Basic introduction to Java

The open source China community team made its first live broadcast, telling the story behind the open source China community in the name of sharing."

What do you need to know about Java? This article will give you a detailed understanding, welcome to collect!

Java basics

▐Three major characteristics of object-oriented

Features: encapsulation, inheritance, polymorphism;

Encapsulation: Abstract things into an object, privatize the properties of the object, and provide some methods that can be accessed by the outside world;

Inheritance: Subclasses extend new data fields or functions, and reuse the attributes and functions of the parent class, with single inheritance and multiple implementations;

Polymorphism: through inheritance (multiple subclasses rewriting the same method), or through interfaces (implementing interfaces and overriding interfaces);

The difference between Java and C++

Differences: C++ supports multiple inheritance and has the concept of pointers, and programmers manage memory by themselves; Java has single inheritance and can use interfaces to implement multiple inheritance. Java does not provide pointers to directly access memory. Program memory is safer, and Java has The JVM automatic memory management mechanism does not require programmers to manually release unused memory.

Polymorphic implementation principles

The underlying implementation of polymorphism is dynamic binding, which associates method calls with method implementations at runtime.

Static binding and dynamic binding:

One is determined at compile time, called static dispatch, such as method overloading;

One is determined at runtime, called dynamic dispatch, such as method overriding (rewriting) and interface implementation.

polymorphic implementation

The stack frame of the current method call (local variable table, operation stack, dynamic connection, return address) will be stored in the virtual machine stack. The implementation process of polymorphism is the process of dynamic dispatch of method calls. If the subclass overrides the method of the parent class, then in the polymorphic call, the dynamic binding process will first determine that the actual type is a subclass, so as to search for the subclass first. method in. This process is the essence of method coverage.

static and final keywords

static: properties and methods can be modified

static modified attributes:

Class-level attributes, shared by all objects, are loaded as the class is loaded (only loaded once), before the object is created; it can be called directly using the class name.

static modification method:

Loaded as the class is loaded; it can be called directly using the class name; in static methods, only static members can be called, and this cannot be used;

final: The keyword is mainly used in three places: variables, methods, and classes.

final modified variable:

If it is a variable of a basic data type, its value cannot be changed once it is initialized;

If it is a reference type variable, it cannot point to another object after it is initialized.

final modification method:

Lock the method to prevent any inheriting class from modifying its meaning (overriding); all private methods in the class are implicitly designated as final.

final modified class:

When final modifies a class, it indicates that this class cannot be inherited. All member methods in a final class are implicitly designated as final methods.

A class cannot be inherited. In addition to the final keyword, the constructor can also be privatized. (inner class is invalid)

Abstract classes and interfaces

Abstract class: A class that contains abstract methods, that is, a class modified with abstract; abstract classes can only be inherited, so final modification cannot be used, and abstract classes cannot be instantiated;

Interface: An interface is an abstract type and a collection of abstract methods. The interface supports multiple inheritance. The methods defined in the interface default to abstract methods modified by public abstract;

Same point:

Neither abstract classes nor interfaces can be instantiated;
Both abstract classes and interfaces can define abstract methods, and subclasses/implementation classes must override these abstract methods;

difference:

Abstract classes have constructors, interfaces have no constructors;
Abstract classes can contain ordinary methods, and interfaces can only modify abstract methods with public abstract (available after Java 8);
Abstract classes can only be inherited singly, interfaces can be inherited multiple times;
Abstract classes can define various types of member variables, and interfaces can only be static constants modified by public static final;

Usage scenarios of abstract classes:

Not only do you want to constrain subclasses to have common behaviors (but don't care how to implement them), but you also want to have default methods and instance variables;

Interface application scenarios:

Constrain multiple implementation classes to have uniform behavior, but it does not care how each implementation class is implemented; there may be no connection between various functions in the implementation class;

Generics and generic erasure

Reference: https://blog.csdn.net/baoyinwang/article/details/107341997

Generics:

The essence of generics is parameterized types. This parameter type can be used in the creation of classes, interfaces and methods, called generic classes, generic interfaces and generic methods respectively.

Generic erasure:

Java's generics are pseudo-generics. When using generics, type parameters are added, which will be removed when the compiler compiles the generated bytecode. This process is called type erasure.

