深入理解PHP原理之Opcodes

最新推荐文章于 2022-09-29 21:38:00 发布

alashan007

最新推荐文章于 2022-09-29 21:38:00 发布

阅读量283

点赞数

分类专栏： PHP

原文链接：https://www.laruence.com/2008/06/18/221.html

版权

PHP 专栏收录该内容

120 篇文章 0 订阅

订阅专栏

Opcode是一种PHP脚本编译后的中间语言，就像Java的ByteCode,或者.NET的MSL，举个例子，比如你写下了如下的PHP代码：

<?php
echo "Hello World";
$a = 1 + 1;
echo $a;
?>

PHP执行这段代码会经过如下4个步骤(确切的来说，应该是PHP的语言引擎Zend)

1.Scanning(Lexing) ,将PHP代码转换为语言片段(Tokens)
2.Parsing, 将Tokens转换成简单而有意义的表达式
3.Compilation, 将表达式编译成Opocdes
4.Execution, 顺次执行Opcodes，每次一条，从而实现PHP脚本的功能。

题外话:现在有的Cache比如APC,可以使得PHP缓存住Opcodes，这样，每次有请求来临的时候，就不需要重复执行前面3步，从而能大幅的提高PHP的执行速度。
那什么是Lexing? 学过编译原理的同学都应该对编译原理中的词法分析步骤有所了解，Lex就是一个词法分析的依据表。 Zend/zend_language_scanner.c会根据Zend/zend_language_scanner.l(Lex文件),来输入的 PHP代码进行词法分析，从而得到一个一个的“词”，PHP4.2开始提供了一个函数叫token_get_all,这个函数就可以讲一段PHP代码 Scanning成Tokens；
如果用这个函数处理我们开头提到的PHP代码，将会得到如下结果:

Array
(
[0] => Array
(
[0] => 367
[1] => Array
(
[0] => 316
[1] => echo
)
[2] => Array
(
[0] => 370
[1] =>
)
[3] => Array
(
[0] => 315
[1] => "Hello World"
)
[4] => ;
[5] => Array
(
[0] => 370
[1] =>
)
[6] => =
[7] => Array
(
[0] => 370
[1] =>
)
[8] => Array
(
[0] => 305
[1] => 1
)
[9] => Array
(
[0] => 370
[1] =>
)
[10] => +
[11] => Array
(
[0] => 370
[1] =>
)
[12] => Array
(
[0] => 305
[1] => 1
)
[13] => ;
[14] => Array
(
[0] => 370
[1] =>
)
[15] => Array
(
[0] => 316
[1] => echo
)
[16] => Array
(
[0] => 370
[1] =>
)
[17] => ;
)

分析这个返回结果我们可以发现，源码中的字符串，字符，空格，都会原样返回。每个源代码中的字符，都会出现在相应的顺序处。而，其他的比如标签，操作符，语句，都会被转换成一个包含俩部分的Array: Token ID (也就是在Zend内部的改Token的对应码，比如,T_ECHO,T_STRING)，和源码中的原来的内容。
接下来，就是Parsing阶段了，Parsing首先会丢弃Tokens Array中的多于的空格，然后将剩余的Tokens转换成一个一个的简单的表达式

1.echo a constant string
2.add two numbers together
3.store the result of the prior expression to a variable
4.echo a variable

然后就改Compilation阶段了，它会把Tokens编译成一个个op_array, 每个op_arrayd包含如下5个部分：

1.Opcode数字的标识，指明了每个op_array的操作类型，比如add , echo
2.结果存放Opcode结果
3.操作数1 给Opcode的操作数
4.操作数2
5.扩展值 1个整形用来区别被重载的操作符

比如，我们的PHP代码会被Parsing成:

* ZEND_ECHO 'Hello World'
* ZEND_ADD ~0 1 1
* ZEND_ASSIGN !0 ~0
* ZEND_ECHO !0

呵呵，你可能会问了，我们的$a去那里了？
恩，这个要介绍操作数了，每个操作数都是由以下俩个部分组成：

a)op_type : 为IS_CONST, IS_TMP_VAR, IS_VAR, IS_UNUSED, or IS_CV
b)u,一个联合体，根据op_type的不同，分别用不同的类型保存了这个操作数的值(const)或者左值(var)

