How to use regex to get things out of hostname

http://www.exampledepot.com/egs/java.lang/HasSubstr.html

just check

if (string.indexOf("abc.com") != -1 }|| string.indexOf("abc.com") != -1 )
{
   //string is containing required substrings
}

Something like this works:
 

String[] t = {"www.abc.com", 
            "www.abc.com.cn", 
            "www.xyz.abc.com", 
            "www.xyz.abc.com.cn",
            "xyz.abc.com", 
            "xyz.abc.com.cn",
        };
        Pattern p = Pattern.compile("((?:[a-z0-9][-a-z0-9]*[a-z0-9]|[a-z0-9])"
                + "(?:\\.com\\.cn|\\.com)$)");
        for (String s : t) {
            Matcher m = p.matcher(s);
            if (m.find()) {
                System.out.println("Found " + m.group(1) + " in " + s);
            } else {
                System.out.println("Not found in " + s);
            }
        }

Select allOpen in new window

Looks a bit ugly because I couldn't find a more elegant way to enforce the "host name can't end or start with a hyphen" rule.

If you need to match more domains, not just ".com.cn" and ".com", then the second part should contain more complicated alternatives, but the idea stays the same.

Try this code

public class TestSubstring {
public static void main(String[] args) {
	String[] string = {"www.abc.com","www.abc.com.cn","www.xyz.abc.com","www.xyz.abc.com.cn","xyz.abc.com","xyz.abc.com.cn"};
	for (String stg: string){
		System.out.println(stg.substring(0, stg.indexOf(".abc.com")));
	}

}
}

Select allOpen in new window

        String [] hosts = {
"www.abc.com",
"www.abc.com.cn",
"www.xyz.abc.com",
"www.xyz.abc.com.cn",
"xyz.abc.com",
"xyz.abc.com.cn"
};

        for(String sh : hosts){

            sh = sh.replaceAll(".*\\.(.+?\\.com)","$1");
            System.out.println("result: " + sh);

        }

Select allOpen in new window

Output:

result: abc.com
result: abc.com.cn
result: abc.com
result: abc.com.cn
result: abc.com
result: abc.com.cn

Select allOpen in new window

@for_yan, this doesn't work for something like "com.cn.com.cn".

 why?  it returns:
result: cn.com.cn

com.cn.com

returns

cn.com

that is what is expected, as I understand.

And certainly for any regex you  can invent
some strign which will break it.

Sorry, I was wrong, it actually works.

No problem.
Though nothing is ideal, I'm sure there is some string which will break it.
Still it helps in great majority of cases

Personally i would use URL.getHost