Types such as List will become List after compilation. All the JVM sees is the List, and the type information attached by generics is invisible to the JVM.

Other types of elements can be added through reflection.

Reflection principles and usage scenarios

Java reflection:

It means that in the running state, for any class, you can know all the properties and methods of this class; and you can call any of its methods;

Reflection principle:

Reflection first obtains the bytecode of the reflection class in Java, and then maps the methods, variables, constructors, etc. in the bytecode to the corresponding Method, Filed, Constructor and other classes.

How to get an instance of Class:

  
  
  
  
  
   
   
   
   
   
   
   
   
   1.类名.class(就是一份字节码)2.Class.forName(String className);根据一个类的全限定名来构建Class对象3.每一个对象多有getClass()方法:obj.getClass();返回对象的真实类型
   
  
   
  
   
  
   
  
  

  
  
  
  
  
   
   
   
   
   
   
   
   
   scenes to be used:

Develop common frameworks - The most important use of reflection is to develop various common frameworks. Many frameworks (such as Spring) are configured (such as configuring JavaBeans, Filters, etc. through XML files). In order to ensure the versatility of the framework, it is necessary to dynamically load different objects or classes and call different methods according to the configuration file during runtime.
Dynamic proxy - In aspect programming (AOP), it is necessary to intercept specific methods. Usually, the dynamic proxy method is chosen. At this time, reflection technology is needed to achieve it.
JDK: spring default dynamic proxy needs to implement the interface;

CGLIB: serializes byte streams through the asm framework, configurable, poor performance;
Custom annotation - the annotation itself only serves as a mark. It needs to use the reflection mechanism to call the annotation interpreter according to the annotation mark and execute the behavior.

Java exception system

Throwable is the super class for all errors or exceptions in the Java language. The next level is divided into Error and Exception.

Error：

Refers to internal errors and resource exhaustion errors of the Java runtime system. The application will not throw objects of this class. If such an error occurs, in addition to informing the user, the rest is to try to terminate the program safely.

Exception包含：RuntimeException、CheckedException；

Programming errors can be divided into three categories: syntax errors, logic errors and operational errors.

Syntax errors (also called compilation errors) are errors that occur during the compilation process and are checked by the compiler to find syntax errors.

A logical error means that the execution result of the program is not in line with expectations. The cause of the error can be located and discovered through debugging.

Runtime errors are errors that cause the program to terminate abnormally, and runtime errors need to be handled through exception handling.

RuntimeException: Runtime exception, the program should try to avoid such exceptions from a logical perspective.

Such as NullPointerException, ClassCastException;

CheckedException: Checked exception, the program uses trycatch to catch and process;

如IOException、SQLException、NotFoundException；

▐ 数据结构

ArrayList和LinkedList

ArrayList：

底层基于数组实现，支持对元素进行快速随机访问，适合随机查找和遍历，不适合插入和删除。（提一句实际上）

默认初始大小为10，当数组容量不够时，会触发扩容机制（扩大到当前的1.5倍），需要将原来数组的数据复制到新的数组中；当从ArrayList的中间位置插入或者删除元素时，需要对数组进行复制、移动、代价比较高。

LinkedList：

底层基于双向链表实现，适合数据的动态插入和删除；
内部提供了List接口中没有定义的方法，用于操作表头和表尾元素，可以当作堆栈、队列和双向队列使用。（比如jdk官方推荐使用基于linkedList的Deque进行堆栈操作）

ArrayList与LinkedList区别：

都是线程不安全的，ArrayList适用于查找的场景，LinkedList适用于增加、删除多的场景。

实现线程安全：

可以使用原生的Vector，或者是Collections.synchronizedList(List list)函数返回一个线程安全的ArrayList集合。

建议使用concurrent并发包下的CopyOnWriteArrayList的。

①Vector:底层通过synchronize修饰保证线程安全，效率较差。

②CopyOnWriteArrayList：写时加锁，使用了一种叫写时复制的方法；读操作是可以不用加锁的。

List遍历快速和安全失败

普通for循环遍历List删除指定元素

  
  
  
  
  
   
   
   
   
   
   
   
   
   for(int i=0; i < list.size(); i++){   if(list.get(i) == 5)        list.remove(i);}

迭代遍历,用list.remove(i)方法删除元素

  
  
  
  
  
   
   
   
   
   