而对于var来说，每个var也不一样
IS_TMP_VAR, 顾名思义，这个是一个临时变量，保存一些op_array的结果，以便接下来的op_array使用，这种的操作数的u保存着一个指向变量表的一个句柄（整数），这种操作数一般用~开头，比如~0,表示变量表的0号未知的临时变量
IS_VAR 这种就是我们一般意义上的变量了,他们以$开头表示
IS_CV 表示ZE2.1/PHP5.1以后的编译器使用的一种cache机制，这种变量保存着被它引用的变量的地址，当一个变量第一次被引用的时候，就会被CV起来，以后对这个变量的引用就不需要再次去查找active符号表了，CV变量以！开头表示。
这么看来，我们的$a被优化成!0了。

补充细节的

转载自：https://blog.csdn.net/qq_34095777/article/details/78940789

首先，介绍 token_get_all()

作用：将提供的源码按PHP标记进行分割，类似与其他字符串分割函数，只是该函数指定了分割模式，你只需要传入字符串就可以。

那么，最重要的就是所谓的PHP标记了。–参见文章的最后

注意：
1.必须是有效的PHP代码字符串才会被分解
2.额外的一个知识点，字符串的四种定界符：
(1) 双引号
(2) 单引号
(3) heredoc, <<<标识符标识符; 相当于双引号，解析其中的变量
(4) nowdoc, <<<’标识符’ 标识符; 相当于单引号，不解析其中的变量

代码：

<?php

$str=<<<'EOF'
<?php
define('T_NEW_LINE', -1);function token_get_all_nl($source){
$new_tokens = array();

// Get the tokens
$tokens = token_get_all($source);

// Split newlines into their own tokens
foreach ($tokens as $token)
{
$token_name = is_array($token) ? $token[0] : null;
$token_data = is_array($token) ? $token[1] : $token;

// Do not split encapsed strings or multiline comments
if ($token_name == T_CONSTANT_ENCAPSED_STRING || substr($token_data, 0, 2) == '/*')
{
$new_tokens[] = array($token_name, $token_data);
continue;
}

// Split the data up by newlines
$split_data = preg_split('#(\r\n|\n)#', $token_data, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY);

foreach ($split_data as $data)
{
if ($data == "\r\n" || $data == "\n")
{
// This is a new line token
$new_tokens[] = array(T_NEW_LINE, $data);
}
else
{
// Add the token under the original token name
$new_tokens[] = is_array($token) ? array($token_name, $data) : $data;
}
}
}

return $new_tokens;
}

function token_name_nl($token)
{
if ($token === T_NEW_LINE)
{
return 'T_NEW_LINE';
}

return token_name($token);
}
?>
EOF;

$resArr = token_get_all($str);

print_r($resArr);
结果片段：

Array
(
[0] => Array
(
[0] => 379
[1] => <?php
[2] => 1
)

[1] => Array
(
[0] => 382
[1] =>

[2] => 1
)

[2] => Array
(
[0] => 319
[1] => define
[2] => 2
)

[3] => (
[4] => Array
(
[0] => 323
[1] => 'T_NEW_LINE'
[2] => 2
)

[5] => ,
[6] => Array
(
[0] => 382
[1] =>
[2] => 2
)

[7] => -
[8] => Array
(
[0] => 317
[1] => 1
[2] => 2
)

[9] => )
[10] => ;
[11] => Array
(
[0] => 346
[1] => function
[2] => 2
)

[12] => Array
(
[0] => 382
[1] =>
[2] => 2
)

[13] => Array
(
[0] => 319
[1] => token_get_all_nl
[2] => 2
)

[14] => (
[15] => Array
(
[0] => 320
[1] => $source
[2] => 2
)

[16] => )
[17] => {
[18] => Array
(
[0] => 382
[1] =>

[2] => 2
)

[19] => Array
(
[0] => 320
[1] => $new_tokens
[2] => 3
)

[20] => Array
(
[0] => 382
[1] =>
[2] => 3
)

[21] => =
[22] => Array
(
[0] => 382
[1] =>
[2] => 3
)

[23] => Array
(
[0] => 368
[1] => array
[2] => 3
)

[24] => (
[25] => )
[26] => ;
[27] => Array
(
[0] => 382
[1] =>

[2] => 3
)

[28] => Array
(
[0] => 377
[1] => // Get the tokens

[2] => 5
)

[29] => Array
(
[0] => 382
[1] =>
[2] => 6
)

[30] => Array
(
[0] => 320
[1] => $tokens
[2] => 6
)

