I was asked during the interview before whether I understand C++ or not, or whether I understand "polymorphism". I have been using Python for scientific research before and I don’t know how to use C++. I have never heard of the word "polymorphism" at all, I have only heard of "polymodality" (for unreasonable connection, deep learning). The full text mainly records several major questions about me: What is polymorphism? What is object-oriented programming? Is Python an Object-Oriented Programming (OOP) language? What is compilation? In addition to object-oriented programming, what other programming paradigms are there?
1. What is polymorphism?
The literal meaning of polymorphism is "multiple forms". In C++, it refers to calling different types of objects through a unified interface to produce different behavioral results. For example: You can command the "animal" to make a scream, but whether it is a cat or a dog barking depends on the type of the actual object.
1. How to achieve polymorphism?
The most common polymorphic form is throughVirtual functionsandinheritAchieved.
Virtual functions: used in base classesvirtualKeyword declaration function, subclasses can be overridden (override) This function.
class Animal {
public:
virtual void sound() { cout << "animals are called" << endl; } // virtual function
};
class Cat : public Animal {
public:
void sound() override { cout << "Meow Meow" << endl; } // Rewrite virtual functions
};
int main() {
Animal* animal = new Cat();
animal->sound(); // Output "meow" instead of "animal call"
}At this time, the base class pointeranimalIt actually points to a subclass object, and the implementation of the subclass will be automatically selected when calling a virtual function.
Seeing this, I vaguely think of the neural network, when Python builds the network, inherits the parent class.。
2. Why do polymorphism need?
- Code reusability: All subclass objects are processed uniformly through the base class interface to avoid duplicate code.
- Extensibility: When adding a new subclass (such as adding a "bird" class to inherit "animals", the original code does not need to be modified, just implement a new virtual function.
- flexibility: When the program is running, you can dynamically adjust the behavior according to the actual object type, such as different characters in the game perform different attack actions.
Imagine you have a universal remote control (base class pointer) that can control TVs of different brands (subclass objects). When the "Start" button is pressed (calling the virtual function), the Sony TV will display the LOGO, and the Samsung TV will play the start music - this is polymorphism: the same operation (start), different objects (TV) produce different behaviors.
By mastering polymorphism, we can write more elegant and easy-to-maintain code, which is alsoObject-Oriented Programmingone of the core advantages.
2. What is object-oriented programming?
When I read C++ popular science posts, I always see that it is object-oriented programming, and I have never understood this concept. Object-oriented programming? So what if I am single, I have no partner? Who are you facing?
1. What is Object Oriented Programming (OOP)?
Object-oriented programming (OOP) is a kind ofObjectAs the core programming paradigm, it encapsulates the data in the program and the logic of the operational data into independent "objects". Each object containsProperties(Data) andMethod(Operation), and define common characteristics of similar objects through classes. For example, a "dog" is a class, and a specific dog (such as "Wangcai") is an object, which has its own color (attribute) and barking (method).
2. What object is it?
An object is the instantiation result of a class and is a concrete thing in the program. For example:
Category: Human (including attributes such as "name", "age", and methods such as "speaking" and "walking").
- Object: Zhang San (the name attribute is "Zhang San" and the age attribute is 20).
- The characteristic of an object is to manage data and behaviors in a unified manner, hiding internal implementation details (encapsulation).
Then I seem to understand. It’s a bit like building a neural network class, passing different parameters through instantiation, changing the number of layers of the network, and obtaining different versions of the network.
3. Why is object-oriented programming required?
- Code reusability: Through inheritance, subclasses can directly replicate the properties and methods of the parent class. For example, after "cats" and "dogs" inherit the "animal" category, there is no need to repeatedly define basic behaviors such as "breathing" and "eating".
- Extensibility: When adding new subclasses (such as "birds" inheriting "animals"), the original code does not need to be modified, just rewrite the virtual function or add new methods.
- Maintenance: Encapsulation protects the internal data of the object and avoids arbitrary external modifications (such as through
privateRestrict access). - Flexibility: Polymorphism allows different objects to respond differently to the same operation. For example, "animals"
sound()Method, the cat object returns "Meow" and the dog object returns "Woom".
3. Is Python an object-oriented programming (OOP) language?
After reading so much, I feel that some parts are very similar to Python, so I have a question: Is Python an object-oriented programming (OOP) language? I remember everyone said it was a scripting language? What is a scripting language?
Python is aFully support for object-oriented programminglanguage. It has three core features of OOP: encapsulation, inheritance and polymorphism.
-
Package: Python pass class (
class) encapsulate data (attributes) and methods (functions) of manipulating data, for example, by__init__Methods initialize object properties and control the visibility of properties by accessing modifiers such as double underscore__. - inherit: Subclasses can inherit the properties and methods of the parent class, for example
class Cat(Animal)expressCatInherited fromAnimal, subclasses can override or extend parent class methods. -
Polymorphism: Objects of different classes can call methods of the same name but exhibit different behaviors. For example,
AnimalClassicsound()Method inDogandCatThe subclasses are implemented as "Wangwang" and "Meow Meow" respectively.
