字节流、字符流

概述

根据数据的类型分为:字节流字符流

  • 字节流 :以字节为单位,读写数据的流。
  • 字符流 :以字符为单位,读写数据的流。

根据数据的流向分为:输入流输出流

  • 输入流 :把数据从其他设备上读取到内存中的流。
  • 输出流 :把数据从内存 中写出到其他设备上的流。

字节流

概述

一切文件数据(文本、图片、视频等)在存储时,都是以二进制数字的形式保存,都一个一个的字节,那么传输时一样如此。所以,字节流可以传输任意文件数据。在操作流的时候,我们要时刻明确,无论使用什么样的流对象,底层传输的始终为二进制数据。

字节输出流(OutputStream)

java.io.OutputStream 抽象类是表示字节输出流的所有类的超类,将指定的字节信息写出到目的地。它定义了字节输出流的基本共性功能方法。

  • public void close() :关闭此输出流并释放与此流相关联的任何系统资源。
  • public void flush() :刷新此输出流并强制任何缓冲的输出字节被写出。
  • public void write(byte[] b):将 b.length字节从指定的字节数组写入此输出流。
  • public void write(byte[] b, int off, int len) :从指定的字节数组写入 len字节,从偏移量 off开始输出到此输出流。
  • public abstract void write(int b) :将指定的字节输出流。

小贴士:

close方法,当完成流的操作时,必须调用此方法,释放系统资源。

FileOutputStream类

java.io.FileOutputStream 类是文件输出流,用于将数据写出到文件。

构造方法
  • public FileOutputStream(File file):创建文件输出流以写入由指定的 File对象表示的文件。
  • public FileOutputStream(String name): 创建文件输出流以指定的名称写入文件。

当你创建一个流对象时,必须传入一个文件路径。该路径下,如果没有这个文件,会创建该文件。如果有这个文件,会清空这个文件的数据。

1
2
3
4
5
6
7
8
9
10
public class FileOutputStreamConstructor throws IOException {
public static void main(String[] args) {
// 使用File对象创建流对象
File file = new File("a.txt");
FileOutputStream fos = new FileOutputStream(file);

// 使用文件名称创建流对象
FileOutputStream fos = new FileOutputStream("b.txt");
}
}
写出字节数据

1. 写出字节write(int b) 方法,每次可以写出一个字节数据,

代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
public class FOSWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileOutputStream fos = new FileOutputStream("fos.txt");
// 写出数据
fos.write(97); // 写出第1个字节
fos.write(98); // 写出第2个字节
fos.write(99); // 写出第3个字节
// 关闭资源
fos.close();
}
}
输出结果:
abc

2. 写出字节数组write(byte[] b),每次可以写出数组中的数据,

代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
public class FOSWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileOutputStream fos = new FileOutputStream("fos.txt");
// 字符串转换为字节数组
byte[] b = "你好!".getBytes();
// 写出字节数组数据
fos.write(b);
// 关闭资源
fos.close();
}
}
输出结果:
你好!

3. 写出指定长度字节数组write(byte[] b, int off, int len) ,每次写出从off索引开始,len个字节,

代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
public class FOSWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileOutputStream fos = new FileOutputStream("fos.txt");
// 字符串转换为字节数组
byte[] b = "abcde".getBytes();
// 写出从索引2开始,2个字节。索引2是c,两个字节,也就是cd。
fos.write(b,2,2);
// 关闭资源
fos.close();
}
}
输出结果:
cd

数据追加续写

经过以上的演示,每次程序运行,创建输出流对象,都会清空目标文件中的数据。如何保留目标文件中数据,还能继续添加新数据呢?

  • public FileOutputStream(File file, boolean append): 创建文件输出流以写入由指定的 File对象表示的文件。
  • public FileOutputStream(String name, boolean append): 创建文件输出流以指定的名称写入文件。

这两个构造方法,参数中都需要传入一个boolean类型的值,true 表示追加数据,false 表示清空原有数据。这样创建的输出流对象,就可以指定是否追加续写了,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
public class FOSWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileOutputStream fos = new FileOutputStream("fos.txt"true);
// 字符串转换为字节数组
byte[] b = "abcde".getBytes();
// 写出从索引2开始,2个字节。索引2是c,两个字节,也就是cd。
fos.write(b);
// 关闭资源
fos.close();
}
}
文件操作前:cd
文件操作后:cdabcde

写出换行

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
public class FOSWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileOutputStream fos = new FileOutputStream("fos.txt");
// 定义字节数组
byte[] words = {97,98,99,100,101};
// 遍历数组
for (int i = 0; i < words.length; i++) {
// 写出一个字节
fos.write(words[i]);
// 写出一个换行, 换行符号转成数组写出
fos.write("\r\n".getBytes());
}
// 关闭资源
fos.close();
}
}

