Url Matching Algorithm Used By Tomcat for web.xml

[Jarryd Folino]

I am trying to set up a simple java program that can do the following:

* Takes in a ' given url' such as '/places/australia'
* Takes in a list of URL mappings such as ["*.do", *jsp, "/places", "/places/aust"]

I want to do this to investigate exactly which given URL's will match against the servlet and filter mappings given in a web.xml file. My initial guess was to use the standard Java Pattern and Matcher classes as per below:

public class Main {

	public static void main(String ... args)
	{
                String urlToTest = "/places/australia";
                for (String urlPatternFromWebXml : new String[]{"*.do", "/places/australia", "*.jsp", "/anything"}) {
                    if (match(urlToTest, urlPatternFromWebXml))
                    {
                    	System.out.println(urlToTest + " does match " + urlPatternFromWebXml);
                    }
                }


	}
	
	public static boolean match(String url, String regex)
	{
		Pattern p = Pattern.compile(regex);
		Matcher m = p.matcher(url);
		return m.find();
	}
	
}

However when I ran the code I got the following exception:

Exception in thread "main" java.util.regex.PatternSyntaxException: Dangling meta character '*' near index 0
*.do
^

This would imply that the standard java regex classes are not used for matching the URL's in web.xml.

Does anyone know of the class library tomcat or a similar container uses to determine if a given url matches the web.xml mappings?

[Frits Walraven]

However when I ran the code I got the following exception: Exception in thread "main" java.util.regex.PatternSyntaxException: Dangling meta character '*' near index 0 *.do

You get the exception because the character "*" has a special meaning in regex. If you want to match you can use ".do".

Note that when you are using regex it will look for all matches in a string and URL mapping works quite differently. Read the Servlet specs 2.4 section SRV.11.1 Use of URL Paths for more information.

Regards,
Frits