1. Why is Python a scripting language? What is a script?
Scripting Language isWritten in text、Explain executionThe programming language is mainly used for automated tasks and rapid development. Its features include:
- No need for compilation: Execute directly through the interpreter line by line, for example, Python's .py file is parsed by the interpreter on the fly.
- Dynamic Type: Variable types are automatically inferred at runtime without explicit declaration.
- Glue language features: Commonly used to integrate other language modules (such as C/C++) to simplify the construction of complex systems.
2. How does Python support both OOP and scripting language features?
Python is a multi-paradigm language that supports object-oriented programming and has the flexibility of a scripting language:
1) OOP's deep integration：
- All data types (such as integers, strings) are essentially objects.
- Implement complex logic encapsulation through classes, suitable for large-scale project development.
2) Convenience of scripting language：
- A single line of code can complete simple tasks (such as
print("Hello World")）。 - Interactive environments (REPL) support instant testing of code snippets.
4. What is compilation?
Then I have questions again? Python does not need to be compiled, so what is compilation? Why do some languages need to be compiled, while others do not need to be compiled?
1. Compilation definition and core process
Compilation is toAdvanced programming languages(such as C, Java) converts the source code toMachine code or intermediate code (such as byte code) that the computer can execute directlyThe process. Its core process includes the following stages:
- Lexical analysis: Decompose the source code into meaningful words (Tokens), such as identifying variable names, operators, etc.
- Grammar analysis: According to the syntax rules of the programming language, build an abstract syntax tree (AST) to check whether the code structure complies with the specification (such as whether the brackets match).
- Semantic analysis and intermediate code generation: Check the logical legitimacy (such as whether the variable types match) and generate an intermediate representation (such as the three address code).
- Code optimization: Perform performance optimization of intermediate code, such as removing redundant calculations or simplifying loop structures.
- Target code generation: Convert the optimized intermediate code to machine code or byte code for a specific platform (such as Java's
.classdocument).
The final product of the compilation isExecutable file(such as .exe in C language) orPlatform-independent intermediate code(such as Java's bytecode). The advantage is that it has high execution efficiency, but it needs to be recompiled for different platforms.
2. Why do some languages need to be compiled, and some do not?
Whether a programming language needs to be compiled depends on itDesign goalsandHow to execute. It is mainly divided into the following two categories:
1) Languages that need to be compiled (compiled language)
Typical representatives：C、C++、Rust。
Cause：
- Performance priority: Directly compiled into machine code, no running-time parsing is required, and the execution speed is fast.
- Hardware-level control: Suitable for scenarios where operating systems, embedded systems, etc. need to directly operate hardware.
- Static type check: Type errors can be found during compilation to improve code security.
2) Languages that do not require compilation (interpreted or mixed languages)
Typical representatives：Python、JavaScript、Ruby。
Cause：
- Explanation: parsing and executing code line by line through the interpreter, without precompiling, making it easy to quickly debug and cross-platform operation.
- Dynamic feature support: Such as Python's dynamic type, reflection and other features, relying on the runtime environment to parse the code.
-
Mixed execution mode：
- Bytecode + virtual machine(such as Java): First compile into intermediate code, then interpreted and executed by the virtual machine, both cross-platform and efficient.
- Instant Compilation(JIT) (such as JavaScript's V8 engine): dynamically compile hot code at runtime to improve execution speed.
3) The core difference between compilation and explanation
| Dimension | Compiled language | Interpreted Language |
|---|---|---|
| How to execute | Compile first and then execute (one-time conversion) | Line by line explanation execution (real-time conversion) |
| Operational efficiency | High (machine code runs directly) | Low (requires runtime parsing) |
| Development efficiency | Low (repeated compilation required) | High (modify and run directly) |
| Cross-platformity | Depend on the platform (recompiled) | Strong (same interpreter supports multiple platforms) |
| Error detection | Syntax/type errors were found during compilation | Only when running can some errors be found |
5. In addition to object-oriented programming, what other programming paradigms are there?
I vaguely remember that in addition to object-oriented programming, there are also process-oriented programming, and I don’t know what it means.
Programming paradigms are different methodologies for solving problems. C language is process-oriented programming. The core idea of process-oriented programming isfunctionAs the basic unit, solve the problem through a series of steps, focusing on "how to do it" rather than "who will do it". Let’s talk about other programming paradigms later.
Comparison between process-oriented and object-oriented:
| Dimension | Process-oriented | Object-oriented |
|---|---|---|
| Basic Unit | function | Classes and Objects |
| Data and logic | Separation (data is passed through parameters) | Encapsulation (data and method are bound to the object) |
| Code reuse | Low (repeated implementation of similar logic) | High (by inheritance and polymorphic multiplexing code) |
| Applicable scale | Small, simple tasks (such as algorithm implementation) | Large and complex systems (such as game engines) |