输出结果:
a
b
c
d
e

字节输入流(InputStream)

java.io.InputStream 抽象类是表示字节输入流的所有类的超类,可以读取字节信息到内存中。它定义了字节输入流的基本共性功能方法。

  • public void close() :关闭此输入流并释放与此流相关联的任何系统资源。
  • public abstract int read(): 从输入流读取数据的下一个字节。
  • public int read(byte[] b): 从输入流中读取一些字节数,并将它们存储到字节数组 b中 。

FileInputStream类

java.io.FileInputStream 类是文件输入流,从文件中读取字节。

构造方法
  • FileInputStream(File file): 通过打开与实际文件的连接来创建一个 FileInputStream ,该文件由文件系统中的 File对象 file命名。
  • FileInputStream(String name): 通过打开与实际文件的连接来创建一个 FileInputStream ,该文件由文件系统中的路径名 name命名。

当你创建一个流对象时,必须传入一个文件路径。该路径下,如果没有该文件,会抛出FileNotFoundException

1
2
3
4
5
6
7
8
9
10
public class FileInputStreamConstructor throws IOException{
public static void main(String[] args) {
// 使用File对象创建流对象
File file = new File("a.txt");
FileInputStream fos = new FileInputStream(file);

// 使用文件名称创建流对象
FileInputStream fos = new FileInputStream("b.txt");
}
}
读取字节数据

读取字节read方法,每次可以读取一个字节的数据,提升为int类型,读取到文件末尾,返回-1,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
public class FISRead {
public static void main(String[] args) throws IOException{
// 使用文件名称创建流对象
FileInputStream fis = new FileInputStream("read.txt");
// 读取数据,返回一个字节
int read = fis.read();
System.out.println((char) read);
read = fis.read();
System.out.println((char) read);
read = fis.read();
System.out.println((char) read);
read = fis.read();
System.out.println((char) read);
read = fis.read();
System.out.println((char) read);
// 读取到末尾,返回-1
read = fis.read();
System.out.println( read);
// 关闭资源
fis.close();
}
}
输出结果:
a
b
c
d
e
-1

循环改进读取方式,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
public class FISRead {
public static void main(String[] args) throws IOException{
// 使用文件名称创建流对象
FileInputStream fis = new FileInputStream("read.txt");
// 定义变量,保存数据
int b ;
// 循环读取
while ((b = fis.read())!=-1) {
System.out.println((char)b);
}
// 关闭资源
fis.close();
}
}
输出结果:
a
b
c
d
e

使用字节数组读取read(byte[] b),每次读取b的长度个字节到数组中,返回读取到的有效字节个数,读取到末尾时,返回-1 ,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
public class FISRead {
public static void main(String[] args) throws IOException{
// 使用文件名称创建流对象.
FileInputStream fis = new FileInputStream("read.txt"); // 文件中为abcde
// 定义变量,作为有效个数
int len ;
// 定义字节数组,作为装字节数据的容器
byte[] b = new byte[2];
// 循环读取
while (( len= fis.read(b))!=-1) {
// 每次读取后,把数组变成字符串打印
System.out.println(new String(b));
}
// 关闭资源
fis.close();
}
}

输出结果:
ab
cd
ed

错误数据d,是由于最后一次读取时,只读取一个字节e,数组中,上次读取的数据没有被完全替换,所以要通过len ,获取有效的字节,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
public class FISRead {
public static void main(String[] args) throws IOException{
// 使用文件名称创建流对象.
FileInputStream fis = new FileInputStream("read.txt"); // 文件中为abcde
// 定义变量,作为有效个数
int len ;
// 定义字节数组,作为装字节数据的容器
byte[] b = new byte[2];
// 循环读取
while (( len= fis.read(b))!=-1) {
// 每次读取后,把数组的有效字节部分,变成字符串打印
System.out.println(new String(b,0,len));// len 每次读取的有效字节个数
}
// 关闭资源
fis.close();
}
}

输出结果:
ab
cd
e

案例练习:图片复制

代码实现:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
public class Copy {
public static void main(String[] args) throws IOException {
// 1.创建流对象
// 1.1 指定数据源
FileInputStream fis = new FileInputStream("D:\\test.jpg");
// 1.2 指定目的地
FileOutputStream fos = new FileOutputStream("test_copy.jpg");

// 2.读写数据
// 2.1 定义数组
byte[] b = new byte[1024];
// 2.2 定义长度
int len;
// 2.3 循环读取
while ((len = fis.read(b))!=-1) {
// 2.4 写出数据
fos.write(b, 0 , len);
}

