之前两天在极客头条上看了一篇帖子，最严谨的校验email地址的正则表达式 ，最近接手的最多的就是yii2.0框架，所以很好奇想看看yii2.0是怎么实现验证规则。
首先，一般在自定义的继承model类中，会有个rules()方法：
public function rules()
{
/*echo "<pre>";
$e=new \Exception();
var_export($e->getTraceAsString());*/
return [
// username and password are both required
[['username', 'password'], 'required'],
// password is validated by validatePassword()
['password', 'validatePassword'],
];
}
通过上面的调用栈信息，我们看到，对于post提交的表单，我们在使用load()加载的时候会调用rules()方法，然后我们再进入load()里面看，load()方法只给属性赋值，不涉及到任何对属性的判断，通过简单的验证就可以看出来，这里就不截图了，所以对属性的判断并不在load()中！
我们获取错误信息是通过getErrors()，那就看下addErrors()的调用栈信息，如下：
里面在调用栈中出现了validateAttribute()方法，
从这次的栈信息中，可以看出来，validateAttribute是对rules里面定义的规则，每一个执行一次，validateAttribute()里面调用了validateValue()方法
public function validateAttribute($model, $attribute)
{
$result = $this->validateValue($model->$attribute);
if (!empty($result)) {
$this->addError($model, $attribute, $result[0], $result[1]);
}
}
所以这里的$this拿到的就是RequiredValidator的句柄了，所以所有对rules的规则的验证都是在validators目录下。
----------------------------------------------------------------------------------下面是对email的验证和php正则---------------------------------------------------------------------------------------------------
文件路径：\yii\validators\EmailValidator
protected function validateValue($value)
{
if (!is_string($value)) {
$valid = false;
} elseif (!preg_match('/^(?P<name>(?:"?([^"]*)"?\s)?)(?:\s+)?(?:(?P<open><?)((?P<local>.+)@(?P<domain>[^>]+))(?P<close>>?))$/i', $value, $matches)) {
$valid = false;
} else {
if ($this->enableIDN) {
$matches['local'] = idn_to_ascii($matches['local']);
$matches['domain'] = idn_to_ascii($matches['domain']);
$value = $matches['name'] . $matches['open'] . $matches['local'] . '@' . $matches['domain'] . $matches['close'];
}
if (strlen($matches['local']) > 64) {
// The maximum total length of a user name or other local-part is 64 octets. RFC 5322 section 4.5.3.1.1
// http://tools.ietf.org/html/rfc5321#section-4.5.3.1.1
$valid = false;
} elseif (strlen($matches['local'] . '@' . $matches['domain']) > 254) {
// There is a restriction in RFC 2821 on the length of an address in MAIL and RCPT commands
// of 254 characters. Since addresses that do not fit in those fields are not normally useful, the
// upper limit on address lengths should normally be considered to be 254.
//
// Dominic Sayers, RFC 3696 erratum 1690
// http://www.rfc-editor.org/errata_search.php?eid=1690
$valid = false;
} else {
$valid = preg_match($this->pattern, $value) || $this->allowName && preg_match($this->fullPattern, $value);
if ($valid && $this->checkDNS) {
$valid = checkdnsrr($matches['domain'], 'MX') || checkdnsrr($matches['domain'], 'A');
}
}
}
return $valid ? null : [$this->message, []];
}所以yii对email的验证使用的也就是这样一段正则
/^(?P<name>(?:"?([^"]*)"?\s)?)(?:\s+)?(?:(?P<open><?)((?P<local>.+)@(?P<domain>[^>]+))(?P<close>>?))$/i $value="1 <[email protected]>";
preg_match('/^(?P<name>(?:"?([^"]*)"?\s)?)(?:\s+)?(?:(?P<open><?)((?P<local>.+)@(?P<domain>[^>]+))(?P<close>>?))$/i', $value, $matches);
var_dump($matches);
1(name): (?P<name>(?:"?([^"]*)"?\s)?) 匹配空格和空格前的字符
2:([^"]*) 匹配除了"之外的任意字符0次或多次
3(open): (?P<open><?) 匹配<
4: ((?P<local>.+)@(?P<domain>[^>]+))(?P<close>>?))
5(local): (?P<local>.+) 匹配任意字符1次到多次
6(domain): (?P<domain>[^>]+) 匹配除了>外的字符1次到多次
7: (?P<close>>?) 匹配>标签
idn_to_ascii 将domain name转换成ASCII形式
补充一些php正则表达式的知识：
chr()将ASCII值转换成字符，
hexdec()把十六进制转换成十进制，
ord()返回字符串的首个字符的ASCII值，
双引号包围十六进制的ASCII码echo输出的时候会转换成字符。
//echo ord($b);
//echo chr('\x7E');
//echo hexdec('\x5C');
(?:pattern)匹配pattern但不获取结果， (pattern)匹配pattern并获取匹配结果，(?:pattern)匹配pattern但不获取匹配结果，(?=pattern)正向预查，(?!pattern)反向预查
预查不消耗字符，也就是a(?=b|c)匹配了ab或ac之后，下次匹配的时候还可以从b或c处开始匹配
var_dump(preg_replace('/^(?:a(a|b))?/','xxx',"aaab"));//xxxab 3个x
var_dump(preg_replace('/^(?:a(?=a|b))?/','xxx',"aaab"));//xxxaab 3个x
var_dump(preg_replace('/(?:a(?=a|b))/','xxx',"aaab"));//xxxxxxxxxb 9个x
var_dump(preg_replace('/(?:a(a|b))?/','xxx',"aaab"));//xxxxxxxxx 9个x
var_dump(preg_replace('/(?:a(?!a|c))/','xxx','aaab'));//aaxxxb
var_dump(preg_replace('/(?!a|c)/','xxx','aaab'));//aaaxxxbxxx