[31] => Array
(
[0] => 382
[1] =>
[2] => 6
)

[32] => =
[33] => Array
(
[0] => 382
[1] =>
[2] => 6
)

[34] => Array
(
[0] => 319
[1] => token_get_all
[2] => 6
)

[35] => (
[36] => Array
(
[0] => 320
[1] => $source
[2] => 6
)

[37] => )
[38] => ;
[39] => Array
(
[0] => 382
[1] =>

[2] => 6
)

[40] => Array
(
[0] => 377
[1] => // Split newlines into their own tokens

[2] => 8
)

[41] => Array
(
[0] => 382
[1] =>
[2] => 9
)

[42] => Array
(
[0] => 334
[1] => foreach
[2] => 9
)

[43] => Array
(
[0] => 382
[1] =>
[2] => 9
)

[44] => (
[45] => Array
(
[0] => 320
[1] => $tokens
[2] => 9
)

[46] => Array
(
[0] => 382
[1] =>
[2] => 9
)

[47] => Array
(
[0] => 338
[1] => as
[2] => 9
)

[48] => Array
(
[0] => 382
[1] =>
[2] => 9
)

[49] => Array
(
[0] => 320
[1] => $token
[2] => 9
)

[50] => )
[51] => Array
(
[0] => 382
[1] =>

[2] => 9
)

[52] => {
[53] => Array
(
[0] => 382
[1] =>

[2] => 10
)

[54] => Array
(
[0] => 320
[1] => $token_name
[2] => 11
)

[55] => Array
(
[0] => 382
[1] =>
[2] => 11
)

[56] => =
[57] => Array
(
[0] => 382
[1] =>
[2] => 11
)

[58] => Array
(
[0] => 319
[1] => is_array
[2] => 11
)

[59] => (
[60] => Array
(
[0] => 320
[1] => $token
[2] => 11
)

[61] => )
[62] => Array
(
[0] => 382
[1] =>
[2] => 11
)

[63] => ?
[64] => Array
(
[0] => 382
[1] =>
[2] => 11
)

[65] => Array
(
[0] => 320
[1] => $token
[2] => 11
)

[66] => [
[67] => Array
(
[0] => 317
[1] => 0
[2] => 11
)

[68] => ]
[69] => Array
(
[0] => 382
[1] =>
[2] => 11
)

[70] => :
[71] => Array
(
[0] => 382
[1] =>
[2] => 11
)

[72] => Array
(
[0] => 319
[1] => null
[2] => 11
)

[73] => ;
[74] => Array
(
[0] => 382
[1] =>

[2] => 11
)

[75] => Array
(
[0] => 320
[1] => $token_data
[2] => 12
)

[76] => Array
(
[0] => 382
[1] =>
[2] => 12
)

[77] => =
[78] => Array
(
[0] => 382
[1] =>
[2] => 12
)

[79] => Array
(
[0] => 319
[1] => is_array
[2] => 12
)

[80] => (
[81] => Array
(
[0] => 320
[1] => $token
[2] => 12
)

[82] => )
[83] => Array
(
[0] => 382
[1] =>
[2] => 12
)

[84] => ?
[85] => Array
(
[0] => 382
[1] =>
[2] => 12
)

[86] => Array
(
[0] => 320
[1] => $token
[2] => 12
)

[87] => [
[88] => Array
(
[0] => 317
[1] => 1
[2] => 12
)

[89] => ]
[90] => Array
(
[0] => 382
[1] =>
[2] => 12
)

[91] => :
[92] => Array
(
[0] => 382
[1] =>
[2] => 12
)

[93] => Array
(
[0] => 320
[1] => $token
[2] => 12
)

[94] => ;
[95] => Array
(
[0] => 382
[1] =>

[2] => 12
)

[96] => Array
(
[0] => 377
[1] => // Do not split encapsed strings or multiline comments

[2] => 14
)

[97] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[98] => Array
(
[0] => 327
[1] => if
[2] => 15
)

[99] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[100] => (
[101] => Array
(
[0] => 320
[1] => $token_name
[2] => 15
)

[102] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[103] => Array
(
[0] => 285
[1] => ==
[2] => 15
)

[104] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[105] => Array
(
[0] => 319
[1] => T_CONSTANT_ENCAPSED_STRING
[2] => 15
)

[106] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[107] => Array
(
[0] => 283
[1] => ||
[2] => 15
)

[108] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[109] => Array
(
[0] => 319
[1] => substr
[2] => 15
)