// 3.关闭资源
fos.close();
fis.close();
}
}

流的关闭原则:先开后关,后开先关。

字符流

当使用字节流读取文本文件时,可能会有一个小问题。就是遇到中文字符时,可能不会显示完整的字符,那是因为一个中文字符可能占用多个字节存储。所以Java提供一些字符流类,以字符为单位读写数据,专门用于处理文本文件。

字符输入流

java.io.Reader抽象类是表示用于读取字符流的所有类的超类,可以读取字符信息到内存中。它定义了字符输入流的基本共性功能方法。

  • public void close() :关闭此流并释放与此流相关联的任何系统资源。
  • public int read(): 从输入流读取一个字符。
  • public int read(char[] cbuf): 从输入流中读取一些字符,并将它们存储到字符数组 cbuf中 。
FileReader类

java.io.FileReader 类是读取字符文件的便利类。构造时使用系统默认的字符编码和默认字节缓冲区。

构造方法
  • FileReader(File file): 创建一个新的 FileReader ,给定要读取的File对象。
  • FileReader(String fileName): 创建一个新的 FileReader ,给定要读取的文件的名称。

当你创建一个流对象时,必须传入一个文件路径。类似于FileInputStream 。

  • 构造举例,代码如下:
1
2
3
4
5
6
7
8
9
10
public class FileReaderConstructor throws IOException{
public static void main(String[] args) {
// 使用File对象创建流对象
File file = new File("a.txt");
FileReader fr = new FileReader(file);

// 使用文件名称创建流对象
FileReader fr = new FileReader("b.txt");
}
}
读取字符数据

读取字符read方法,每次可以读取一个字符的数据,提升为int类型,读取到文件末尾,返回-1,循环读取,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
public class FRRead {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileReader fr = new FileReader("read.txt");
// 定义变量,保存数据
int b ;
// 循环读取
while ((b = fr.read())!=-1) {
System.out.println((char)b);
}
// 关闭资源
fr.close();
}
}

使用字符数组读取read(char[] cbuf),每次读取b的长度个字符到数组中,返回读取到的有效字符个数,读取到末尾时,返回-1 ,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
public class FRRead {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileReader fr = new FileReader("read.txt");
// 定义变量,保存有效字符个数
int len ;
// 定义字符数组,作为装字符数据的容器
char[] cbuf = new char[2];
// 循环读取
while ((len = fr.read(cbuf))!=-1) {
System.out.println(new String(cbuf));
}
// 关闭资源
fr.close();
}
}

获取有效的字符改进,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
public class FISRead {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileReader fr = new FileReader("read.txt");
// 定义变量,保存有效字符个数
int len ;
// 定义字符数组,作为装字符数据的容器
char[] cbuf = new char[2];
// 循环读取
while ((len = fr.read(cbuf))!=-1) {
System.out.println(new String(cbuf,0,len));
}
// 关闭资源
fr.close();
}
}

字符输出流

java.io.Writer 抽象类是表示用于写出字符流的所有类的超类,将指定的字符信息写出到目的地。它定义了字节输出流的基本共性功能方法。

  • void write(int c) 写入单个字符。
  • void write(char[] cbuf) 写入字符数组。
  • abstract void write(char[] cbuf, int off, int len) 写入字符数组的某一部分,off数组的开始索引,len写的字符个数。
  • void write(String str) 写入字符串。
  • void write(String str, int off, int len) 写入字符串的某一部分,off字符串的开始索引,len写的字符个数。
  • void flush() 刷新该流的缓冲。
  • void close() 关闭此流,但要先刷新它。
FileWriter类

java.io.FileWriter 类是写出字符到文件的便利类。构造时使用系统默认的字符编码和默认字节缓冲区。

构造方法
  • FileWriter(File file): 创建一个新的 FileWriter,给定要读取的File对象。
  • FileWriter(String fileName): 创建一个新的 FileWriter,给定要读取的文件的名称。

当你创建一个流对象时,必须传入一个文件路径,类似于FileOutputStream。

  • 构造举例,代码如下:
1
2
3
4
5
6
7
8
9
10
public class FileWriterConstructor {
public static void main(String[] args) throws IOException {
// 使用File对象创建流对象
File file = new File("a.txt");
FileWriter fw = new FileWriter(file);

// 使用文件名称创建流对象
FileWriter fw = new FileWriter("b.txt");
}
}
写出数据

写出字符write(int b) 方法,每次可以写出一个字符数据,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
public class FWWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileWriter fw = new FileWriter("fw.txt");
// 写出数据
fw.write(97); // 写出第1个字符
fw.write('b'); // 写出第2个字符
fw.write('C'); // 写出第3个字符
fw.write(30000); // 写出第4个字符,中文编码表中30000对应一个汉字。

