How to match repeated patterns?

I would like to match:

some.name.separa开发者_如何学编程ted.by.dots

But I don't have any idea how.

I can match a single part like this

 \w+\.

How can I say "repeat that"

Try the following:

\w+(?:\.\w+)+

The + after (?: ... ) tell it to match what is inside the parenthesis one or more times.

Note that \w only matches ASCII characters, so a word like café wouldn't be matches by \w+, let alone words/text containing Unicode.

EDIT

The difference between [...] and (?:...) is that [...] always matches a single character. It is called a "character set" or "character class". So, [abc] does not match the string "abc", but matches one of the characters a, b or c.

The fact that \w+[\.\w+]* also matches your string is because [\.\w+] matches a . or a character from \w, which is then repeated zero or more time by the * after it. But, \w+[\.\w+]* will therefor also match strings like aaaaa or aaa............

The (?:...) is, as I already mentioned, simply used to group characters (and possible repeat those groups).

More info on character sets: http://www.regular-expressions.info/charclass.html

More info on groups: http://www.regular-expressions.info/brackets.html

EDIT II

Here's an example in Java (seeing you post mostly Java answers):

import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class Main {
    public static void main(String[] args) {
        String text = "some.text.here only but not Some other " + 
                "there some.name.separated.by.dots and.we are done!";
        Pattern p = Pattern.compile("\\w+(?:\\.\\w+)+");
        Matcher m = p.matcher(text);
        while(m.find()) {
            System.out.println(m.group());
        }
    }
}

which will produce:

some.text.here
some.name.separated.by.dots
and.we

Note that m.group(0) and m.group() are equivalent: meaning "the entire match".

This will also work:

(\w+(\.|$))+

You can use ? to match 0 or 1 of the preceeding parts, * to match 0 to any amount of the preceeding parts, and + to match at least one of the preceeding parts.

So (\w\.)? will match w. and a blank, (\w\.)* will match r.2.5.3.1.s.r.g.s. and a blank, and (\w\.)+ will match any of the above but not a blank.

If you want to match something like your example, you'll need to do (\w+\.)+, which means 'match at least one non whitespace, then a period, and match at least one of these'.

(\w+\.)+

Apparently, the body has to be at least 30 characters. I hope this is enough.