[110] => (
[111] => Array
(
[0] => 320
[1] => $token_data
[2] => 15
)

[112] => ,
[113] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[114] => Array
(
[0] => 317
[1] => 0
[2] => 15
)

[115] => ,
[116] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[117] => Array
(
[0] => 317
[1] => 2
[2] => 15
)

[118] => )
[119] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[120] => Array
(
[0] => 285
[1] => ==
[2] => 15
)

[121] => Array
(
[0] => 382
[1] =>
[2] => 15
)

[122] => Array
(
[0] => 323
[1] => '/*'
[2] => 15
)

[123] => )
[124] => Array
(
[0] => 382
[1] =>

[2] => 15
)

[125] => {
[126] => Array
(
[0] => 382
[1] =>

[2] => 16
)

[127] => Array
(
[0] => 320
[1] => $new_tokens
[2] => 17
)

[128] => [
[129] => ]
[130] => Array
(
[0] => 382
[1] =>
[2] => 17
)

[131] => =
[132] => Array
(
[0] => 382
[1] =>
[2] => 17
)

[133] => Array
(
[0] => 368
[1] => array
[2] => 17
)

[134] => (
[135] => Array
(
[0] => 320
[1] => $token_name
[2] => 17
)

[136] => ,
[137] => Array
(
[0] => 382
[1] =>
[2] => 17
)

[138] => Array
(
[0] => 320
[1] => $token_data
[2] => 17
)

[139] => )
[140] => ;
[141] => Array
(
[0] => 382
[1] =>

[2] => 17
)

[142] => Array
(
[0] => 344
[1] => continue
[2] => 18
)

[143] => ;
[144] => Array
(
[0] => 382
[1] =>

[2] => 18
)

[145] => }
[146] => Array
(
[0] => 382
[1] =>

[2] => 19
)

[147] => Array
(
[0] => 377
[1] => // Split the data up by newlines

[2] => 21
)

[148] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[149] => Array
(
[0] => 320
[1] => $split_data
[2] => 22
)

[150] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[151] => =
[152] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[153] => Array
(
[0] => 319
[1] => preg_split
[2] => 22
)

[154] => (
[155] => Array
(
[0] => 323
[1] => '#(\r\n|\n)#'
[2] => 22
)

[156] => ,
[157] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[158] => Array
(
[0] => 320
[1] => $token_data
[2] => 22
)

[159] => ,
[160] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[161] => -
[162] => Array
(
[0] => 317
[1] => 1
[2] => 22
)

[163] => ,
[164] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[165] => Array
(
[0] => 319
[1] => PREG_SPLIT_DELIM_CAPTURE
[2] => 22
)

[166] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[167] => |
[168] => Array
(
[0] => 382
[1] =>
[2] => 22
)

[169] => Array
(
[0] => 319
[1] => PREG_SPLIT_NO_EMPTY
[2] => 22
)

[170] => )
[171] => ;
[172] => Array
(
[0] => 382
[1] =>

[2] => 22
)

[173] => Array
(
[0] => 334
[1] => foreach
[2] => 24
)

[174] => Array
(
[0] => 382
[1] =>
[2] => 24
)

[175] => (
[176] => Array
(
[0] => 320
[1] => $split_data
[2] => 24
)

[177] => Array
(
[0] => 382
[1] =>
[2] => 24
)

[178] => Array
(
[0] => 338
[1] => as
[2] => 24
)

[179] => Array
(
[0] => 382
[1] =>
[2] => 24
)

[180] => Array
(
[0] => 320
[1] => $data
[2] => 24
)

[181] => )
[182] => Array
(
[0] => 382
[1] =>

[2] => 24
)

[183] => {
[184] => Array
(
[0] => 382
[1] =>

[2] => 25
)

[185] => Array
(
[0] => 327
[1] => if
[2] => 26
)

[186] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[187] => (
[188] => Array
(
[0] => 320
[1] => $data
[2] => 26
)

[189] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[190] => Array
(
[0] => 285
[1] => ==
[2] => 26
)

[191] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[192] => Array
(
[0] => 323
[1] => "\r\n"
[2] => 26
)

[193] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[194] => Array
(
[0] => 283
[1] => ||
[2] => 26
)

[195] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[196] => Array
(
[0] => 320
[1] => $data
[2] => 26
)

[197] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[198] => Array
(
[0] => 285
[1] => ==
[2] => 26
)

