正则零宽断言 Regex.Replace(string,string,string)

今天终于明白了

Regex.Replace(string input,string pattern,string replacement)

(1)

input: http://www.xxx.com/news/5/

pattern:.*?xxx\.com/news/\d+/

replacement:news.aspx

结果news.aspx

因为input按照pattern匹配的结果为http://www.xxx.com/news/5/,然后我们用news.aspx替换

最终结果就是news.aspx

(2)

input:  abcedft http://www.xxx.com/news/5/6/7YYYYY

pattern: http://(.*/?)\.*xxx.com/news/(\d+/)*(\d+)/*
replacement: /news.aspx
结果为 abcedft/news.aspxYYYYY
因为input 中匹配的为 http://www.xxx.com/news/5/6/7

(3)

input:http://www.xxx.com/news/5/

pattern:http://(.*?)\.xxx.com/news/(\d+)/

replacement:news.aspx?id=$2

结果为:news.aspx?id=5

因为input按照pattern匹配的结果为http://www.xxx.com/news/5/,同时把(.*?)和(\d+)捕获到组里

然后用replacement来替换，$2代表的是获取第2个捕获组的值

零宽断言

(?<=exp)匹配exp后面的位置

(?=exp)匹配exp前面的位置

例如：

(?<=^|>)[^<>]+(?=<|$)

上面正则的意思就是匹配以[^<>]+这个开头或[^<>]+前面是>，同时[^<>]+后面是<或以[^<>]+结果的值

例子：

string test = "abc def <img src=\"124.jpg\"> ghi";
Regex reg = new Regex(@"(?<=^|>)[^<>]+(?=<|$)");
string result = reg.Replace(test, delegate(Match m) { return m.Value.Replace(" ", "&nbsp;").Replace("\n", "<br>"); });
richTextBox2.Text = result;