Sometimes we need to calculate the number of words contained in a string. For pure English strings, The number of words is equal to the length of the string, use
The strlen function can be used to obtain it, but what if the string contains Chinese characters? mb_strlen can be implemented, but unfortunately the extension is not installed, so implement it yourself.
php has an extension that is generally required. We can use mb_strlen to get the number of words in a string. The usage is generally as follows:
$len = mb_strlen("你是我的小苹果","utf-8");
Get the string length as expected: 7.
What if the mb extension is not installed? Implement it yourself.
We must first understand the fact that strings are composed of characters, and characters are represented by bytes. Each English character is a byte, corresponding to an ASCII code. The ASCII code of an English character is less than 128 , which is 0x80 in hexadecimal. When the ascii code of a byte exceeds 127, it means that the current byte is not a complete character.
For example
$str = "你是我的小苹果";
can get the first byte, let’s take a look at what it is:
php > $str = "你是我的小苹果"; php > echo $str{0}; �
is gibberish, it's just
Copy the code The code is as follows: You
One of the bytes of a word, that is,
Copy the code The code is as follows: You
This character is composed of more than one byte, let’s try this:
php > echo $str{0}.$str{1}.$str{2};
You
As you can see, if the three bytes are connected together and output, it becomes a complete
Copy the code The code is as follows: You
.
As for why there are three bytes here instead of two or four? This depends on the encoding of the string. My console here is utf8 encoded by default. In PHP, a utf8 character is expressed in three bytes. If it is gbk encoded, it will be two bytes. As for the relationship between encoding and bytes, this topic is relatively large and cannot be covered in one article. Please refer to this article: Character encoding notes: ascii, unicode and utf8.
Knowing this, we can write a word count check function ourselves. The general process is as follows:
1. A for loop traverses the bytes 2. Determine whether the byte encoding is >= 0x80, if so, skip N bytes
I wrote a simple function that can determine the length of gbk or utf8 string, for reference only:
<?php function mbstrlen($str,$encoding="utf8") { if (($len = strlen($str)) == 0) { return 0; } $encoding = strtolower($encoding); if ($encoding == "utf8" or $encoding == "utf-8") { $step = 3; } elseif ($encoding == "gbk" or $encoding == "gb2312") { $step = 2; } else { return false; } $count = 0; for ($i=0; $i<$len; $i++) { $count++; //如果字节码大于127,则根据编码跳几个字节 if (ord($str{$i}) >= 0x80) { $i = $i + $step - 1;//之所以减去1,因为for循环本身还要$i++ } } return $count; } echo mbstrlen(iconv("utf-8","gbk","你是我的小苹果"),"gbk"); echo mbstrlen("你是我的小苹果");