[199] => Array
(
[0] => 382
[1] =>
[2] => 26
)

[200] => Array
(
[0] => 323
[1] => "\n"
[2] => 26
)

[201] => )
[202] => Array
(
[0] => 382
[1] =>

[2] => 26
)

[203] => {
[204] => Array
(
[0] => 382
[1] =>

[2] => 27
)

[205] => Array
(
[0] => 377
[1] => // This is a new line token

[2] => 28
)

[206] => Array
(
[0] => 382
[1] =>
[2] => 29
)

[207] => Array
(
[0] => 320
[1] => $new_tokens
[2] => 29
)

[208] => [
[209] => ]
[210] => Array
(
[0] => 382
[1] =>
[2] => 29
)

[211] => =
[212] => Array
(
[0] => 382
[1] =>
[2] => 29
)

[213] => Array
(
[0] => 368
[1] => array
[2] => 29
)

[214] => (
[215] => Array
(
[0] => 319
[1] => T_NEW_LINE
[2] => 29
)

[216] => ,
[217] => Array
(
[0] => 382
[1] =>
[2] => 29
)

[218] => Array
(
[0] => 320
[1] => $data
[2] => 29
)

[219] => )
[220] => ;
[221] => Array
(
[0] => 382
[1] =>

[2] => 29
)

[222] => }
[223] => Array
(
[0] => 382
[1] =>

[2] => 30
)

[224] => Array
(
[0] => 309
[1] => else
[2] => 31
)

[225] => Array
(
[0] => 382
[1] =>

[2] => 31
)

[226] => {
[227] => Array
(
[0] => 382
[1] =>

[2] => 32
)

[228] => Array
(
[0] => 377
[1] => // Add the token under the original token name

[2] => 33
)

[229] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[230] => Array
(
[0] => 320
[1] => $new_tokens
[2] => 34
)

[231] => [
[232] => ]
[233] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[234] => =
[235] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[236] => Array
(
[0] => 319
[1] => is_array
[2] => 34
)

[237] => (
[238] => Array
(
[0] => 320
[1] => $token
[2] => 34
)

[239] => )
[240] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[241] => ?
[242] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[243] => Array
(
[0] => 368
[1] => array
[2] => 34
)

[244] => (
[245] => Array
(
[0] => 320
[1] => $token_name
[2] => 34
)

[246] => ,
[247] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[248] => Array
(
[0] => 320
[1] => $data
[2] => 34
)

[249] => )
[250] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[251] => :
[252] => Array
(
[0] => 382
[1] =>
[2] => 34
)

[253] => Array
(
[0] => 320
[1] => $data
[2] => 34
)

[254] => ;
[255] => Array
(
[0] => 382
[1] =>

[2] => 34
)

[256] => }
[257] => Array
(
[0] => 382
[1] =>

[2] => 35
)

[258] => }
[259] => Array
(
[0] => 382
[1] =>

[2] => 36
)

[260] => }
[261] => Array
(
[0] => 382
[1] =>

[2] => 37
)

[262] => Array
(
[0] => 348
[1] => return
[2] => 39
)

[263] => Array
(
[0] => 382
[1] =>
[2] => 39
)

[264] => Array
(
[0] => 320
[1] => $new_tokens
[2] => 39
)

[265] => ;
[266] => Array
(
[0] => 382
[1] =>

[2] => 39
)

[267] => }
[268] => Array
(
[0] => 382
[1] =>

[2] => 40
)

[269] => Array
(
[0] => 346
[1] => function
[2] => 42
)

[270] => Array
(
[0] => 382
[1] =>
[2] => 42
)

[271] => Array
(
[0] => 319
[1] => token_name_nl
[2] => 42
)

[272] => (
[273] => Array
(
[0] => 320
[1] => $token
[2] => 42
)

[274] => )
[275] => Array
(
[0] => 382
[1] =>

[2] => 42
)

[276] => {
[277] => Array
(
[0] => 382
[1] =>

[2] => 43
)

[278] => Array
(
[0] => 327
[1] => if
[2] => 44
)

[279] => Array
(
[0] => 382
[1] =>
[2] => 44
)

[280] => (
[281] => Array
(
[0] => 320
[1] => $token
[2] => 44
)

[282] => Array
(
[0] => 382
[1] =>
[2] => 44
)

[283] => Array
(
[0] => 287
[1] => ===
[2] => 44
)

[284] => Array
(
[0] => 382
[1] =>
[2] => 44
)

