当解析一个文件时,PHP 会寻找起始和结束标记,也就是 <?php 和 ?>,这告诉 PHP 开始和停止解析二者之间的代码。此种解析方式使得 PHP 可以被嵌入到各种不同的文档中去,而任何起始和结束标记之外的部分都会被 PHP 解析器忽略。
PHP 也允许使用短标记 <? 和 ?>,但不鼓励使用。只有通过激活 php.ini 中的 short_open_tag 配置指令或者在编译 PHP 时使用了配置选项 --enable-short-tags 时才能使用短标记。
如果文件内容是纯 PHP 代码,最好在文件末尾删除 PHP 结束标记。这可以避免在 PHP 结束标记之后万一意外加入了空格或者换行符,会导致 PHP 开始输出这些空白,而脚本中此时并无输出的意图。
<?php
echo "Hello world";
// ... more code
echo "Last statement";
// 脚本至此结束,并无 PHP 结束标记
<?php[whitespace]", not just "<?php"
This is absolutely correct, but the wording may confuse some developers less familiar with the extent of the term "[whitespace]".
Whitespace, in this context, would be any character that generated vertical or horizontal space, including tabs ( \t ), newlines ( \n ), and carriage returns ( \r ), as well as a space character ( \s ). So reusing purkrt's example:
<?php/*blah*/ echo "a"?>
would not work, as mentioned, but :
<?php /*php followed by space*/ echo "a"?>
will work, as well as :
<?php
/*php followed by end-of-line*/ echo "a"?>
and :
<?php
/*php followed by tab*/ echo "a"
?>
I just wanted to clarify this to prevent anyone from misreading purkrt's note to mean that a the opening tag --even when being on its own line--required a space ( \s ) character. The following would work but is not at all necessary or how the earlier comment should be interpreted :
<?php
/*php followed by a space and end-of-line*/ echo "a"
?>
<?php
// Code will end here ?> This is output as literal text.
<?php
# Same with this method of commenting ?> This is output as literal text.
However they do not have an effect in C-style comments:
<?php
/* Code will not end here ?> as closing tags are ignored inside C-style comments. */