Regex matching URL with www and no consecutive dots

Question

Can you help me with regex?

I have line

"Sites www.google.com и www.ridd.rdd..com good."

After parse I'v get this type of line:

"Sites http://www.google.com и www.ridd.rdd..com good."

Problem with checking consecutive points. To sites with an error (with two points in a row) "http//:" should not be appended.

My regex:

 Matcher matchr = Pattern.compile("w{3}(\\.\\w+)+[a-z]{2,6}").matcher(text);

        while (matchr.find()) {
            text = text.replace(matchr.group(0), "http://" + matchr.group(0));
        }

        System.out.println(text);

http://stackoverflow.com/questions/27745/getting-parts-of-a-url-regex — light_ray, Apr 06 '16 at 08:58
So are you trying to add "http://" to the start of all the url strings? — jazibobs, Apr 06 '16 at 08:58
See [this demo](http://ideone.com/STFdHe), please explain what output is expected if it does not work for you. — Wiktor Stribiżew, Apr 06 '16 at 09:01
@WiktorStribiżew to sites with an error (with two points in a row) should not append the "http"//" — Yahor Urbanovich, Apr 06 '16 at 09:03
How can you distinguish between `"Sites www.google.com и www.ridd.rdd..com good."` and `"Sites www.ridd.rdd..com и www.google.com."`? What are URL boundaries? BTW, try [`text.replaceAll("\\b(?!\\S*\\.{2})w{3}(\\.+\\w+)+[a-z]{2,6}\\b", "http://$0")`](http://ideone.com/vdVVzv). — Wiktor Stribiżew, Apr 06 '16 at 09:09
Before regex: "Sites www.google.com и www.ridd.rdd..com good." After: "Sites http://www.google.com и www.ridd.rdd..com good."; http added only www.[site].[prefix] — Yahor Urbanovich, Apr 06 '16 at 09:11
The question got closed with [*Getting parts of a URL (Regex)*](http://stackoverflow.com/questions/27745/getting-parts-of-a-url-regex) that is totally unrelated to the current problem. — Wiktor Stribiżew, Apr 06 '16 at 09:15

score 1 · Answer 1 · answered Apr 06 '16 at 09:23

Your regex w{3}(\\.\\w+)+[a-z]{2,6} matches a part of the second bad "URL", www.ridd.rdd..com. So, you need to make sure the substring you match has no consecutive dots. You may use word boundaries and a negative lookahead (?!\S*\.{2}).

Use

String text = "Sites www.google.com и www.ridd.rdd..com good.";
text = text.replaceAll("\\b(?!\\S*\\.{2})w{3}(\\.\\w+)+[a-z]{2,6}\\b", "http://$0");
// => Sites http://www.google.com и www.ridd.rdd..com good.

See the IDEONE demo

Pattern explanation:

\\b - leading word boundary
(?!\\S*\\.{2}) - there should not be any consecutive dots in the non-whitespace chunk to follow
w{3} - match www
(\\.\\w+)+ - 1+ sequences of . followed with 1+ alphanumeric or underscore characters
[a-z]{2,6} - make sure there are 2 to 6 a-z letters...
\\b - at the end of this "word"

Glad to help, please consider accepting the answer since it worked for you. — Wiktor Stribiżew, Apr 06 '16 at 09:26

Regex matching URL with www and no consecutive dots

1 Answers1