star on github or to fork repository github/sidristij/dotnetbook.