[285] => Array
(
[0] => 319
[1] => T_NEW_LINE
[2] => 44
)

[286] => )
[287] => Array
(
[0] => 382
[1] =>

[2] => 44
)

[288] => {
[289] => Array
(
[0] => 382
[1] =>

[2] => 45
)

[290] => Array
(
[0] => 348
[1] => return
[2] => 46
)

[291] => Array
(
[0] => 382
[1] =>
[2] => 46
)

[292] => Array
(
[0] => 323
[1] => 'T_NEW_LINE'
[2] => 46
)

[293] => ;
[294] => Array
(
[0] => 382
[1] =>

[2] => 46
)

[295] => }
[296] => Array
(
[0] => 382
[1] =>

[2] => 47
)

[297] => Array
(
[0] => 348
[1] => return
[2] => 49
)

[298] => Array
(
[0] => 382
[1] =>
[2] => 49
)

[299] => Array
(
[0] => 319
[1] => token_name
[2] => 49
)

[300] => (
[301] => Array
(
[0] => 320
[1] => $token
[2] => 49
)

[302] => )
[303] => ;
[304] => Array
(
[0] => 382
[1] =>

[2] => 49
)

[305] => }
[306] => Array
(
[0] => 382
[1] =>

[2] => 50
)

[307] => Array
(
[0] => 381
[1] => ?>
[2] => 51
)

)