   
   
   
   Iterator<Integer> it = list.iterator();while(it.hasNext()){    Integer value = it.next();    if(value == 5){        list.remove(value);    }}

foreach遍历List删除元素

  
  
  
  
  
   
   
   
   
   
   
   
   
   for(Integer i:list){    if(i==3) list.remove(i);}

fail—fast：快速失败

当异常产生时，直接抛出异常，程序终止。

fail-fast主要是体现在当我们在遍历集合元素的时候，经常会使用迭代器，但在迭代器遍历元素的过程中，如果集合的结构（modCount）被改变的话，就会抛出异常ConcurrentModificationException，防止继续遍历。这就是所谓的快速失败机制。

fail—safe：安全失败

采用安全失败机制的集合容器，在遍历时不是直接在集合内容上访问的，而是先复制原有集合内容，在拷贝的集合上进行遍历。由于在遍历过程中对原集合所作的修改并不能被迭代器检测到，所以不会触发ConcurrentModificationException。

缺点：基于拷贝内容的优点是避免了ConcurrentModificationException，但同样地，迭代器并不能访问到修改后的内容，即：迭代器遍历的是开始遍历那一刻拿到的集合拷贝，在遍历期间原集合发生的修改迭代器是不知道的。

场景：java.util.concurrent包下的容器都是安全失败，可以在多线程下并发使用，并发修改。

详细介绍HashMap

角度：数据结构+扩容情况+put查找的详细过程+哈希函数+容量为什么始终都是2^N，JDK1.7与1.8的区别。

参考：https://www.jianshu.com/p/9fe4cb316c05

数据结构：

HashMap在底层数据结构上采用了数组＋链表＋红黑树，通过散列映射来存储键值对数据。

扩容情况：

默认的负载因子是0.75，如果数组中已经存储的元素个数大于数组长度的75%，将会引发扩容操作。

【1】创建一个长度为原来数组长度两倍的新数组。

【2】1.7采用Entry的重新hash运算，1.8采用高于运算。

put操作步骤：

1、判断数组是否为空，为空进行初始化;

2、不为空，则计算 key的hash值，通过(n - 1) & hash计算应当存放在数组中的下标 index;

3、查看table[index] 是否存在数据，没有数据就构造一个Node节点存放在 table[index] 中；

4、存在数据，说明发生了hash冲突(存在二个节点key的hash值一样), 继续判断key是否相等，相等，用新的value替换原数据；

5、若不相等，判断当前节点类型是不是树型节点，如果是树型节点，创造树型节点插入红黑树中；

6、若不是红黑树，创建普通Node加入链表中；判断链表长度是否大于8，大于则将链表转换为红黑树；

7、插入完成之后判断当前节点数是否大于阈值，若大于，则扩容为原数组的二倍；

哈希函数：

通过hash函数（优质因子31循环累加）先拿到key的hashcode，是一个32位的值，然后让hashcode的高16位和低16位进行异或操作。该函数也称为扰动函数，做到尽可能降低hash碰撞，通过尾插法进行插入。

容量为什么始终都是2^N：

先做对数组的⻓度取模运算，得到的余数才能⽤来要存放的位置也就是对应的数组下标。这个数组下标的计算⽅法是“ (n - 1) & hash ”。（n代表数组⻓度）。方便数组的扩容和增删改时的取模。

JDK1.7与1.8的区别：

JDK1.7 HashMap：

底层是 数组和链表 结合在⼀起使⽤也就是链表散列。如果相同的话，直接覆盖，不相同就通过拉链法解决冲突。扩容翻转时顺序不一致使用头插法会产生死循环，导致cpu100%

JDK1.8 HashMap：

底层数据结构上采用了数组＋链表＋红黑树；当链表⻓度⼤于阈值（默认为 8-泊松分布），数组的⻓度大于 64时，链表将转化为红⿊树，以减少搜索时间。（解决了tomcat臭名昭著的url参数dos攻击问题）

ConcurrentHashMap

可以通过ConcurrentHashMap和Hashtable来实现线程安全；Hashtable 是原始API类，通过synchronize同步修饰，效率低下；ConcurrentHashMap通过分段锁实现，效率较比Hashtable要好。

