Sinopsis
You may already know that C++ (pronounced ‘see plus plus’) is so named because it was designed by Bjarne Stroustrup as a successor to an earlier (and still widely used) language called C. In C, ++ means ‘increment’ and, in mathematical terms, to increment means to obtain something’s successor. Therefore, you could interpret the name as meaning ‘the successor of C’. Like the concept of a successor in mathematics, that does not imply replacement. If you already know C, you need to recognize that C++ is a new and different language, even though much C source code will compile as C++. Usually the result of a successful compilation of C source code with a C++ compiler will be a program that behaves exactly like the one produced by a C compiler. However, that is not always true. In the early 1980s, Bjarne Stroustrup designed an extension to C that he called ‘C with classes’. If you are interested in the history of how that personal tool grew up to become the most widely used programming language in the world and one that has fired the imaginations of many people you will have to look elsewhere. (A good place to start would be with The Design and Evolution of C++ [Stroustrup 1994].) This book is about programming in C++ as the ISO/IEC 14882:2003 Standard defines it, that is, Standard C++ as it was specified in 2003 (which is the first official standard, with various corrections that were made between 1998 and 2003).
C++ is one of the most widely used programming languages in the world. It is also one of the largest programming languages ever designed. Bjarne Stroustrup specified that one of the design criteria of the language is that there should be no room for a lower-level language between C++ and native machine code. Very few programmers ever use C++’s lowest level, and many do not even know that it has an asm keyword, which allows support for writing code in assembler. The incorporation of C into C++ was an important design decision. On the positive side, it made it easy for C programmers to transfer to C++. Having made the transfer they could, at least in theory, incrementally add to their C++ skills and understanding. On the negative side, it has tied C++ to a number of features of C’s design that experience has shown to be, at best, problematical. It has also caused problems to many who have moved from C to C++, because they have made the transition from a C to a C++ compiler without actually making the mental transition from C to C++. They are still C programmers at heart. There is nothing wrong with that, but it does provide a roadblock to their becoming fluent C++ programmers. If you are a C programmer you may find studying modern C++ tougher than you would if your first language were something else. At the high end of C++ we find tools that allow innovators to do metaprogramming, that is, source code that generates source code.Wewill not be exploring that in this book, but it is worth noting that in learning C++ you are learning a language that supports the most innovative development of programming currently around. In between assembler support and support for metaprogramming, C++ provides tools for procedural programming, object-based programming, object-oriented programming (I will explain the difference later when you know enough C++ to appreciate the differences), and generic programming. With care, you can even do some functional programming.
Alongside the raw power of the core of the C++ language, the Standard C++ Library supports a wide range of things that programmers commonly want to do. We have learnt a great deal over the last few years, and were we to start writing a library today we might produce a substantially different one. However, what we have is better than anything provided previously in any widely used programming language. In addition, much of the library has been designed for extension: it is designed so that new components can easily be added and work correctly with standard components. On the other hand the Standard Library currently lacks many of the components that users of more recent languages such as Java, C#, and Python have come to expect. It is both one of the strengths and one of the weaknesses of C++ that it does not dictate a methodology or paradigm. When people first learn C++ this can be a problem, because the range of choice requires understanding of the implications of those choices. If the newcomer already knows another programming language, they will naturally try to discover how to write their first language in C++ terms. They will think in their first language and try to translate into C++. Such is the range of C++ that they can often get a close approximation, but that usually does not lead to good C++.
C++ can be viewed as everyone’s second language. Mastery of C++ requires that you leave behind the crutch of your first language. That is hard and you will make many mistakes along the way. However, the result will be that you are a much better programmer both in C++ and in any other programming language you already know or choose to learn later.
C++ has a wide range of operators. Most of them can be extended to include user-defined types. With the potential for redefinition comes the responsibility to use such a facility wisely. The intention was that it should be possible to add types such as complex numbers, matrices, quaternions, etc. and provide the operators that a domain specialist would expect and find intuitive. Unfortunately, some programmers take the availability of a mechanism as a challenge to find creative ways of using it. The result is that their code becomes ever more obscure.
C++ is a living language. By this I mean two things. The first is that the very best users continue to develop new idioms and other ways to use it. The entire growth of metaprogramming in C++ started one evening when a group of experts realized that the template technology of C++ (designed to support generic programming) was a Turing-complete programming language in its own right, one that was implemented at compile time. C++ was not designed for metaprogramming, so using it is often ugly, but it has enabled experts to explore the potential of metaprogramming.
The second way in which C++ is a living language is that it is subject to periodic change. Even as I write, those responsible for the definition of C++ (WG21, an ISO standards committee) are working on changes that will eventually come into effect at the end of this decade. Some of those changes are to make C++ easier to write and to learn, some are aimed at cleaning up inconsistencies, and some will be aimed at further extending the power of the language. At the time of writing it is impossible to predict exactly what will be added and what changes will be introduced. I know that providing better support for metaprogramming is one of the potential additions to C++.
Content
- Introduction
- Overview of C++
- Getting Started
- Fundamental Types, Operators, and Simple Variables
- Looping and Making Decisions
- Namespaces and the C++ Standard Library
- Writing Functions in C++
- Behavior, Sequence Points, and Order of Evaluation
- Generic Functions
- User-Defined Types, Part 1: typedef and enum
- User-Defined Types, Part 2: Simple classes (value types)
- User-Defined Types, Part 3: Simple classes (homogeneous entity types)
- Pointers, Smart Pointers, Iterators, and Dynamic Instances
- User-Defined Types, Part 4: Class hierarchies, polymorphism, inheritance, and subtypes
- Dynamic Object Creation and Polymorphic Objects
- Streams, Files, and Persistence
- Exceptions
- Overloading Operators and Conversion Operators
- Containers, Iterators, and Algorithms
- Something Old, Something New
.jpg)