/*
【注意】关闭资源时,与FileOutputStream不同。
如果不关闭,数据只是保存到缓冲区,并未保存到文件。
*/
// fw.close();
}
}
输出结果:
abC田
关闭和刷新

因为内置缓冲区的原因,如果不关闭输出流,无法写出字符到文件中。但是关闭的流对象,是无法继续写出数据的。如果我们既想写出数据,又想继续使用流,就需要flush 方法了。

  • flush :刷新缓冲区,流对象可以继续使用。
  • close :先刷新缓冲区,然后通知系统释放资源。流对象不可以再被使用了。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
public class FWWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileWriter fw = new FileWriter("fw.txt");
// 写出数据,通过flush
fw.write('刷'); // 写出第1个字符
fw.flush();
fw.write('新'); // 继续写出第2个字符,写出成功
fw.flush();

// 写出数据,通过close
fw.write('关'); // 写出第1个字符
fw.close();
fw.write('闭'); // 继续写出第2个字符,【报错】java.io.IOException: Stream closed
fw.close();
}
}
写出数据

写出字符数组write(char[] cbuf)write(char[] cbuf, int off, int len) ,每次可以写出字符数组中的数据,用法类似FileOutputStream,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
public class FWWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileWriter fw = new FileWriter("fw.txt");
// 字符串转换为字节数组
char[] chars = "黑马程序员".toCharArray();

// 写出字符数组
fw.write(chars); // 黑马程序员

// 写出从索引2开始,2个字节。索引2是'程',两个字节,也就是'程序'。
fw.write(b,2,2); // 程序

// 关闭资源
fos.close();
}
}

写出字符串write(String str)write(String str, int off, int len) ,每次可以写出字符串中的数据,更为方便,代码使用演示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
public class FWWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象
FileWriter fw = new FileWriter("fw.txt");
// 字符串
String msg = "黑马程序员";

// 写出字符数组
fw.write(msg); //黑马程序员

// 写出从索引2开始,2个字节。索引2是'程',两个字节,也就是'程序'。
fw.write(msg,2,2); // 程序

// 关闭资源
fos.close();
}
}

续写和换行:操作类似于FileOutputStream。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
public class FWWrite {
public static void main(String[] args) throws IOException {
// 使用文件名称创建流对象,可以续写数据
FileWriter fw = new FileWriter("fw.txt"true);
// 写出字符串
fw.write("我叫");
// 写出换行
fw.write("\r\n");
// 写出字符串
fw.write("仝少坤");
// 关闭资源
fw.close();
}
}
输出结果:
我叫
仝少坤

Properties类

可以作为工具类的配置文件使用

public Properties() :创建一个空的属性列表。

基本的存储方法
  • public Object setProperty(String key, String value) : 保存一对属性。
  • public String getProperty(String key) :使用此属性列表中指定的键搜索属性值。
  • public Set<String> stringPropertyNames() :所有键的名称的集合。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
public class ProDemo {
public static void main(String[] args) throws FileNotFoundException {
// 创建属性集对象
Properties properties = new Properties();
// 添加键值对元素
properties.setProperty("filename", "a.txt");
properties.setProperty("length", "209385038");
properties.setProperty("location", "D:\\a.txt");
// 打印属性集对象
System.out.println(properties);
// 通过键,获取属性值
System.out.println(properties.getProperty("filename"));
System.out.println(properties.getProperty("length"));
System.out.println(properties.getProperty("location"));

// 遍历属性集,获取所有键的集合
Set<String> strings = properties.stringPropertyNames();
// 打印键值对
for (String key : strings ) {
System.out.println(key+" -- "+properties.getProperty(key));
}
}
}
输出结果:
{filename=a.txt, length=209385038, location=D:\a.txt}
a.txt
209385038
D:\a.txt
filename -- a.txt
length -- 209385038
location -- D:\a.txt

public void load(InputStream inStream): 从字节输入流中读取键值对。

参数中使用了字节输入流,通过流对象,可以关联到某文件上,这样就能够加载文本中的数据了。文本数据格式:

1
2
3
filename=a.txt
length=209385038
location=D:\a.txt

加载代码:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
public class ProDemo2 {
public static void main(String[] args) throws FileNotFoundException {
// 创建属性集对象
Properties pro = new Properties();
// 加载文本中信息到属性集
pro.load(new FileInputStream("read.txt"));
// 遍历集合并打印
Set<String> strings = pro.stringPropertyNames();
for (String key : strings ) {
System.out.println(key+" -- "+pro.getProperty(key));
}
}
}
输出结果:
filename -- a.txt
length -- 209385038
location -- D:\a.txt