This project contains a Ragel language file that provides a robust URL Java parser (or at least, it aims to be!). See: http://tools.ietf.org/html/rfc3987
You need Ragel. You can get your platform binaries from the website URL or build from the source using gcc.
To manually generate the URL Java parser, change directory at the location of the URLParser.rl file and execute the following command-line (whereas the ragel executable should be available in your environment PATH variable):
ragel -J URLParser.rl
To generate the URL Java parser with Maven, you will need version 3.0.4 at minimum (Ragel compilation does not work on lower versions). Execute the following command-line at the project's root directory:
mvn compile
The generated class should be located in the target/generated-sources/ragel/org/caron/ragel/url directory.
String url = ...;
Url parsedUrl = urlParser.parse(url);
System.out.println(parsedUrl);
To execute the tests and verify the coverage of this grammar, you can dig into the UrlParserTest class and execute it with Maven. To get the XML reports:
mvn test
To get the HTML reports, you need to get the Maven site phase executed (due to this bug):
mvn site
If you dont need a robust implementation, this regexp might be enough:
String regex = "^(([^:/?#.]+):)?(//)?(([^:/]*)?(\\:([^/]*))?\\@)?(([^/:]+)|\\[[^/\\]]+\\])?(:(\\d*))?(/[^?#]*)(\\?([^#]*))?(#(.*))?";
Pattern p = Pattern.compile(regex);
Matcher m = p.matcher(INPUT); // get a matcher object
Whereas each capturing groups would correspond to:
$2 = scheme
$5 = username
$7 = password
$8 = server
$11 = port
$12 = path
$14 = query
$16 = fragment
Unit-tests for this regular expression is provided, if you want to verify the support coverage or add your own tests.