通过结果可以看到，得到一个二维数组

$resArr[0][0] 数字代号，需要使用token_name转换为下面表格中的代号
比如：
token_name(308); // T_STRING

$resArr[0][1] 分割的片段

代号   语法   参考
T_ABSTRACT   abstract   抽象类（自 PHP 5.0.0 起可用）
T_AND_EQUAL   &=   赋值运算符
T_ARRAY   array()   array()，数组语法
T_ARRAY_CAST   (array)类型转换
T_AS   as   foreach
T_BAD_CHARACTER       在 ASCII 32 之前除了 \t (0x09), \n (0x0a) 和 \r (0x0d) 以外的任何字符
T_BOOLEAN_AND   &&   逻辑运算符
T_BOOLEAN_OR   |   逻辑运算符
T_BOOL_CAST       (bool) 或 (boolean) 类型转换
T_BREAK   break   break
T_CASE   case   switch
T_CATCH   catch   异常处理（自 PHP 5.0.0 起可用）
T_CHARACTER
T_CLASS   class   类和对象
T_CLONE   clone   类和对象。仅用于 PHP 5。
T_CLOSE_TAG   ?> 或 %>
T_COMMENT   // 或 #   以及 PHP 5 下的 /* */ 注释
T_CONCAT_EQUAL   .=   赋值运算符
T_CONST   const
T_CONSTANT_ENCAPSED_STRING   “foo” 或 ‘bar’字符串语法
T_CONTINUE   continue
T_CURLY_OPEN
T_DEC   - -   递增/递减运算符
T_DECLARE   declare   declare
T_DEFAULT   default   switch
T_DIV_EQUAL   /=   赋值运算符
T_DNUMBER   0.12，等等浮点数
T_DOC_COMMENT   /** */   PHPDoc 风格的注释（仅用于 PHP 5）
T_DO   do   do..while
T_DOLLAR_OPEN_CURLY_BRACES   ${   复杂变量解析语法
T_DOUBLE_ARROW   =>   数组语法
T_DOUBLE_CAST   (real),(double),(float)   类型转换
T_DOUBLE_COLON   ::   见下面的 T_PAAMAYIM_NEKUDOTAYIM
T_ECHO   echo   echo
T_ELSE   else   else
T_ELSEIF   elseif   elseif
T_EMPTY   empty   empty()
T_ENCAPSED_AND_WHITESPACE
T_ENDDECLARE   enddeclare   declare，替代语法
T_ENDFOR   endfor   for，替代语法
T_ENDFOREACH   endforeach   foreach，替代语法
T_ENDIF   endif   if，替代语法
T_ENDSWITCH   endswitch   switch，替代语法
T_ENDWHILE   endwhile   while，替代语法
T_END_HEREDOC   heredoc   语法
T_EVAL   eval()   eval()
T_EXIT   exit 或 die   exit(), die()
T_EXTENDS   extends   extends，类和对象
T_FILE   __FILE__   常量
T_FINAL   final   Final 关键字（自 PHP 5.0.0 起可用）
T_FOR   for   for
T_FOREACH   foreach   foreach
T_FUNCTION   function 或 cfunction   函数
T_GLOBAL   global   变量范围
T_HALT_COMPILER   __halt_compiler()   __halt_compiler（自 PHP 5.1.0 起可用）
T_IF   if   if
T_IMPLEMENTS   implements   对象接口（自 PHP 5.0.0 起可用）
T_INC   ++   递增／递减运算符
T_INCLUDE   include()   include
T_INCLUDE_ONCE   include_once()   include_once
T_INLINE_HTML
T_INSTANCEOF   instanceof   type operators（仅 PHP5）
T_INT_CAST   (int) 或 (integer)   类型转换
T_INTERFACE   interface   对象接口（自 PHP 5.0.0 起可用）
T_ISSET   isset()   isset()
T_IS_EQUAL   ==   比较运算符
T_IS_GREATER_OR_EQUAL   >=   比较运算符
T_IS_IDENTICAL   ===   比较运算符
T_IS_NOT_EQUAL   != or <>   比较运算符
T_IS_NOT_IDENTICAL   !==   比较运算符
T_IS_SMALLER_OR_EQUAL   <=   比较运算符
T_LINE   __LINE__   常量
T_LIST   list()   list()
T_LNUMBER   123，012，0x1ac，等等   整型
T_LOGICAL_AND   and   逻辑运算符
T_LOGICAL_OR   or   逻辑运算符
T_LOGICAL_XOR   xor   逻辑运算符
T_MINUS_EQUAL   -=   赋值运算符
T_ML_COMMENT   /* 和 */   注释（仅用于 PHP 4）
T_MOD_EQUAL   %=   赋值运算符
T_MUL_EQUAL   *=   赋值运算符
T_NEW new 类和对象
T_NUM_STRING
T_OBJECT_CAST   (object)   类型转换
T_OBJECT_OPERATOR   ->   类和对象
T_OPEN_TAG   <?php, <? 或 <%   从 HTML 转义
T_OPEN_TAG_WITH_ECHO   <?= or <%=   从 HTML 转义
T_OR_EQUAL   =   赋值运算符
T_PAAMAYIM_NEKUDOTAYIM   ::   ::也定义为 T_DOUBLE_COLON。
T_PLUS_EQUAL   +=   赋值运算符
T_PRINT   print()   print
T_PRIVATE   private   类和对象。仅用于 PHP 5。
T_PUBLIC   public   类和对象。仅用于 PHP 5。
T_PROTECTED   protected   类和对象。仅用于 PHP 5。
T_REQUIRE   require()   require
T_REQUIRE_ONCE   require_once()   require_once
T_RETURN   return   返回值
T_SL   <<   位运算符
T_SL_EQUAL   <<=   赋值运算符
T_SR   >>   位运算符
T_SR_EQUAL   >>=   赋值运算符
T_START_HEREDOC   <<<   heredoc 语法
T_STATIC   static   变量范围
T_STRING
T_STRING_CAST   (string)   类型转换
T_STRING_VARNAME
T_SWITCH   switch   switch
T_THROW   throw   异常处理（自 PHP 5.0.0 起可用）
T_TRY   try   异常处理（自 PHP 5.0.0 起可用）
T_UNSET   unset()   unset()
T_UNSET_CAST   (unset)   （无文档; 类型为 NULL）
T_USE   use   （未实现）
T_VAR   var   类和对象
T_VARIABLE   $foo   变量
T_WHILE   while   while，do..while
T_WHITESPACE
T_XOR_EQUAL   ^=   赋值运算符
T_FUNC_C   __FUNCTION__   常量，自 PHP 4.3.0 起可用
T_CLASS_C   __CLASS__   常量，自 PHP 4.3.0 起可用

alashan007

关注

0
点赞
踩
0

收藏

觉得还不错? 一键收藏
0
评论
深入理解PHP原理之Opcodes

最近要给Yahoo的同事们做一个关于PHP和Apache处理请求的内部机制的讲座，刚好写了些关于Opcodes的文字，就发上来了，这个文章基于 Sara Golemon大师的《Understanding OPcode》Opcode是一种PHP脚本编译后的中间语言，就像Java的ByteCode,或者.NET的MSL，举个例子，比如你写下了如下的PHP代码：<?php echo "Hello World"; $a = 1 + 1; echo $a; ?>...
复制链接

扫一扫

专栏目录