Get the second level domain of an URL (java)

Question

I am wondering if there is a parser or library in java for extracting the second level domain (SLD) in an URL - or failing that an algo or regex for doing the same. For example:

URI uri = new URI("http://www.mydomain.ltd.uk/blah/some/page.html");

String host = uri.getHost();

System.out.println(host);

which prints:

mydomain.ltd.uk

Now what I'd like to do is robustly identify the SLD ("ltd.uk") component. Any ideas?

Edit: I'm ideally looking for a general solution, so I'd match ".uk" in "police.uk", ".co.uk" in "bbc.co.uk" and ".com" in "amazon.com".

Thanks

Answer 1

After reeading everything here, the correct solution should be (with guava)

InternetDomainName.from(uriHost).topPrivateDomain().toString();

errors when using Guava to get the private domain name

Answer 2

Don't know your purpose but Second-Level Domain may not mean much to you. You probably need to find public suffix and the domain right below it is what you are looking for.

Apache Http Component (HttpClient 4) comes with classes to handle this,

org.apache.http.impl.cookie.PublicSuffixFilter
org.apache.http.impl.cookie.PublicSuffixListParser

You need to download the public suffix list from here,

http://mxr.mozilla.org/mozilla-central/source/netwerk/dns/effective_tld_names.dat?raw=1

Answer 3

After looking at these answers and not being satisfied by them I used the class com.google.common.net.InternetDomainName to subtract the public parts of a domain name from all the parts:

Set<String> nonePublicDomainParts(String uriHost) {
    InternetDomainName fullDomainName = InternetDomainName.from(uriHost);
    InternetDomainName publicDomainName = fullDomainName.publicSuffix();
    Set<String> nonePublicParts = new HashSet<String>(fullDomainName.parts());
    nonePublicParts.removeAll(publicDomainName.parts());
    return nonePublicParts;
}

That class is on maven in the guava library:

    <dependency>
        <groupId>com.google.guava</groupId>
        <artifactId>guava</artifactId>
        <version>10.0.1</version>
        <scope>compile</scope>
    </dependency>

Internally this class is using a TldPatterns.class which is package private and has the list of top level domains baked into it.

Interestingly, if you look at that classes source at the link below it explicitly lists "police.uk" as a private domain name. This is correct as police.uk is a private domain controlled by the police; else criminals.police.uk will be emailing you asking for your credit card details in relation to their ongoing investigations into card fraud ;)

http://code.google.com/p/guava-libraries/source/browse/guava/src/com/google/common/net/TldPatterns.java?spec=svn8c3cc7e67132f8dcaae4bd214736a8ddf6611769&r=8c3cc7e67132f8dcaae4bd214736a8ddf6611769