ConcurrentHashMap的底层实现：

JDK1.7的ConcurrentHashMap底层采⽤分段的数组+链表实现；采用 分段锁（Sagment）对整个桶数组进⾏了分割分段(Segment默认16个)，每⼀把锁只锁容器其中⼀部分数据，多线程访问容器⾥不同数据段的数据，就不会存在锁竞争，提⾼并发访问率。

JDK1.8的 ConcurrentHashMap采⽤的数据结构跟HashMap1.8的结构⼀样，数组+链表/红⿊树；摒弃了Segment的概念，⽽是直接⽤ Node 数组+链表+红⿊树的数据结构来实现，通过并发控制synchronized和CAS来操作保证线程的安全。

序列化和反序列化

序列化的意思就是将对象的状态转化成字节流，以后可以通过这些值再生成相同状态的对象。对象序列化是对象持久化的一种实现方法，它是将对象的属性和方法转化为一种序列化的形式用于存储和传输。反序列化就是根据这些保存的信息重建对象的过程。

序列化：将java对象转化为字节序列的过程。

反序列化：将字节序列转化为java对象的过程。

优点：

实现了数据的持久化，通过序列化可以把数据永久地保存到硬盘上（通常存放在文件里）Redis的RDB
利用序列化实现远程通信，即在网络上传送对象的字节序列。Google的protoBuf。

反序列化失败的场景：

序列化ID：serialVersionUID不一致的时候，导致反序列化失败。

String

String使用数组存储内容，数组使用final修饰，因此String定义的字符串的值也是不可变的。

StringBuffer对方法加了同步锁，线程安全，效率略低于StringBuilder。

▐ 设计模式与原则

单例模式

某个类只能生成一个实例，该实例全局访问，例如Spring容器里一级缓存里的单例池。

优点：

唯一访问：如生成唯一序列化的场景、或者spring默认的bean类型。

提高性能：频繁实例化创建销毁或者耗时耗资源的场景，如连接池、线程池。

缺点：

不适合有状态且需变更的；

实现方式：

饿汉式：线程安全速度快；

懒汉式：双重检测锁，第一次减少锁的开销、第二次防止重复、volatile防止重排序导致实例化未完成；

静态内部类：线程安全利用率高；

枚举：effictiveJAVA推荐，反射也无法破坏；

工厂模式

定义一个用于创建产品的接口，由子类决定生产何种产品。

优点：解耦：提供参数即可获取产品，通过配置文件可以不修改代码增加具体产品。

缺点：每增加一个产品就得新增一个产品类。

抽象工厂模式

提供一个接口，用于创建相关或者依赖对象的家族，并由此进行约束。

优点：可以在类的内部对产品族进行约束。

缺点：假如产品族中需要增加一个新的产品，则几乎所有的工厂类都需要进行修改。

▐ 构造方法

构造方法可以被重载，只有当类中没有显性声明任何构造方法时，才会有默认构造方法。

构造方法没有返回值，构造方法的作用是创建新对象。

▐ 初始化块

静态初始化块的优先级最高，会最先执行，在非静态初始化块之前执行。

静态初始化块会在类第一次被加载时最先执行，因此在main方法之前。

▐ This

关键字this代表当前对象的引用。当前对象指的是调用类中的属性或方法的对象。

关键字this不可以在静态方法中使用。静态方法不依赖于类的具体对象的引用。

▐ 重写和重载的区别

重载指在同一个类中定义多个方法，这些方法名称相同，签名不同。

重写指在子类中的方法的名称和签名都和父类相同，使用override注解。

▐ Object类方法

toString默认是个指针，一般需要重写；

equals比较对象是否相同，默认和==功能一致；

hashCode散列码，equals则hashCode相同，所以重写equals必须重写hashCode；

finalize用于垃圾回收之前做的遗嘱，默认空，子类需重写；

clone深拷贝，类需实现cloneable的接口；

getClass反射获取对象元数据，包括类名、方法；

notify、wait用于线程通知和唤醒；

▐ 基本数据类型和包装类

¤ 拓展阅读 ¤

3DXR技术 | 终端技术 | 音视频技术

服务端技术 | 技术质量 | 数据算法

本文分享自微信公众号 - 大淘宝技术（AlibabaMTT）。
如有侵权，请联系 [email protected] 删除。
本文参与“OSC源创计划”，欢迎正在阅读的你也加入，一起分享。